46394_FM_00i_xviii.qxd
2/26/10
2:30 PM
Page i
Epidemiology and Prevention of Cardiovascular Diseases: A Global ChaLLEnge second edition
Darwin R. Labarthe, MD, MPH, PhD
46394_FM_00i_xviii.qxd
3/5/10
2:23 PM
Page ii
World Headquarters Jones and Bartlett Publishers 40 Tall Pine Drive Sudbury, MA 01776 978-443-5000
[email protected] www.jbpub.com
Jones and Bartlett Publishers Canada 6339 Ormindale Way Mississauga, Ontario L5V 1J2 Canada
Jones and Bartlett Publishers International Barb House, Barb Mews London W6 7PA United Kingdom
Jones and Bartlett’s books and products are available through most bookstores and online booksellers. To contact Jones and Bartlett Publishers directly, call 800-832-0034, fax 978-443-8000, or visit our website, www.jbpub.com. Substantial discounts on bulk quantities of Jones and Bartlett’s publications are available to corporations, professional associations, and other qualified organizations. For details and specific discount information, contact the special sales department at Jones and Bartlett via the above contact information or send an email to
[email protected]. Copyright © 2011 by Jones and Bartlett Publishers, LLC All rights reserved. No part of the material protected by this copyright may be reproduced or utilized in any form, electronic or mechanical, including photocopying, recording, or by any information storage and retrieval system, without written permission from the copyright owner. This publication is designed to provide accurate and authoritative information in regard to the Subject Matter covered. It is sold with the understanding that the publisher is not engaged in rendering legal, accounting, or other professional service. If legal advice or other expert assistance is required, the service of a competent professional person should be sought. This book was written by Darwin R. Labarthe in his private capacity. No official support or endorsement by the Centers for Disease Control and Prevention, Department of Health and Human Services is intended, nor should be inferred. Production Credits Publisher: Michael Brown Editorial Assistant: Catie Heverling Editorial Assistant: Teresa Reilly Production Manager: Tracey Chapman Senior Marketing Manager: Sophie Fleck Manufacturing and Inventory Control Supervisor: Amy Bacus Composition: Auburn Associates, Inc. Cover Design: Kristin E. Parker Cover Image: fruit: © Daniel Gilbey/Dreamstime.com; family: © Pavel Losevsky/Dreamstime.com; no smoking: © Rosengaard/Dreamstime.com Printing and Binding: Courier Stoughton Cover Printing: Courier Stoughton Library of Congress Cataloging-in-Publication Data Labarthe, Darwin. Epidemiology and prevention of cardiovascular diseases : a global challenge / Darwin Labarthe. — 2nd ed. p. ; cm. Includes bibliographical references and index. ISBN-13: 978-0-7637-4689-6 (pbk.) ISBN-10: 0-7637-4689-4 (pbk.) 1. Cardiovascular system—Diseases—Prevention. 2. Cardiovascular system—Diseases—Epidemiology. 3. Cardiovascular system—Diseases—Etiology. I. Title. [DNLM: 1. Cardiovascular Diseases—epidemiology. 2. Cardiovascular Diseases—prevention & control. WG 120 L113e 2011] RA645.C34L33 2011 614.5'91—dc22 2009044199 6048 Printed in the United States of America 14 13 12 11 10 10 9 8
7
6
5
4
3
2
1
46394_FM_00i_xviii.qxd
2/26/10
2:30 PM
Page iii
Contents Foreword xi Preface xiii Dedication xv Acknowledgments
xvii
PART I—A PUBLIC HEALTH PERSPECTIVE Chapter 1
1
Cardiovascular Diseases: A Global Public Health Challenge
3
Summary 3 The Epidemiology and Prevention of Cardiovascular Diseases: Definition and Scope 3 The Basis of Public Health Concern Worldwide 5 Rates of Occurrence in Selected Populations and Changes in Recent Decades 10 Current Burdens of Major Cardiovascular Diseases in the United States and the World 12 Opportunities for Prevention 15 References 17
Chapter 2
Distributions and Disparities
19
Summary 19 Introduction 19 Age and Life Stages 22 Sex or Gender 26 Race or Ethnicity 27 Geography or Place 29 Person, Place, and Time 32 Conclusion 35 References 36
iii
46394_FM_00i_xviii.qxd
iv
2/26/10
2:30 PM
Page iv
CONTENTS
PART II—THE MAJOR CARDIOVASCULAR DISEASES Chapter 3
Atherosclerosis
39
41
Summary 41 Introduction 41 Methods of Measurement––Invasive and Noninvasive 42 Manifestations––Above and Below the “Clinical Horizon” 43 Mechanisms of Atherogenesis 46 Person, Place, and Time 47 Atherosclerosis in Childhood, Youth, and Early Adulthood 49 Prevention and Treatment of Atherosclerosis 54 Current Issues 55 References 55
Chapter 4
Coronary Heart Disease
59
Summary 59 Introduction 59 Background 62 Population Studies––Definition and Classification, Diagnostic Algorithms, and Criteria Rates 65 Risks 74 Trends and Explanations 79 Forecasts 83 Current Issues 83 References 83 Appendix 4-A: Key to Population Abbreviations Used by the WHO MONICA Project
Chapter 5
Stroke
89
Summary 89 Introduction 89 Background 91 Population Studies––Definition, Classification, and Diagnostic Methods Rates 93 Risks 101 Trends 104 Forecasts 107 Current Issues 107 References 107
92
63
87
46394_FM_00i_xviii.qxd
2/26/10
2:30 PM
Page v
CONTENTS
Chapter 6
Related Conditions
111
Summary 111 Introduction 111 Peripheral Arterial Disease 113 Aortic Aneurysm 120 Chronic Heart Failure 123 Deep Vein Thrombosis and Pulmonary Embolism Arrhythmias 135 References 136
PART III—THE MAIN DETERMINANTS Chapter 7
132
139
Genes and Environment
141
Summary 141 Introduction 142 Concepts and Strategies of Genetic Epidemiology 144 Family History 146 Gene–Environment Interaction 150 Cardiovascular Applications of Genomic Epidemiology 152 Current Issues 155 References 156
Chapter 8
Dietary Imbalance
159
Summary 159 Introduction 159 Concepts and Definitions of Dietary Patterns Measurement 165 Determinants 168 Distribution 170 Cardiovascular-Related Effects of Diet 172 Prevention and Control 180 Current Issues 185 References 186
Chapter 9
Physical Inactivity
191
Summary 191 Introduction 191 Concepts and Definitions Measurement 194
192
161
v
46394_FM_00i_xviii.qxd
vi
2/26/10
2:30 PM
Page vi
CONTENTS
Determinants 194 Mechanisms 196 Distribution 197 Cardiovascular-Related Effects Prevention and Control 211 Current Issues 217 References 218
Chapter 10
Obesity
204
223
Summary 223 Introduction 224 Concepts and Definitions 224 Measurement 229 Determinants 229 Mechanisms 233 Distribution 236 Rates and Risks 244 Prevention and Control 253 Current Issues 262 References 262
Chapter 11
Adverse Blood Lipid Profile Summary 269 Introduction 269 Concepts and Definitions 270 Measurement 273 Determinants 274 Mechanisms 275 Distribution 277 Relation to Rates and Risks 283 Relation to Other Factors 290 Prevention and Control 292 Current Issues 302 References 304
Chapter 12
High Blood Pressure Summary 311 Introduction 312 Concepts and Definitions Measurement 316
311
312
269
46394_FM_00i_xviii.qxd
2/26/10
2:30 PM
Page vii
CONTENTS
Determinants 317 Mechanisms 322 Distribution 323 Relation to Rates and Risks 336 Relation to Other Factors 339 Prevention and Control 341 Current Issues 351 References 352
Chapter 13
Diabetes and the Metabolic Syndrome Summary 361 Introduction 362 Concepts and Definitions 363 Measurement 366 Determinants 367 Mechanisms 368 Relation to Other Factors 369 Distribution 369 Rates and Risks 375 Prevention and Control 382 Current Issues 387 References 389
Chapter 14
Smoking and Other Tobacco Use Summary 395 Introduction 395 Concepts and Definitions 396 Measurement 397 Determinants 398 Mechanisms 398 Distribution 400 Relation to Rates and Risks 404 Relation to Other Factors 414 Prevention and Control 414 Current Issues 424 References 425
Chapter 15
Other Personal Factors Summary 431 Alcohol Consumption
433
431
395
361
vii
46394_FM_00i_xviii.qxd
viii
2/26/10
2:30 PM
Page viii
CONTENTS
Adverse Psychosocial Factors 447 Hemostatic Factors 464 Evolving and Emerging Factors 476 References 491
Chapter 16
Social and Physical Environment
503
Summary 503 Introduction 503 Social Status 505 Changes in Social Conditions 512 Particulate Air Pollution 522 Neighborhood Characteristics 524 Current Issues 527 References 528
PART IV—CAUSATION AND PREVENTION: THEORY, PRACTICE, AND FURTHER RESEARCH Chapter 17
What Causes Cardiovascular Diseases?
535
Summary 535 Introduction 535 Causal Judgment 536 Causal Constructs 540 Causation of Atherosclerotic and Hypertensive Diseases Conclusion 547 Current Issues 547 References 548
Chapter 18
Strategies of Prevention
551
Summary 551 Introduction 551 Concepts and Language of Prevention 552 Strategies of Prevention 557 Intervention Approaches 560 A Developing Country Perspective 561 Current Issues 563 References 565
543
533
46394_FM_00i_xviii.qxd
2/26/10
2:30 PM
Page ix
CONTENTS
Chapter 19
Evidence and Decision Making
567
Summary 567 Introduction 567 Nature of Evidence 569 Evidence-Based Decision Making 570 Approaches to Evaluation of Evidence 573 Current Issues 587 References 587
Chapter 20
Recommendations, Guidelines, and Policies
591
Summary 591 Introduction 592 Clinical Guidelines 594 Community Guidelines 607 Public Policies 609 Current Issues 613 References 614
Chapter 21
The Case for Prevention
619
Summary 619 Introduction 619 Experience with Multifactor Primary Prevention 621 The Burden of Risk 636 Economic Considerations 640 Models for Explanation and Prediction 642 Visions of Success in CVD and Chronic Disease Prevention Counter-Arguments 648 Current Issues 649 References 649
Chapter 22
Taking Action
646
657
Summary 657 Introduction: Calls to Action 657 Overview: Goals, Strategies, and Action Plans 658 Case Study: A Public Health Action Plan to Prevent Heart Disease and Stroke Obstacles to Taking Action 669 Current Issues 673 References 675
665
ix
46394_FM_00i_xviii.qxd
x
2/26/10
2:30 PM
Page x
CONTENTS
Chapter 23
Epidemiology and a CVD Prevention Research Agenda Summary 679 Introduction 680 Concepts of Epidemiology 681 Goals 682 Strategies of Investigation 683 Proposed Research Agendas 685 Capacity Requirements 690 Populomics: The Population Context of Research on Health Current Issues 693 References 693
INDEX
697
679
692
46394_FM_00i_xviii.qxd
2/26/10
2:30 PM
Page xi
Foreword
P
ublication of this second edition of Dr. Darwin Labarthe’s invaluable book, Epidemiology and Prevention of Cardiovascular Diseases: A Global Challenge is most opportune: As Dr. Labarthe emphasizes, the CVD prevention effort at this juncture—50 years down the road—confronts both “. . . considerable challenges and immense opportunities . . . .” On the one hand, the CVD epidemic persists; on a global scale it is waxing—indeed a challenge. And the challenge holds also for countries like the United States, where epidemic CVD persists despite substantial declines in coronary/stroke death rates during the latter decades of the 20th century. The current situation in the United States is problematic— as this book documents—making the challenges considerable indeed: tapering or cessation in recent years of down trends in CVD mortality and in major CVD risk factor levels (e.g., for saturated fat and cholesterol intakes, and for diet-dependent serum cholesterol and blood pressure); epidemic obesity with its consequences, including rampant incidence of diabetes and other obesity-driven metabolic CVD risk factors; unabated high salt intake; overall dietary and physical activity patterns still generally adverse populationwide including among children and teenagers; all too many still smoking, all too many teenagers becoming smokers; even more so for lower socioeconomic strata of all ethnicities, hence paltry proportions of all strata at low CVD risk––a critical index––and little or no evidence of a sustained upward slope in this index; in the media, especially TV, an on-going flood of promotions of foods/beverages harmful for heart health; in medical practice, overwhelming reliance on a high risk strategy (reactive, not proactive) to cope with these challenges––a focus on detection of people who already have a high level of the established major CVD risk factors and their long-term treatment with medications (in 2008, 320.4 million prescriptions for antihypertensive and 139.6 million prescriptions for antihypercholesterolemic drugs, as reported by the AARP). However useful for patients already at high CVD risk, this limited one-sided strategy relying on pills as the remedy begs the basic issue: Epidemics are due to population-wide exposures to new ways of
life for which the human species has not been adapted over the 2–4 million years of hominid/hominoid evolution; their roots are mass “. . . disturbances of human culture . . .” (Rudolf Virchow)—generalizations fully applicable/valid for the CVD epidemic, as this monograph details. To end the CVD epidemic, the sine qua non is rectification of the multiple disturbances in human culture causing it—a proposition repeatedly verified as valid by the history of conquest of earlier epidemics (e.g., tuberculosis, pellagra, rickets). The opportunities to conquer the CVD epidemic are indeed immense. First and foremost, prerequisite knowledge concerning the etiology of the CVD epidemic: the data base (already substantial 50 years ago) is now vast—extensive concordant data, worldwide in scope, accumulated over decades by epidemiology and every other research methodology available to medicine. Critical detailed information on the multiple causes of epidemic CVD is in hand—and for prevention of mass disease, such information on causation, the “question of questions,” is decisive. We know in depth what needs to be done—at every level of prevention—to break links in the chain of causation, including for primary and primordial prevention, i.e., the prevention in the first place (from preconception on) of the adverse lifestyles and the lifestylerelated established major risk factors. Crucial to this effort are improved eating patterns, Mediterranean and East-Asian style cuisines updated for this century, especially as to lower salt, plus moderation in intakes of alcohol, fats, and total calories—along with regular frequent exercise and non-smoking. The update for the 21st century—derived from the research achievements of the last 50 years—gives an enhanced nutrient intake pattern: as earlier, low in saturated fats and cholesterol; reduced in total fats; enhanced in polyunsaturated fats; calorie controlled; plus free of trans fats; much lower in salt; reduced in sugars especially separated sugars (e.g., from sweetened beverages); enhanced in total protein, especially vegetable protein (lower in animal protein from meats); for those who drink, moderate (not excessive) in alcohol; enhanced in potassium/calcium/magnesium/phosphorus/ non-heme iron, the vitamins, and fiber (from whole xi
46394_FM_00i_xviii.qxd
xii
2/26/10
2:30 PM
Page xii
FOREWORD
grains/beans/vegetables/fruits). As this enumeration indicates, the research findings (including from population-based observational studies and clinical trials) document the multifaceted dietary imbalances–– concurrent excesses and inadequacies––now implicated in the etiology of the CVD epidemic and its diet-related major metabolic risk factors. Consequent recommendations enable variegated approaches to modern delightful eating styles assuring avoidance/rectification of these imbalances. These eating styles, along with regular frequent exercise, offer the potential for all population strata (socioeconomic/ethnic) to prevent/check/correct the nowadays still usual development of adverse levels of major metabolic risk factors: serum total/LDL/ VLDL/HDL cholesterol; blood pressure; plasma glucose; weight; and they go beyond these merits, since adverse eating patterns produce excess CVD risk over and above their adverse influences on these metabolic risk factors. So the opportunities are truly immense, as Dr. Labarthe emphasizes. Their scope encompasses potential for realization of the critical goal: continuous progressive enhancement in the coming years/decades of the percent of the population at low risk, so that for most people—not just a small minority—CVD risk is miniscule; they are freed of the burden of epidemic CVD, with consequent enhanced longevity with health. High stakes indeed! The opportunities are immense also because on a world scale and in several regions of the world, public policy is in place at the national level (including in the United States), policy committed to the accomplishment of CVD prevention through a two-pronged strategy (population-wide and high risk) emphasizing improved lifestyles. In a few places, public policy specifically includes priority for achieving the decisive goal of progressively increasing the percentage of the population at low risk. In the United States, substantial funds have recently been allocated—specifically to the national Centers for Disease Control and Prevention (CDC)—for the CVD prevention effort. Opportunities are immense also for this effort because many countries—ranging from Finland to Japan to the United Kingdom and the United States— have already over decades accrued extensive positive experiences with sustained public health efforts to improve lifestyles, thereby control lifestyle-related major CVD risk factors, and contribute to CVD
prevention/control. Repeatedly, the public has been responsive and substantial progress (albeit incomplete) has been achieved, despite opposition from special interests (including sectors of the food and beverage industries, the big tobacco companies)––e.g., in the United States, sizable declines in intakes of saturated and trans fats, total fats, cholesterol; the related decrease in adult population average serum cholesterol from about 240 mg/dl 50 years ago to about 200 mg/dl by the year 2000, achieving a national public health goal; marked falls in the prevalence of cigarette smoking; associated declines––in the order of 50% or more—in mortality from CHD and stroke, with consequent addition of years to life expectancy for young, middle-aged, and older adults. And, in several countries, as well as internationally, there are significant social movements in place, supporting/encouraging the effort, bringing together health professionals and lay leaders in effective alliances. As Dr. Labarthe notes, this too is an important component for a successful prevention effort—important today for CVD, as it was in the 19th century for TB control. All these are indeed solid bases for accomplishment of next key tasks. As noted repeatedly, this book is replete with many-sided up-to-date information invaluable for every person concerned with the CVD prevention effort. It is a fitting product of Dr. Labarthe’s extraordinary capacities and experiences over decades— as a colleague, teacher, researcher, public health leader—in academia, at the CDC, at local/national/ international learning venues, including the seminal US and International Ten Day Teaching Seminars on CVD Epidemiology and Prevention he has effectively led for years. On a personal note, over 40 years ago when I authored an early monograph on this same subject, it was my privilege for it to have a Foreword by Paul Dudley White, MD––distinguished cardiologist, statesman, world leader, humanist/humanitarian. Its opening sentence read, “Dr. Jeremiah Stamler has written the book on Preventive Cardiology that I would liked to have written.” Today these words are mine in regard to this volume by Darwin Labarthe. Jeremiah Stamler, MD Professor Emeritus Feinberg School of Medicine Northwestern University Chicago, Illinois
46394_FM_00i_xviii.qxd
2/26/10
2:30 PM
Page xiii
Preface The central messages of this second edition are these:
change on the context and content of the book is substantial; however, it is necessary to underscore the disclaimer that the views expressed throughout are personal and are not intended to represent the official position of the US Centers for Disease Control and Prevention (CDC) or the US Department of Health and Human Services (DHHS). The book has changed principally in presenting greater emphasis on public health in cardiovascular disease prevention while retaining its epidemiologic content. The intent has been to increase the book’s value for both epidemiologists and public health professionals by bringing the original content up to date in Parts I–III and expanding discussion of how epidemiology is translated into policy and practice in Part IV. Currency has been achieved by including more than 1600 citations and nearly 400 tables and figures, many from recent sources; rewriting the chapters on the major cardiovascular diseases and their determinants; introducing a chapter on genomic epidemiology; and expanding discussion of the global dimensions of CVD. For many sources, URLs are included to permit continuing access for interested readers. New chapters in Part IV address strategies of prevention as part of a recently developed action framework; the nature of evidence for prevention, and methods for its evaluation as practiced by several leading authoritative bodies; current national, regional, and global recommendations, guidelines, and policies for prevention of CVD and other major chronic diseases; the case for CVD prevention at individual and population levels; and action plans adopted for implementation in the United States, Europe, South Asia, and worldwide. Expansion in these areas had one regrettable cost—lack of updates on rheumatic heart disease, Chagas’ disease, congenital heart disease, and Kawasaki disease—which are treated only in the first edition. Features retained in the second edition include the basic structure in which the public health perspective is introduced in Part I; the major atherosclerotic and hypertensive diseases are discussed in Part II;
(1) Cardiovascular diseases remain the foremost causes of preventable death globally and continue to grow in prominence, because of their attendant burden, disparities, and costs. (2) Epidemiology has contributed immeasurably to a vast body of knowledge about the causes and means of prevention of these and related conditions, but this knowledge has yet to be applied on a sufficient scale to confer its potential societal benefit. (3) Public health is accountable for putting this knowledge more fully to work by setting goals, devising strategic plans and policies, implementing targeted actions, and documenting their impact in improving the health of populations. These messages are consistent with the content of the first edition but have gained force from developments in the intervening decade: increased awareness of the global burden of cardiovascular diseases, with their immense social and economic consequences; a growing sense of need to integrate approaches to cardiovascular diseases with prevention of other chronic or noncommunicable diseases, with transformation of health systems to address them coherently; and the ever more urgent goal to reduce the mounting burden, disparities, and costs of these diseases. Epidemiology, through its applications in development, adoption, and implementation of health policy and in public health practice, is fundamental to achieving this goal. It is the author’s hope that this new edition will contribute to this effort. The difference of a decade is due importantly to advances in science and practice that better inform our understanding of the need and opportunity for effective action. From the start of this past decade, the perspective of the author, too, has advanced—from that of the academic epidemiologist to that of the public health practitioner. This was a result of undertaking a US governmental role at the federal level and the greatly enhanced public health experience afforded by this opportunity. The influence of this xiii
46394_FM_00i_xviii.qxd
xiv
2/26/10
2:30 PM
Page xiv
PREFACE
their main determinants are reviewed sequentially in Part III; and implications of this evidence for theory, practice, and research are discussed in Part IV, which concludes with a chapter on CVD epidemiology of the future—the rich and varied research opportunities presented and the place of epidemiology as the core discipline of “populomics,” the scientific foundation of population health. A historic perspective is also retained, although this is not meant to recount the history of the field, which is being done in a far more effective way elsewhere. The purpose instead is to illustrate wherever appropriate the key studies that, from early in the development of CVD epidemiology, have made fundamental and lasting contributions to our current knowledge. Throughout the book, the unifying approach of a single author has the advantage of a consistent presentation and coherent interpretation across the many topics addressed. There is room for differing opinion and further exploration of many topics raised. The content reflects one person’s perspective and in no
case represents an exhaustive systematic review, although those of others are cited extensively. Closing each chapter is a more or less speculative suggestion of current issues most important for further discussion. In these ways the text is intended to stimulate thinking and debate. The author welcomes comments, queries, and criticisms from readers. A Chinese proverb says, “Teachers open the door but you must enter by yourself.”1 It is hoped that the material that follows will open many doors for students and practitioners of CVD prevention and public health, revealing a world of opportunity for fulfilling our highest obligation: to assure conditions in which people can be healthy.
Reference 1. Schiller D. The Little Zen Companion. New York: Workman Publishing; 1994.
46394_FM_00i_xviii.qxd
2/26/10
2:30 PM
Page xv
Dedication This book is dedicated to all whose work is reflected here and to all who will contribute to advances in the understanding of cardiovascular diseases and reduction of the public health burden they represent throughout the world.
xv
46394_FM_00i_xviii.qxd
2/26/10
2:30 PM
Page xvi
46394_FM_00i_xviii.qxd
2/26/10
2:30 PM
Page xvii
Acknowledgments George Howard, William Kannel, Thomas Kottke, Ian MacMahon, Henry McGill, Kenneth Pelletier, Douglas Schocken, Richie Sharrett, Jeremiah Stamler, Elaine Stone, Jack Strong, and Thomas Thom. Several peer reviewers provided valuable comment and suggestions. Although their anonymity precludes personal recognition, it is hoped that they will see their input reflected in the final product. Merrily Labarthe deserves boundless gratitude for her support and forbearance throughout this project.
Technical assistance in bibliographic work was provided with diligence and skill by T. Christopher Bond; the permissions process was supported by Tiffany Lynn Williams and by Anthony Omokheowa Anani and Elohor Anani. Contributions of critical points of information were made by Patty Borhani, Ross Brownson, Michele Casper, Elizabeth Barrett-Connor, Rory Collins, Leonard Cook, Jeffrey Cutler, Jack Farquhar, Lawrence Green, Nancy Haase, Millicent and Ian Higgins,
xvii
46394_FM_00i_xviii.qxd
2/26/10
2:30 PM
Page xviii
46894_CH01_001_018.qxd
2/25/10
3:22 PM
Page 1
A Public Health Perspective
1
P A R T
1
46894_CH01_001_018.qxd
2/25/10
3:22 PM
Page 2
46894_CH01_001_018.qxd
2/25/10
3:22 PM
Page 3
C H A P T E R
1 Cardiovascular Diseases: A Global Public Health Challenge vention of cardiovascular diseases define a significant public health challenge.
SUMMARY Cardiovascular diseases comprise especially the major disorders of the heart and the arterial circulation supplying the heart, brain, and peripheral tissues. Their common occurrence in most populations and the great attendant mortality, loss of independence, impaired quality of life, and social and economic costs are compelling reasons for public health concern. The epidemiology and prevention of these diseases involve the understanding of their causes, identification of means of prevention, and monitoring of populations to assess the changing burden of these diseases and the measurable impact of interventions to control them. Together, the cardiovascular or circulatory diseases have figured prominently in the large shifts among causes of death, especially in industrial societies, during the 20th century. During this period they have become the predominant cause of death in many countries and in the world as a whole. The “theory of epidemiologic transition” offers an interpretation of these shifts. It may have special implications for developing countries, as increasing proportions of these populations attain older ages, social changes unfold, and disease patterns change. Evidence indicates that cardiovascular diseases are already epidemic in low- and middle-income as well as high-income regions of the world and have become deep-rooted in most societies in recent decades. Cardiovascular epidemiology has documented the nature and extent of the major atherosclerotic and hypertensive diseases as global phenomena. It has contributed substantially to establishing their underlying causal factors. It has also identified the potential for prevention on a population-wide scale, including prevention of the risk factors themselves. Together, the global burden and immense opportunities for pre-
THE EPIDEMIOLOGY AND PREVENTION OF CARDIOVASCULAR DISEASES: DEFINITION AND SCOPE The cardiovascular diseases, or diseases of the heart and blood vessels, comprise many conditions that vary widely in manifestations and in public health importance. The present focus is chiefly on the atherosclerotic and hypertensive diseases. These are the cardiovascular conditions that develop on the basis of longstanding disease of the walls of arteries, especially in the heart, brain, and lower extremities, or of the aorta or as a consequence of persistently high blood pressure. In many but not all populations, these underlying processes—both atherosclerosis and hypertension—coexist. Heart attacks and strokes are very common manifestations of these conditions and are the chief contributors to their major public health importance. Also of public health concern are certain conditions affecting the venous circulation as well as disturbances of cardiac rhythm. After the presentation of a public health perspective on the atherosclerotic and hypertensive diseases in Part I, these and several related conditions will be described in detail in Part II. Their main determinants have become scientifically established over a half century or more through research in populations as well as clinical and laboratory studies. This population research is the main subject of Part III. The contribution of cardiovascular epidemiology to strategies of prevention of these conditions, the evidence and rationale for public health approaches to prevention, and plans of action regarding public health 3
46894_CH01_001_018.qxd
4
2/25/10
3:22 PM
Page 4
CHAPTER 1 CARDIOVASCULAR DISEASES
policy, practice, and research are addressed in Part IV, the concluding section. Preceding a broad overview that addresses some historical and global dimensions of the atherosclerotic and hypertensive diseases, two notes may be useful for clarification. First, terminology varies widely as the conditions of concern are defined and discussed throughout the extensive literature of this field. For example, “CVD” may refer to cardiovascular disease, meaning specifically disease of the heart or its blood supply, or to cerebrovascular disease, affecting the circulation of the brain. A major condition of the heart, coronary heart disease, may be abbreviated as “CHD,” “IHD” (ischemic heart disease), or “CAD” (coronary artery disease). Throughout the book, terminology of original sources is used in presentation of tables, figures, and sometimes text. It is hoped that fidelity to the source will not result in confusion for the reader. In general, the expressions “cardiovascular disease or diseases” or “heart disease and stroke” are used here as equivalent terms and refer to the full spectrum of the atherosclerotic and hypertensive diseases. Second, an underlying view of epidemiology in general and of cardiovascular epidemiology in particular doubtless gives shape to the organization and presentation of what follows. One premise is that the utility of epidemiology, and a great part of its societal value, depends on application of its findings toward improvement of the public’s health. Translated into cardiovascular epidemiology, the premise of this book is that the work of a half-century or more to understand the causes and discover the means of prevention of cardiovascular diseases on a national and global scale establishes not only the possibility but also the public health responsibility to seek effective action based on this science.
Table 1-1
An overview follows that addresses issues of classification; the nature and magnitude of the public health challenge; rates and burdens of cardiovascular diseases past, present, and future; and concepts of cardiovascular disease development and prevention. The Cardiovascular Diseases—An International Classification What is the scope of cardiovascular diseases? Definition and classification of the cardiovascular diseases, as with other conditions, have evolved with changing concepts of disease and take many forms, in part because of different purposes. In epidemiology, special value attaches to a classification that is standardized and in common use in many or most hospitals, medical practices, states, countries, and regions of the world. In this way, some confidence is justified that reference to the same condition in different information sources corresponds to the same reality. The leading source of such a classification for use throughout the world is the World Health Organization publication, now in its 10th revision, the International Statistical Classification of Diseases and Related Health Problems (ICD 10).1 Published in 1992, ICD 10 presents the category of Diseases of the Circulatory System as shown in Table 1-1. Each three-character code in this classification has an alphabetic initial followed by two digits. The alpha code for this category is the letter I, with blocks of digits from 00-02 to 90-95 that distinguish 10 classes. Within a block, each two-digit code corresponds to a distinct subset of that class. Greater detail can be provided by use of an additional decimal place. For example, for “ischemic heart diseases” (I20-I25) the code I21 identifies acute myocardial in-
Diseases of the Circulatory System (I00–I99)
I00–I02
Acute rheumatic fever
I05–I09
Chronic rheumatic heart diseases
I10–I15
Hypertensive diseases
I20–I25
Ischemic heart diseases
I26–I28
Pulmonary heart disease and diseases of pulmonary circulation
I30–I52
Other forms of heart disease
I60–I69
Cerebrovascular diseases
I70–I79
Diseases of arteries, arterioles, and capillaries
I80–I89
Diseases of veins, lymphatic vessels, and lymph nodes, not elsewhere classified
I90–I95
Other and unspecified disorders of the circulatory system
Note: Classification excludes congenital malformations, transient cerebral ischemic attacks and related syndromes, and certain others.
46894_CH01_001_018.qxd
2/25/10
3:22 PM
Page 5
THE BASIS OF PUBLIC HEALTH CONCERN WORLDWIDE
pertensive diseases and selected venous system disorders (essentially I10-I15, I20-I25, I26, I60-I69, and parts of I70-I79, I80-I89, and G45) are discussed. In the chapters specific to a given condition, further detail of the ICD 10 codes will be noted, and earlier ICD codes will be referenced as needed. This is because ICD 10 has been widely implemented only recently; the data currently available on the conditions of interest therefore represent largely one or another of the previous versions, each of which was current for about a decade.
farction, or heart attack; codes I21.1 through I21.4 represent the anatomic location of the damage within the heart; and I21.9 is used for cases where location is not specified. Thus a case record, whether in the form of a death certificate or a hospital discharge summary, can potentially be coded in a consistent way, and cases with the same code can be collected and treated statistically as representing the same kind of circulatory event. The validity of such analyses depends, of course, on the quality of information available and the nosologic coding procedures applied. The conditions listed in Table 1-1 are those judged by the writers of ICD 10 to be classified best as diseases of the circulatory system. The conditions addressed in the following chapters are mainly I20-I25, I60-I69, and I10-I15. One additional category not shown in Table 1-1 also receives attention— transient cerebral ischemic attacks (TIA) or “light strokes.” For present purposes, these events are considered to belong with the cerebrovascular diseases or circulatory conditions affecting the brain. More detailed classification of these conditions is addressed in Chapters 3–6, in which the atherosclerotic and hy-
THE BASIS OF PUBLIC HEALTH CONCERN WORLDWIDE The Magnitude of the Problem Why do the cardiovascular diseases, taken together, warrant epidemiologic attention? The answer lies in part in the very large proportion of deaths, throughout the world, attributed to cardiovascular conditions. Figure 1-1 presents the percentages of deaths
World Total Perinatal diseases Accidents 6% 5% Cancer 9%
All other diseases 24%
Infectious and parasitic diseases 14%
CVD 23% Respiratory diseases (including TB) 18%
Industrial Countries
Developing Countries
Accidents 5% Cancer 6% Perinatal diseases 7%
Respiratory diseases (including TB) 7.5% Accidents 7%
All other diseases 27%
All other diseases 18.5%
Cancer 19%
CVD 16% Respiratory diseases (including TB) 21%
Infectious and parasitic diseases 18%
5
CVD 48%
Note: CVD, cardiovascular disease; TB, tuberculosis. Of the total deaths 78% are in developing countries.
Figure 1-1 Relative Contributions of Cardiovascular Death to Total Mortality in Developing and Industrial Countries and the World, 1980. Source: From Disease Control Priorities in Developing Countries, edited by DT Jamison et al., © 1993 by World Bank. Reproduced with permission of World Bank in the format Textbook via Copyright Clearance Center.
46894_CH01_001_018.qxd
6
2/25/10
3:22 PM
Page 6
CHAPTER 1 CARDIOVASCULAR DISEASES
due to cardiovascular diseases (“CVD” in the figure) in the world as a whole (23%) and separately in developing countries (16%) and industrial countries (48%) as already recognized in 1980.2 (This inclusive category comprises all cardiovascular deaths, which is heavily dominated by ischemic and hypertensive heart diseases and cerebrovascular disease.) Such estimates are subject to reservations, especially for the developing countries, as to the completeness of death registration and accuracy of cause-of-death assignment. Even allowing for these concerns, there is substantial support for the view that cardiovascular diseases have been increasing in frequency for some decades to constitute a rising public health problem of developing countries. In fact, the proportion of all deaths worldwide occurring in these populations is so great that the majority of cardiovascular deaths worldwide occur in developing countries.
the 20th century, as shown for the United States, over the period 1900–1970 in Figure 1-2.3 The importance of this early analysis of mortality in the United States is its contribution to a theory of populationwide changes in patterns of disease formulated in the 1970s, discussed below. The figure serves well even now to illustrate the changes in death rates, or the numbers of deaths per 100,000 population per year, due to multiple categories of causes over several decades. The relative shift for heart disease resulted from both an absolute increase in the rate of heart disease deaths (from a little more than 100 deaths to about 400 deaths per 100,000 population) and concurrent major decreases in other causes of death, especially in tuberculosis and other infectious diseases. Even before the 1920s, heart disease and stroke together exhibited mortality greater than that from any other category. In 2004 (the most recent year for which final mortality data were available at the time of writing), they still accounted for more than 35% of all deaths in the United States, as described in subsequent chapters.4
Changing Patterns of Mortality in the United States In industrialized countries, the prominence of heart disease among causes of death rose sharply during 2000 1000 INFECTIOUS DISEASES 500
Deaths per 100,000 Population
HEART DISEASE
CANCER STROKE
100
VIOLENT/ACCIDENTAL DISEASES OF EARLY INFANCY
50
10 TUBERCULOSIS 5
UNITED STATES 1900–1970 1
1900
1910
1920
1930
1940
1950
1960
1970
Years
Figure 1-2 Secular Trends for Cardiovascular Disease and Other Cause-Specific Death Rates in the United States, 1900–1970. Source: From the Population Bulletin. © 1977, Courtesy of the Population Reference Bureau, Inc., Washington, DC.
46894_CH01_001_018.qxd
2/25/10
3:22 PM
Page 7
THE BASIS OF PUBLIC HEALTH CONCERN WORLDWIDE
Such vast shifts in causes of death stimulate strong epidemiologic interest because they must reflect profound changes in the factors that influence health and disease. Factors identified or confirmed through their association with such trends may constitute clues to causation or point to potential interventions. As will be seen, major increases or decreases in disease occurrence have until recently proven to be difficult to explain in retrospect. Methods of data analysis that incorporate information on factors related to prevention and treatment are currently used to sort out the contributions of these influences (see Chapter 4). But the experience of the United States indicated in Figure 1-2 is not unique, and such changes in patterns of mortality may be in progress in many countries. This creates the possibility of observing these changes—to the extent they may escape effective control measures—as they actually unfold in some populations. The likelihood of this is suggested both by epidemiologic theory and by a number of observations illustrated in the following section. The “Theory of Epidemiologic Transition” The theoretical basis for this view is that of “epidemiologic transition,” formulated by Omran in an analysis of long-term patterns of mortality in human societies and first published in 1971.5 According to Omran, “Conceptually, the theory of epidemiologic transition focuses on the complex change in patterns of health and disease and on the interactions between
these patterns and their demographic, economic, and sociologic determinants and consequences …”5, p 509 Omran distinguished three stages of progression, historically over centuries, in the dominant patterns of mortality: the “age of pestilence and famine,” the “age of receding pandemics,” and the “age of degenerative and man-made diseases.” In a later extension of the theory, Olshansky and Ault have proposed a fourth stage, the “age of delayed degenerative diseases.”6 These four stages of the epidemiologic transition are indicated in a representation by Gaziano and others published in the second edition of a major World Bank publication, Disease Control Priorities in Developing Countries, in 2006 (Table 1-2).7 These authors elaborate on an earlier version presented by Pearson and others2 by expanding the description of each stage and adding information on life expectancy, the percentage of the world’s population in each stage, and the regions affected. In addition, they note “CHF”—chronic heart failure—among dominant forms of cardiovascular diseases present in the fourth phase (see Chapter 6). One point of particular interest in this and the previous version of the table is the “percentage of deaths attributable to CVD.” These estimates of proportionate mortality due to cardiovascular diseases in each stage were introduced by Pearson and colleagues as approximations based on their judgment (T.A. Pearson, personal communication, 2004). The shift from very low to much higher frequencies of circulatory diseases
Table 1-2
The Epidemiologic Transition Phase of Deaths from Epidemiologic Circulatory Transition Disease (%) Age of pestilence 5–10 and famine
7
Circulatory Problems Rheumatic heart disease; infectious and deficiencyinduced cardiomyopathies
Risk Factors Uncontrolled infection; deficiency conditions
Age of receding pandemics
10–35
As above, plus hypertensive heart disease and hemorrhagic stroke
High-salt diet leading to hypertension; increased smoking
Age of degenerative and man-made diseases
35–55
All forms of stroke; ischemic heart disease
Atherosclerosis from fatty diets; sedentary lifestyle; smoking
Age of delayed degenerative diseases
Probably under 50
Stroke and ischemic heart diseasea
Education and behavioral changes leading to lower levels of risk factors
Note: Omran introduced the concept of epidemiologic transition with discussion of phase 1, 2, and 3. Olshansky and Ault added the concept of a fourth phase. a
At older ages. Represents a smaller proportion of deaths.
Source: From Disease Control Priorities in Developing Countries, edited by DT Jamison et al. Copyright © 1993 The International Bank for Reconstruction and Development/The World Bank. Used by permission of Oxford University Press, Inc.
46894_CH01_001_018.qxd
8
2/25/10
3:22 PM
Page 8
CHAPTER 1 CARDIOVASCULAR DISEASES
as causes of death is much like that demonstrated in Figure 1-2. As shown in Table 1-1, under this theory circulatory diseases increase from a minor proportion of all deaths in the first stage to become the predominant cause of death in the third stage; finally, they may decrease slightly in relative importance, though still perhaps representing the largest single category of deaths. Characteristic shifts also occur in the predominance of particular forms of circulatory disease over the successive phases: first, solely rheumatic heart disease and cardiomyopathies; second, these and also hypertensive heart disease and hemorrhagic stroke; third, all forms of stroke plus ischemic heart disease; and fourth, the latter causes persisting but occurring at older ages and as a somewhat reduced proportion of all deaths. In the United States and other countries undergoing industrialization in the 19th and 20th centuries, this epidemiologic transition is already far advanced into the third or fourth stage. Developing countries, however, have widely been thought to remain in the
first or second stage because proportionate mortality from circulatory conditions remained low; but absolute rates reached levels of concern even while dominance of communicable diseases persisted. Two decades and more ago, demographic changes already under way were thought capable of producing more rapid transition in these countries than was experienced by the already-industrialized countries. Dodu, of the World Health Organization Cardiovascular Diseases Unit in Geneva, wrote in 1988 of the emergence of cardiovascular diseases in developing countries.8 He presented data (Figure 1-3), based on Omran’s work, to show that the percentage of deaths due to cardiovascular diseases (and cancer) in a population increases as life expectancy at birth increases. This would be anticipated because cardiovascular disease death rates are very much higher for successively older age groups in adulthood. As shown by Dodu, when the average person attains age 61 or 62 years, cardiovascular diseases are expected to predominate over infectious diseases as a cause of death.
Figure 1-3 Percentages of Deaths Due to Cardiovascular Diseases (CVD), Cancer (CAN), and Infections, in Relation to Life Expectancy at Birth. Source: From Cardiology, Vol 75, Emergence of Cardiovascular Diseases in Developing Countries, SRA Dodu, © 1988 S Karger AG, Basel. Reproduced with permission from United Nations Secretariat, p 58.
46894_CH01_001_018.qxd
2/25/10
3:22 PM
Page 9
THE BASIS OF PUBLIC HEALTH CONCERN WORLDWIDE
Dodu further demonstrated, as shown in Figure 1-4, that life expectancy at birth increased sharply in developing regions of the world in the third quarter of the 20th century. Further, this increase was projected to continue, so that by the year 2000 even Africa would attain an average life expectancy at birth of nearly 60 years. This value would be exceeded slightly by that for the population of South Asia but by far for East Asia and Latin America. This changing demographic picture alone, influenced partly by recession of infectious diseases as a cause of neonatal and infant mortality, leads to an expectation of an increasing proportion of deaths from circulatory diseases, in accordance with the theory of epidemio-
9
logic transition. As a concrete example, Dodu cited the experience of Singapore, where in 30 years (1948–1979) life expectancy increased from about 40 years to 70 years, and cardiovascular diseases shifted from only 5% to more than 30% of all deaths. Economic Considerations Worldwide public health concern about cardiovascular diseases is partly because of the high frequency of occurrence of these diseases as a cause of death. This reality continues in industrialized countries and is increasingly recognized in developing countries. In addition, among the personal and social costs of cardiovascular diseases, both fatal and nonfatal, are their
Figure 1-4 Life Expectancy in Relation to Calendar Time, by Region of the World, 1950–2030. Source: From Cardiology, Vol 75, Emergence of Cardiovascular Diseases in Developing Countries, SRA Dodu, © 1988 S Karger AG, Basel. Reproduced with permission from United Nations Secretariat, p 62.
46894_CH01_001_018.qxd
10
2/25/10
3:22 PM
Page 10
CHAPTER 1 CARDIOVASCULAR DISEASES
economic costs, which are increasingly important. In the United States in the year 2009, for example, the cost of medical care alone for cardiovascular diseases is projected to be $313.8 billion, with additional indirect costs due to disability and death in working years of $161.5 billion, making the total one-year economic burden $475.3 billion.4 The World Health Organization estimated the economic burden of lost income (analogous to the indirect costs above) due to heart disease, stroke, and diabetes for nine countries across a spectrum of income levels, over the years 2005–2015.9 For China, such losses were projected to reach 18.3 billion international dollars for 2005 and 131.8 billion dollars in 2015. These macroeconomic dimensions of cardiovascular diseases lend importance to intensified efforts to prevent these diseases in populations throughout the world, in both industrialized and developing countries.
RATES OF OCCURRENCE IN SELECTED POPULATIONS AND CHANGES IN RECENT DECADES World Bank Regions, 1985 The status of the major regions of the world with respect to mortality from circulatory system diseases toward the end of the 20th century is summarized in Table 1-3.2 For the total world population and for each geographic/economic area distinguished by the World Bank, Table 1-3 indicates for 1985 the total numbers of deaths (in thousands), the percentages of the total due to circulatory diseases (as discussed
above), and the death rate (per 100,000 population, adjusted for differences between regions in age composition) for all circulatory diseases and for two of the component categories, ischemic heart disease and cerebrovascular disease. The percentages of deaths from circulatory diseases were highest for the industrial economies, both market economies such as the United States (46%) and nonmarket ones such as the countries of the former Soviet Union (47%). They were only one-half to one-quarter as high for the remaining four regions (10–22%). The total death rate was highest for the industrial nonmarket economies (357 per 100,000). This group of countries experienced a 60% higher death rate from “ischemic disease” and a 65% higher death rate from cerebrovascular disease than did the industrial market economies. Scrutiny of Table 1-3 reveals something of a paradox. The lower percentages of deaths due to circulatory diseases in all of the nonindustrial regions might be taken as consistent with the epidemiologic transition and give the impression that circulatory diseases are not yet important in these regions. However, this interpretation is refuted by the actual death rates for ischemic disease and cerebrovascular disease. These rates were nearly as high or higher for nonindustrial as for industrial market economies—such as the United States. Asia was again exceptional, in this instance by having the lowest mortality for “ischemic disease” of any region (46 per 100,000) and the highest for cerebrovascular disease (91 per 100,000) outside the industrial nonmarket region. Clearly, both categories of circulatory disease were as well established in the nonindustrial regions of the world as in the industrial market economies as measured by mortality experience as
Table 1-3
Estimated Mortality from Circulatory System Diseases, World Bank, 1985 Age-Standardized Death Rate (per 100,000 Population)a Region Deaths Total Ischemic Cerebrovascular (Thousands) Deaths (%) Total Disease Disease Industrial market economies 3355 46 235 99 59 Industrial nonmarket economies 2220 47 357 164 106 Latin America and the Caribbean 691 22 222 69 57 Sub-Saharan Africa 756 10 273 85 74 Middle East and North Africa 602 14 250 82 68 Asia 3841 17 195 46 91 Total 11,465 23 243 84 81 a
Rates are standardized using the 1985 world age structure.
Source: From Disease Control Priorities in Developing Countries, edited by DT Jamison et al. Copyright © 1993 The International Bank for Reconstruction and Development/The World Bank. Used by permission of Oxford University Press, Inc.
46894_CH01_001_018.qxd
2/25/10
3:22 PM
Page 11
RATES OF OCCURRENCE IN SELECTED POPULATIONS AND CHANGES IN RECENT DECADES
early as 1985, more than two decades ago. These rates were simply dominated in these regions by persistent high mortality from noncirculatory causes, resulting in low proportionate mortality despite substantial circulatory death rates. The epidemiologic transition had evidently taken a different course in these regions than in the industrialized parts of the world. Change from 1988–1998 in 15 Countries As of the late 20th century, marked differences—2- to 3-fold in Table 1-2—in mortality from specific cardiovascular causes were demonstrable in populations comprising major economic and geographic regions of the world. The mortality pattern of a single country such as the United States was seen to have evolved throughout much of the last century. But a country may experience striking change in ischemic heart disease mortality within a decade or less, as shown in Figure 1-5.7 In this figure, data on ischemic heart disease mortality are presented as the percentage change in death rates among people aged 35–74 years at death,
from 1988 to 1998 for 15 selected countries. The changes in rates are shown for both males and females in each country. The resulting picture is one of a continuous gradient of change that ranges from ⫹62% to ⫺49% for males in Croatia and Denmark, respectively, and from ⫹61% to ⫺52% for females in Croatia and Australia, respectively. That the adverse, upward changes were clustered in Eastern European countries of the former Soviet Union is noteworthy, as is the exceptional situation of Hungary with virtually no change. Also noteworthy is the similarity in patterns of change—in both direction and magnitude—for males and females in the same countries. These differences over time in changes in mortality are of great epidemiologic interest because of the between-population dimension of variation that they represent and resulting possibilities for comparative investigation. Deeper insight into the causes of these large population changes and the potential for prevention or control of the underlying epidemic processes would be anticipated as a result.
62%
Croatia
61%
Kazakhstan
56%
Belarus
53%
30%
Ukraine
49%
38%
Romania
26% 26% 210%
Japan
22%
Hungary 215%
Greece
229%
Portugal
28% 2% 211% 219%
229% 239%
Netherlands
Luxembourg
36%
229% 230%
United States
Sweden
11
240% 220%
243% 243%
Australia
252%
246%
Denmark
246%
249%
Males Females
Figure 1-5 Percentage Change in Ischemic Heart Disease Death Rates in People Age 35–74, 1988–1998, Selected Countries. Source: Reproduced with permission from Disease Control Priorities in Developing Countries, 2nd edition, edited by DT Jamison et al., © 2006 by World Bank. Courtesy of the International Bank for Reconstruction and Development/The World Bank.
46894_CH01_001_018.qxd
12
2/25/10
3:22 PM
Page 12
CHAPTER 1 CARDIOVASCULAR DISEASES
CURRENT BURDENS OF MAJOR CARDIOVASCULAR DISEASES IN THE UNITED STATES AND THE WORLD US Mortality and Morbidity Among Adults, Children, and Youth at Mid-Decade, 2000–2010 In the United States, not only vital statistics but also extensive survey data from several sources contribute to an ongoing assessment of the cardiovascular disease burden. The Centers for Disease Control and Prevention (CDC) operate multiple surveillance systems, including those of the National Center for Health Statistics. The National Health and Nutrition Examination Survey (NHANES) began with periodic surveys first conducted in 1960–1962 and has now become a continuous data collection process with a complete new probability sample of the civilian, noninstitutionalized population of the nation as a whole every 2 years. The Behavioral Risk Factor Surveillance System (BRFSS) is a telephone interview survey conducted annually in a probability sample of each of the 50 states and the District of Columbia. Numerous other systems collect data on specific behaviors, hospital and ambulatory medical care, and other related topics. The National Heart, Lung and Blood Institute conducts continuing multicenter population studies across the adult age range—most renowned being the Framingham Heart Study—and including the Coronary Artery Risk Development in Young Adults Study (CARDIA), the Atherosclerosis Risk in Communities Study (ARIC), the Cardiovascular Health Study (CHS), and others. Each year the American Heart Association, through the work of a committee representing these
agencies and the broader cardiovascular epidemiology community, compiles data from these and other sources into an extensive update published online (www.americanheart.org) as well as in the journal Circulation. The 2008 update concluded with summary tables illustrated by Table 1-4a, for adults, and Table 1-4b, for children and youth.4 The estimated numbers and percentages of persons in the United States as a whole and numbers or percentages by sex within categories of race/ethnicity are indicated. With few exceptions these data represent the years 2006 or 2007. To appreciate these summary data fully requires familiarity with the design and methods of each source, for which references are provided in the publication. It is important to note certain limitations, however. Not all population groups of interest are represented, such as American Indian/Alaska Natives or Asians, for whom limited available data are provided in the body of the report. Prevalence estimates are based on sample surveys and projected to the population as a whole, with whatever limitations of sampling error and bias from nonparticipation may apply. Reference to “new and recurrent” CHD and strokes reflects inability from available data to distinguish between first events and recurrences, and numbers of events could include multiple events in the same individual. Among persons age 20 or older, more than 80 million are estimated to have some form of cardiovascular disease (Table 1-4a). Included in this total are CHD, stroke, high blood pressure (HBP), and heart failure (HF), as well as other conditions. Coronary heart disease is by far the dominant condition with respect to reported deaths (more than 450,000), with
Prevalence of Selected Cardiovascular Conditions in Adults,a by Sex and Race/Ethnicity, United States, 2006 High Blood Heart Total CVDb CHDc Stroke Pressure Failure 80.0 16.8 6.5 73.6 5.7 All Adults (⫻106) Females (%) Whites 33.3 6.6 3.2 30.3 1.8 Blacks 45.9 9.0 4.1 43.9 4.2 Mexican Americans 32.5 6.3 3.8 30.4 1.4 Males (%) Whites 37.8 8.8 2.3 34.1 3.1 Blacks 45.9 9.6 3.9 44.4 4.2 Mexican Americans 26.1 5.4 2.1 23.1 2.1 a Ages ⱖ 20 years. Table 1-4a
b
CVD, cardiovascular diseases.
c
CHD, coronary heart disease.
Source: Data from Heart Disease and Stroke Statistics—2009 Update. A Report from the American Heart Association Statistics Committee and Stroke Statistics Committee. D Lloyd-Jones et al., © 2008. Courtesy of the American Heart Association/American Stroke Association.
46894_CH01_001_018.qxd
2/25/10
3:22 PM
Page 13
CURRENT BURDENS OF MAJOR CARDIOVASCULAR DISEASES IN THE UNITED STATES AND THE WORLD
Table 1-4b
13
Prevalence of Selected Cardiovascular Risk Factors in Children and Youth, by Sex and Race/Ethnicity, United States, 2007a Mean Total Meet Current Current Cholesterol Physical Activity Cigarette Concentration Recommendations Smoking (%) (mg/dl) (%) Overweight (%) (Grades 9–12)b (Ages 12–19 Years)c (Grades 9–12)b (Ages 2–19 Years)c
Females White Black Hispanicb or Mexican Americanc Males White Black Hispanicb or Mexican Americanc
22.5 8.4 14.6
165.0 162.8 163.1
27.9 21.0 21.9
29.5 39.2 35.0
23.8 14.9 18.7
154.5 161.7 158.2
46.1 41.3 38.6
31.9 30.8 40.8
a
Mean total cholesterol concentration from the National Health and Nutrition Examination Survey, 2005–2006; overweight from the National Health and Nutrition Examination Survey, 2006. bData for Hispanics. cData for Mexican Americans. Source: Data from Heart Disease and Stroke Statistics—2009 Update. A Report from the American Heart Association Statistics Committee and Stroke Statistics Committee. D Lloyd-Jones et al., © 2008. Courtesy of the American Heart Association/American Stroke Association.
stroke deaths occurring about one-third as frequently (150,000). More than one-half of the total cardiovascular disease deaths are due to coronary heart disease. Because the total prevalence for the four specified conditions would be 100.1 million, it is evident that some persons are affected by multiple conditions. High blood pressure predominates in the prevalence estimates overall and in each sex-race/ethnicity group and differs notably in prevalence among groups— from 23.1% in persons identified as Mexican American males to 43.9% in Black females. Differences between Blacks and Whites in prevalence of total CVD and several components parallel their differences in high blood pressure. The importance of these prevalence figures is that they represent the numbers of persons who continue to live with each condition, perhaps having disability or incurring substantial medical care costs, as well as high risk of recurring cardiovascular events. They also underscore the fact that mortality data, which have been considered alone up to this point in the discussion, do not provide a complete picture of the cardiovascular diseases. Information of other kinds is needed for adequate assessment of their importance in the population. It is striking that large proportions of the population—children and youth as well as adults (Tables 1-4a and 1-4b)—are affected by the indicated risk factors: tobacco use, high blood cholesterol, physical inactivity, being overweight or obesity, and diabetes. Each of these and several other factors are examined in detail in Part III.
World Income Groups and Regions, 2001 To update the global experience of cardiovascular mortality from the previous pictures of 1980 and 1985 (Figure 1-1 and Table 1-3), the recent World Bank publication, Global Burden of Disease and Risk Factors, provides estimated ischemic heart disease and cerebrovascular disease mortality as numbers of deaths and percentages of all deaths in 2001 for economic (highand low- and middle-income) and geographic regions of the world (Table 1-5).10 It is important to recall discussion of proportionate mortality in the context of the theory of epidemiologic transition, above. Here, a portion of circulatory mortality is represented by the sum of percentages of death in the two major categories shown: among low- and middle-income countries, from 47.9% in Europe and Central Asia to 6.5% in subSaharan Africa, versus 27.2% in the aggregate of highincome countries. Europe and Central Asia include the nonmarket industrial economies, in which death rates from these causes were highest among world regions in the 1985 World Bank data (Table 1-2). Together, these two components of circulatory mortality in 2001 accounted for 2,714,000 deaths in the region. But even in sub-Saharan Africa, 698,000 deaths were attributed to these two causes. They also represented more than 10 million deaths in low- and middle-income countries overall, and just over 2 million in high-income countries. These estimates indicate a substantial public health burden from ischemic heart disease and cerebrovascular disease in the developing world by the year 2001; other estimates date the establishment of these conditions in developing countries earlier.
46894_CH01_001_018.qxd
14
2/25/10
3:22 PM
Page 14
CHAPTER 1 CARDIOVASCULAR DISEASES
Table 1-5
Percentages and Numbers of Deaths Attributed to Ischemic Heart Disease and Cerebrovascular Disease, by Broad Income Group and by Region Among Low- and Middle-Income Countries, 2001 Income Group and Ischemic Heart Cerebrovascular Region Disease Disease (%) N (ⴛ1000) (%) N (ⴛ1000) Low- and Middle-Income - East Asia and Pacific 8.8 1151 14.6 1902 - Europe and Central Asia 29.7 1685 18.2 1029 - Latin America and the Caribbean 10.9 358 8.2 267 - Middle East and North Africa 16.9 323 6.8 130 - South Asia 13.6 1838 6.8 923 - Sub-Saharan Africa 3.2 343 3.3 355 All 11.8 5699 9.5 4608 High-Income 17.3 1364 9.9 781 Source: Data from Global Burden of Diseases Study, edited by AD Lopez. © 2006. The International Bank for Reconstruction and Development/ The World Bank.
Projected Cardiovascular Contributions to the Global Burden of Disease Early assessments of the global burden of cardiovascular diseases focused on available mortality data, which are often the most readily found health indicator albeit with important and sometimes severe limitations. Two significant new approaches were undertaken with a major initiative under leadership of C.J.L. Murray and A.D. Lopez, the Global Burden of Disease and Injury Series that began in 1988.11 First, it represented a major new investment in making country-level estimates for the world population both for a baseline year, 1990, and projected to 2020, thus providing insight for health policy that might reduce or avert the anticipated burdens of specific diseases. Second, the project went beyond mortality data to estimate burdens due to disability and to deaths within the working years. Extensive discussion of methods and detailed presentation of country- and regionspecific data occupy several volumes of published material from this study, including the 2006 publication that provided the data for Table 1-5. Deaths, Years Lost, and Disability Three aspects of the projections for ischemic heart disease and stroke, from 1990 to 2020, are presented
here (Table 1-6). First is the relative position of these two conditions among all major causes of death worldwide, as estimated for each of these years. Ischemic heart disease and cerebrovascular disease were found to be the first and second leading causes of death worldwide as of 1990 and were projected to remain in this rank 30 years later, in 2020. Second, years of life lost (YLL) were estimated, taking into account predicted ages at death in relation to an assumed life expectancy of 82.5 years. In these computations, ischemic heart disease was projected to advance from fourth to first rank and cerebrovascular disease from seventh to third rank between 1990 and 2020. Not only the fact of death but also the age at death is taken into account in estimating the burden. Third, in addition to years of life lost, years lived with disability of given severity is combined to yield the measure of disability-adjusted life years (DALYs). When attributed to specific causes, these can then similarly be ranked as to their contribution to disease burden for a given population. The results for ischemic heart disease and stroke were, respectively, increases from fifth to first rank and from sixth to fourth rank from 1990 to 2020. It is noteworthy that both conditions contribute importantly to disease burden in terms
Table 1-6
Rank Order of Ischemic Heart Disease and Cerebrovascular Disease as Causes of Death, Years of Life Lost (YLL), and Disability-Adjusted Life Years (DALYs), 1990 and 2020 Death YLLs DALYs Condition 1990 2020 1990 2020 1990 2020 Ischemic heart disease 1 1 4 1 5 1 Cerebrovascular disease 2 2 7 3 6 4 Source: Data from Global Burden of Diseases Study, edited by CJL Murray and AD Lopez, © 1996. Harvard School of Public Health.
46894_CH01_001_018.qxd
2/25/10
3:22 PM
Page 15
OPPORTUNITIES FOR PREVENTION
not only of death but also of reduced life expectancy and disability. Ischemic heart disease ranks first among all health conditions in all three measures by 2020 scarcely more than a decade from now. The World Health Organization has developed maps of the global distribution of DALYs lost to coronary heart disease and to stroke that are accessible at: http://www.who.int/cardiovascular_diseases/resources/ atlas/en/index.html. Productive Years of Life Lost In a further approach to gauging the population impact of death and disability attributable to cardiovascular diseases, Leeder and colleagues reported a subsequent analysis, A Race Against Time: The Challenge of Cardiovascular Disease in Developing Countries.12 The study was undertaken in an effort to put cardiovascular diseases on the world map after it was neglected, as were the chronic diseases overall, in a major report on macroeconomics and global health. Its message was that death and disability from cardiovascular diseases will strike working age populations with a devastating impact on economic development of low- and middle-income countries unless effective public health action is taken urgently. The strategy of A Race Against Time was to examine “productive years of life lost” by focusing on the projected cardiovascular disease mortality occurring in the workforce at ages from 35 to 64 years. The impact was estimated for the year 2000 and projected to 2030, in five countries—Brazil, South Africa, Russia, China, and India. The United States and Portugal were assessed as comparison countries, the former with markedly declining cardiovascular mortality and the latter with the lowest rates among the high-income countries of Europe. Details of the methods are presented in the report, and the central findings are summarized in Table 1-7. Table 1-7
Country Brazil S. Africa Russia China India US Portugal
15
The specific age group of interest, with sufficiently high cardiovascular disease mortality to be significantly affected, was persons from 35 to 64 years of age. The 30-year projections indicate a major increase in years of life lost from 2000 to 2030 for India and China and a similar relative (though much lesser absolute) increase for Brazil. The 2030 rate for the United States of 1661 years of life lost/100,000 population is not markedly less than that for China, 1863/100,000—but the relative population sizes of the countries contribute to a five-fold greater impact in China. Among 10-year age groups within the workforce, the 45–54 year age group generally experiences the heaviest burden. Among the authors’ conclusions is this central point: “that without concerted, ongoing intervention to prevent the precursors and reverse the negative effects of CVD in developing countries, a global health crisis in the current workforces (and later among the elderly) of those countries will occur––and sooner, rather than later.”12, p 84
OPPORTUNITIES FOR PREVENTION An overview of the major cardiovascular diseases as a public health challenge would be incomplete without recognizing opportunities for prevention. It is the potential impact of effective public health action that makes the challenge more than an academic interest and a matter of urgent national and global health policy. In Part IV, concepts and strategies of prevention, supporting evidence and the case for prevention, and finally a plan of public health action to prevent cardiovascular diseases on a population level are discussed in some detail. But briefly, here, before the major conditions themselves and their determinants are addressed in Parts II and III, a closing note on prevention is included for perspective.
Productive Years of Life Lost (Thousands) Due to Cardiovascular Diseases by Decade of Age and Overall, 35–64 Years, in Selected Countries, Years 2000 and 2030
Age 35–44 2000 358 112 976 1551 2260 481 13
2030 487 125 740 1768 3691 443 12
Age 45–54 2000 457 123 1427 3070 3959 714 16
2030 740 157 1420 3695 7790 741 22
Age 55–64 2000 246 67 911 2046 3002 437 12
2030 514 110 1012 4998 6456 789 19
Age 35–64 2000 1061 302 3314 6667 9221 1631 41
a
Rate/100,000 population at age 35–64.
Source: Data from A Race Against Time, © 2004, The Trustees of Columbia University in the City of New York.
2030 1742 392 3208 10460 17937 1,972 53
Ratea in 2030 1957 2667 5887 1863 3707 1661 1317
46894_CH01_001_018.qxd
16
2/25/10
3:22 PM
Page 16
CHAPTER 1 CARDIOVASCULAR DISEASES
The past half-century, in which cardiovascular epidemiology came into being, saw not only extensive documentation of the occurrence of cardiovascular diseases throughout the world but also successful investigation of its underlying causes and means of prevention. Strategies of prevention were derived from the concept of risk factors, introduced in 1961 by Dawber and colleagues in the Framingham Heart Study.13 Rose articulated most clearly, in 1981 and after, the idea of two complementary approaches to shifting adverse population distributions of risk factors toward more favorable ones, by a “high-risk” strategy of intensive intervention targeting those at the extreme of risk and a “mass“ or “population-wide” strategy to shift the whole distribution toward lower risk.14 Strasser, meanwhile, had proposed in 1978 what he termed “primordial prevention.”15 This was conceived as a means of preventing, on a worldwide front, the epidemics of risk factors themselves that Rose’s strategies were devised to reverse. In 2004, Stamler and colleagues reviewed the history of research that established the major risk factors—serum total cholesterol, blood pressure, cigarette smoking, body mass index, diabetes, and, the “pivotal” factor, adverse diet. He emphasized the concept of maintaining low risk, or absence of risk factors, in increasing proportions of the population through “safe improvements in population lifestyles, especially dietary habits from childhood on.”16,17 The concept of primordial prevention is clearly embedded in the idea of maintaining low risk, on a population-wide basis and beginning in childhood. Growing recognition of the global dimensions of the cardiovascular disease epidemic, evidenced for example in the attention paid by the World Bank, has stimulated efforts to place cardiovascular diseases and, more broadly, chronic diseases on national and global health agendas. This effort has itself been a challenge. In the United States, for example, a report in the mid-1990s on this nation’s investment in chronic disease prevention indicated that less than 3% of the aggregate budgets of state health departments, where constitutional responsibility for public health resides, and a similarly small proportion of public health personnel were dedicated to chronic disease prevention.18 A Race Against Time, published in 2004, was a rejoinder to the World Health Organization Commission on Macroeconomics and Health that failed to acknowledge the role of chronic diseases among major health-related impediments to economic development.12 The editor of The Lancet in 2005 addressed “The neglected epidemic of chronic disease” in introducing a set of reports exhorting
health policy-makers throughout the world, and especially in India and China, to take meaningful action in this area.19–23 It is significant that the recently expanded report Global Burden of Disease and Risk Factors incorporates a detailed assessment of risk factors, as well as diseases on a national, regional, and global scale.24 Among the extensive data presented are estimates of the contributions of selected risk factors to the burden (in DALYs) and mortality due to ischemic heart disease and stroke, both worldwide and separately for high- and low- and middle-income regions (Table 1-8). The population attributable fraction (PAF) for each risk factor and outcome represents the percentage of “burden” (in DALYs) or mortality that would be avoided if the lowest population risk, or “theoretical-minimum-risk exposure distribution,” rather than the actual or assumed distribution were present for the population in question. Each factor is considered separately, although factors frequently overlap in their occurrence. As a result, the cumulative percentages for multiple risk factors may exceed 100%. Mazzati and coauthors present details of methods for estimating joint effects of multiple risk factors (Joint PAF). For the risk factors considered individually, high blood pressure stands out as the leading factor for stroke and is nearly equivalent to high cholesterol for ischemic heart disease. For example, if the blood pressure distribution of the high-income region were reduced to the theoretical-minimum-risk exposure (estimated to be 115 mm Hg with a standard deviation of 6 mm Hg), the burden of stroke in DALYs would be expected to be 56% lower for the region. Attributable fractions for blood pressure and cholesterol, as well as for smoking and alcohol use, are higher in high-income than in low- and middleincome regions, although the differences in PAF for high blood pressure are negligible. Low fruit and vegetable intake and physical inactivity contribute more strongly to risk of ischemic heart disease than to stroke. Urban air pollution, though minor in relation to other factors, does contribute to both outcomes, to a greater degree in the low- and middle-income than the high-income region. Importantly, the joint contribution of these risk factors accounts for the great majority (80%) of ischemic heart disease as well as the majority (60–70%) of stroke, with only minor differences between economic regions. This glimpse of evidence based on epidemiologic observations suggests a vast potential for prevention, if public health strategies can be devised, implemented, and sustained to preserve low risk or restore
46894_CH01_001_018.qxd
2/25/10
3:22 PM
Page 17
REFERENCES
17
Table 1-8
Individual and Joint Contributions of Risk Factors to Ischemic Heart Disease and Stroke, World Total and by Broad Income Group Ischemic Heart Disease Stroke Factor PAF for individual factor (%) PAF for individual factor (%) HighLow- and Middle- World HighLow- and Middle- World Income Income Income Income High blood pressure 48 44 45 56 54 54 High cholesterol 57 46 48 25 15 16 Overweight and obesity 27 16 18 20 10 12 Low fruit and vegetable intake 19 30 28 9 11 11 Physical inactivity 21 21 21 8 6 7 Smoking 23 15 17 21 12 13 Alcohol use 13 4 2 11 5 3 Urban air pollution 1 2 2 1 4 3 -------------------------------------------------------------------------------------------------------------------Joint PAF—burden (%) 84 80 80 68 64 65 Joint PAF—mortality (%) 80 78 79 54 61 60 Source: Data from Global Burden of Diseases Study, edited by AD Lopez. © 2006. The International Bank for Reconstruction and Development/ The World Bank.
the more favorable distributions of risk that can be presumed to have existed historically. Here, then, is the global challenge, to be addressed country by country: to recognize and acknowledge the immense burden of chronic diseases, and of cardiovascular diseases in particular, and the need for concerted public health action to achieve the demonstrated potential for major reductions in risk. REFERENCES 1. World Health Organization. International Statistical Classification of Diseases and Related Health Problems. 10th rev. Geneva (Switzerland): World Health Organization; 1992. 2. Pearson TA, Jamison DT, Trejo-Gutierrez J. Cardiovascular disease. In: Jamison DT, Mosley WH, Measham AR, Bobadilla JL, eds. Disease Control Priorities in Developing Countries. Oxford (England): Oxford University Press; 1993:577–594. 3. Omran AR. Epidemiologic transition in the United States: the health factor in population change. Population Bulletin. 1977;32:1–42. 4. Lloyd-Jones D, Adams R, Carnethon M, et al. Heart disease and stroke statistics—2009 update. A report from the American Heart Association Statistics Committee and Stroke Statistics Subcommittee. Circulation. 2009;119:e1–e161.
5. Omran AR. The epidemiological transition: a theory of the epidemiology of population change. Milbank Q. 1971;49:509–538. 6. Olshansky SJ, Ault AB. The fourth stage of the epidemiologic transition: the age of delayed degenerative diseases. Milbank Q. 1986;64: 355–391. 7. Gaziano TA, Reddy KS, Paccaud F, Horton S, Chaturvedi V. Cardiovascular disease. In: Jamison DT, Breman JG, Measham AR, Alleyne G, Claeson M, Evans DB, et al., eds. Disease Control Priorities in Developing Countries. 2nd ed. Washington, DC: International Bank for Reconstruction and Development/The World Bank; 2006:645–662. 8. Dodu SRA. Emergence of cardiovascular diseases in developing countries. Cardiol. 1988; 75:56–64. 9. World Health Organization. Preventing Chronic Diseases: A Vital Investment. Geneva (Switzerland): World Health Organization; 2005. 10. Mathers CD, Lopez AD, Murray CJL. The burden of disease and mortality by condition: data, methods, and results for 2001. In: Lopez AD, Mathers CD, Ezzati M, Jamison DT, Murray CJL, eds. Global Burden of Disease and Risk Factors. Washington, DC: The
46894_CH01_001_018.qxd
18
2/25/10
3:22 PM
Page 18
CHAPTER 1 CARDIOVASCULAR DISEASES
International Bank for Reconstruction and Development/The World Bank; 2006:45–240. 11. Murray CJL, Lopez AD. Alternative visions of the future: projecting mortality and disability, 1990–2020. In: Murray CJL, Lopez AD, eds. The Global Burden of Disease: A Comprehensive Assessment of Mortality and Disability from Diseases, Injuries, and Risk Factors in 1990 and Projected to 2020. Boston, MA: The Harvard School of Public Health; 1996. 12. Leeder S, Raymond S, Greenberg H. A Race Against Time: The Challenge of Cardiovascular Disease in Developing Countries. New York: The Trustees of Columbia University in the City of New York; 2004. 13. Kannel WB, Dawber TR, Kagan A, Revotskie N, Stokes III, J. Factors of risk in the development of coronary heart disease—six-year follow-up experience: the Framingham Study. Ann Intern Med. 1961;55:33–50. 14. Rose G. Strategy of prevention: lessons from cardiovascular disease. Br Med J. 1981;282: 1847–1851. 15. Strasser T. Reflections on cardiovascular diseases. Interdisc Sci Rev. 1978;3:225–230. 16. Stamler J. Established major coronary risk factors: historical overview. In: Marmot M, Elliott P, eds. Coronary Heart Disease Epidemiology: From Aetiology to Public Health. 2nd ed. Oxford (England): Oxford University Press; 2005:18–31. 17. Stamler J, Neaton JD, Garside DB, Daviglus ML. Current status: six established major risk factors—and low risk. In: Marmot M, Elliott P, eds. Coronary Heart Disease Epidemiology:
From Aetiology to Public Health. 2nd ed. Oxford (England): Oxford University Press; 2005:32–70. 18. National Center for Chronic Disease Prevention and Health Promotion, Centers for Disease Control and Prevention. Unrealized Prevention Opportunities: Reducing the Health and Economic Burden of Chronic Disease. Bethesda, MD: Public Health Service, US Dept of Health and Human Services; 1997. 19. Horton R. The neglected epidemic of chronic disease. Lancet. 2005;366(9496):1514. 20. Strong K, Mathers C, Leeder S, Beaglehole R. Preventing chronic diseases: how many lives can we save? Lancet. 2005;366(9496): 1578–1582. 21. Epping-Jordan JE, Galea G, Tukuitonga C, Beaglehole R. Preventing chronic disease: taking stepwise action. Lancet. 2005;366(9497): 1667–1671. 22. Reddy KS, Shah B, Varghese C, Ramadoss A. Responding to the threat of chronic diseases in India. Lancet. 2005;366(9498):1744–1749. 23. Wang L, Kong L, Bai Y, Burton R. Preventing chronic disease in China. Lancet. 2005;366 (9499):1821–1824. 24. Mazzati E, Vander Hoorn S, Lopez AD, et al. Comparative quantification of mortality and burden of disease attributable to selected risk factors. In: Lopez AD, Mathers CD, Ezzati M, Jamison DT, Murray CJL, eds. Global Burden of Disease and Risk Factors. Washington, DC: The International Bank for Reconstruction and Development/The World Bank; 2006:241–396.
46894_CH02_019_038.qxd
2/25/10
3:23 PM
Page 19
C H A P T E R
2 Distributions and Disparities mension to the picture of a far-reaching epidemic at local, national, and global levels. Data shown in Chapter 1 illustrate variation in cardiovascular disease burden among the World Bank regions. Such distributions are widely regarded as indicating global disparities in health, disfavoring low- and middle-income countries. Several aspects of place and time are considered as background for topics that follow.
SUMMARY Age, sex, and race are three personal characteristics used in describing the epidemiology of virtually every disease and condition. They are part of the epidemiology of atherosclerotic and hypertensive diseases and each of their underlying determinants. From an epidemiologic perspective, distributions by age, sex, and race demonstrate variation or patterns of interest in occurrence of diseases or determinants within or between populations. These patterns are regarded as “clues to causation” or signs of potential for prevention. From a public health perspective, such variation often represents something more: disparities, or inequities, in health and disease. Here the perception is that one or more groups within a society— elderly, women, or racial/ethnic minorities, for example—may bear a disproportionate, and intrinsically unfair, burden. Each of these perspectives contributes to shaping public health approaches to prevention, in cardiovascular disease as elsewhere. Although age, sex, and race are commonplace in epidemiology, each of them still deserves attention from the start as to points of definition, ascertainment, and interpretation. The triad of person, place, and time (where “person” is represented by age, sex, race, and other personal characteristics) is also an epidemiologic convention. Adding place, or geography, to the picture of cardiovascular diseases reveals further patterns that convey local, regional, or global variations of both epidemiologic and public health importance. Why is one region, such as the southeastern United States, more heavily afflicted by stroke than elsewhere in the same country? Mapping occurrence of cardiovascular diseases over time adds another important di-
INTRODUCTION Epidemiology and Public Health Knowing the distribution of a disease within or between populations is basic to its epidemiologic understanding. Epidemiology is frequently characterized as study of the distribution of disease in relation to person, place, and time. One tenet of epidemiology is that patterns in the distribution of a disease may suggest “clues to causation” and possibilities for prevention. Epidemiology documents the occurrence of cardiovascular diseases in relation to personal characteristics, beginning with age, sex, and race (each to be defined later). Taking into account other personal characteristics as well as place and time extends the picture of disease distribution and deepens understanding of its occurrence. Some examples follow. Death rates for coronary heart disease in the United States in 2004 are shown in Figure 2-1.1 They present a striking pattern of variation in rates with age. A several-fold increase is apparent across age categories, from youngest (ages 35–44 years) to oldest (ages 75–84 years). Clear patterns are also evident by sex: The rates are greater for males than for females in each age category. By race, the rates are
19
46894_CH02_019_038.qxd
20
2/25/10
3:23 PM
Page 20
CHAPTER 2 DISTRIBUTIONS AND DISPARITIES
2,000 Black Male
Deaths/ 100,000 Population
White Male Black Female
1,500
White Female
1,000
500
0 35244
45254
552 64 Age (Years)
65274
75284
Figure 2-1 Death Rates for Coronary Heart Disease by Age, Race, and Sex, U.S., 2004. Source: Morbidity & Mortality: 2007 Chart Book on Cardiovascular, Lung and Blood Diseases. National Institutes of Health, National Heart, Lung and Blood Institute. June 2007.
greater for Blacks than for Whites in each age category and for both sexes. On careful inspection, the picture reveals another significant feature: any given death rate, for example 500 or 1000 deaths/100,000 population, is reached by men at an age level roughly 10 years younger than the age at which women reach the same rate. This “lag in CHD death rates” for women is a commonly observed pattern that remains to be adequately explained. A geographic pattern is also apparent when death rates are compared among the states. In Figure 2-2, categories of death rates for cardiovascular diseases (here including coronary heart disease, stroke, and other cardiovascular causes of death, excepting congenital heart disease) are represented by states for the combined years 2001–2003.1 The map reveals a concentration of highest cardiovascular mortality mainly in southeastern states but extending into the north, a picture that might stimulate a number of speculations about the cause. In these and many other examples to follow, the influences of age, sex, race, and place, or geography, are described. Variation over time, the third aspect of the triad, was illustrated in Chapter 1 by decadeslong trends in cardiovascular mortality. Each of these dimensions of “distributions” will be discussed below from an epidemiologic perspective, as background for addressing the major cardiovascular diseases and their determinants.
From a public health perspective, distributions of disease have a further implication. Some groups within a population, or some populations, experience a greater disease burden than do others. Such differences in disease burden are often considered to represent “disparities,” in a sense beyond the simple fact of variation in disease frequency. According to Braveman, “Health disparities/inequalities are potentially avoidable differences in health (or in health risks that policy can influence) between groups of people who are more and less advantaged socially; these differences systematically place socially disadvantaged groups at further disadvantage on [sic] health.”2, p 180 This sense of “disparities” has a particular meaning in public health. It goes beyond mere variation in disease frequency, or the epidemiologic concept of distribution, as expressed by the National Institutes of Health: “Health disparities are differences in the incidence, prevalence, mortality, and burden of diseases and other adverse health conditions that exist among specific population groups in the United States.”3 Explicit attention to who is affected most, and the special concern when—as is very often the case—it is “socially disadvantaged” groups, leads to consideration of “health equity,” or justice, in the arena of health and society. When addressed in British or European literature, social or socioeconomic position is often understood as the measure of relative disadvantage.2
46894_CH02_019_038.qxd
2/25/10
3:23 PM
Page 21
INTRODUCTION
21
Age-Adjusted Death Rates for Cardiovascular Diseases* by State, U.S., 2001⫺2003
Deaths /100,000 Population 336⫺425 (14) 300⫺335 (12) 280⫺299 (13) 235⫺279 (11)
* Excludes congenital malformations of the circulatory system.
Figure 2-2 Age-Adjusted Death Rates for Cardiovascular Diseases by State, US, 2001–2003. Source: Morbidity & Mortality: 2007 Chart Book on Cardiovascular, Lung and Blood Diseases. National Institutes of Health, National Heart, Lung and Blood Institute. June 2007.
In the United States, health objectives for the nation are presented for each decade in successive Healthy People reports. Eliminating health disparities has become an overarching goal, second only to increasing life expectancy and quality of life for all. Disparities are defined in relation to several population characteristics: “The second goal of Healthy People 2010 is to eliminate health disparities among segments of the population, including differences that occur by gender, race or ethnicity, education or income, disability, geographic location, or sexual orientation.”4, p 11 In addition to this broad goal, in each of several areas such as heart disease and stroke (Focus Area 12), disparities are addressed in relation to specific objectives for the decade. The 2010 objectives include targets to be achieved for each demographic group, usually as defined by age, sex, and race. The more stringent targets are often set for one or more of the racial/ethnic minorities. Special efforts to achieve the targets for these groups are implied, above and beyond those required to improve health for the population as a whole.
This public health perspective can be developed further, as reviewed by Kumanyika and Morssink.5 They argue against defining health disparities chiefly in terms of excess deaths from any particular cause, as is often done. They suggest that this approach is inadequate to address the wide-ranging public health consequences of social disadvantage. In part this leads to an undue focus on healthcare services and individual-level approaches, whereas a broader concept of health of the population would hold greater promise. The focus on cause-specific excess deaths is represented as limited in scope in contrast with a population health focus that invokes a wide range of community attributes in their social and environmental context. The public health perspective on disparities and its implications are raised here alongside the epidemiologic perspective on distributions because these fundamental characteristics—age, sex, and race, and person, place, and time—are important from both points of view. They provide the basic frame of reference for understanding the distribution, determinants,
46894_CH02_019_038.qxd
22
2/25/10
3:23 PM
Page 22
CHAPTER 2 DISTRIBUTIONS AND DISPARITIES
and means of prevention of the atherosclerotic and hypertensive diseases, and they point to issues regarding public health action. Dimensions of Population Diversity Two additional points warrant comment before each of these basic characteristics is discussed in greater detail. The first is population diversity with respect to age, sex, and race. More adequate epidemiologic data are needed on cardiovascular diseases for population groups underrepresented in health surveys and other data sources. If data are not sufficient to describe the health status of specific groups, the nature and extent of variation and disparities remain unknown, with consequent limitations on both epidemiology and public health response. Validity of generalizing results of intervention studies from one population group to others is another concern. Whether results for White men of middle age are applicable to women, non-Whites, or older or younger persons is often a matter of judgment without directly relevant data. Broader representation of women as well as men and minorities as well as Whites in future studies can address this issue in the long run, but in the interim a dilemma remains regarding whether a particular intervention should be applied generally on the presumption of universal benefit and safety or whether this should await direct evidence within each specific group of concern. For both reasons, each of these aspects of population diversity has received increasing emphasis in connection with health research in general, including both clinical and population studies of cardiovascular diseases. Specifically in cardiovascular epidemiology, many of the early population studies focused on middle-aged White men. There are important counterexamples, such as the Framingham Heart Study of men and women, the Tecumseh Study of a whole community at all ages, and the Evans County Study in Black adults. But a landmark report in 1978 on the associations of cigarette smoking, blood pressure, and cholesterol concentrations with coronary heart disease event rates, from the US National Pooling Project, was necessarily based on data exclusively for White men aged 40–59 years at entry to the respective studies.6 This was the only age-sexrace stratum among contributing studies with sufficient numbers of events for detailed analysis. Although based on rational design considerations of the time, the insufficiency of corresponding data for women became a prominent issue in more recent years. The need for much more extensive data for
older and younger persons, for women, and for nonWhite groups became much more fully appreciated in the late 1980s and since. This recognition, reinforced by social and political influences, has led to explicit emphasis in policies of the US National Institutes of Health, for example, on inclusion of women and minorities in clinical and population research.7 Modifiability of These Factors One further aspect of age, sex, and race is a common presumption that each of these is “unmodifiable,” contrasting in principle with such factors as dietary patterns or physical activity. But differences in disease distribution by any of these characteristics could reflect underlying social or environmental factors— social conditions, behavioral patterns, specific exposures, or others. To this extent, health patterns by age, sex, or race may in fact point to modifiable characteristics. For example, a rise in population mean levels of blood pressure was long regarded as a natural or inevitable concomitant of aging. But epidemiologic studies have repeatedly shown that populations differ widely in degree of rise in blood pressure with age, a pattern that is altogether absent from some populations. Thus age, as it predicts blood pressure levels, is not strictly “unmodifiable.” It is instead a marker for other factors whose modification can counter a tendency for blood pressure to rise with age. Similar relationships for sex and race suggest the value of identifying “cofactors” that may permit mitigating apparent effects of age, sex, or race. It may be useful to bear in mind these concepts of population diversity and modifiability of the “unmodifiable” factors, as age, sex, and race are examined more closely.
AGE AND LIFE STAGES Definition and Classification A few considerations suggest that “age” is not always so simple an attribute as it might appear. The range of ages relevant to cardiovascular diseases extends throughout the life span, from conception to the oldest attained ages: from determination of genetic makeup and the course of fetal development to the oldest ages with the greatest risks of coronary events, strokes, and other major cardiovascular conditions. Quantitative measures of gestational age are usually expressed in weeks from the mother’s last menstrual period to a specified date; in the early postnatal period
46894_CH02_019_038.qxd
2/25/10
3:23 PM
Page 23
AGE AND LIFE STAGES
in days, weeks, or months since birth; and thereafter in completed years, that is, age in years at the most recent birthday. Categories of age are expressed either in class intervals of the quantitative measures or in any of several qualitative terms, including the preadult categories of fetus, newborn, infant, child, adolescent, or youth (with some overlap) and young, middle-aged, older, or elderly adults. Often these terms are used without definition, so their actual age correspondence is unclear. Other terms related to physiological or social aspects of age include developmental age (as assessed by secondary sexual characteristics or skeletal age), age at menarche or menopause, age at majority or attainment of legal responsibility, reproductive age, retirement age, and others. Ascertainment Age is generally ascertained by determining from available records or by questionnaire the age of the subject or respondent, in years, as of the last birthday. A few special considerations sometimes arise. Ascertainment of gestational age is subject to uncertainty in the timing of the last menstrual period. Date of birth by day, month, and year is usually known and can often be confirmed from a birth certificate or other official record (although there are populations for which documentation is unavailable). Knowledge of the actual birth date is obviously more precise than age at last birthday, given that exact age as of any given date can then be calculated and expressed in decimal years (e.g., 10.3 years). This level of accuracy can be important in studies in childhood and adolescence. In this period, change in characteristics of interest can be so rapid that age at last birthday is too imprecise for sound interpretation of the results. In some circumstances, such as death or incapacity, knowledge of an individual’s age may depend on the report of a relative or other surrogate informant, whose reliability may be unknown. In general, however, accuracy of classification by age is expected to be satisfactory absent intentional misrepresentation. Age Adjustment and Standardization Most health-related phenomena are strongly age dependent. Therefore, groups with dissimilar age composition typically exhibit different patterns of these conditions. Accordingly, any differences in such patterns between populations might merely reflect underlying dissimilarities in age distributions and offer only spurious hints of more meaningful differences. In most populations, to use a now familiar example,
23
blood pressure does increase with age in adulthood; if the prevalence of high blood pressure were compared between two populations with different proportions of older members, a finding of a higher prevalence in the population with the greater proportion of older adults would be expected and uninformative. For this reason, comparisons within specific age groups are more informative than “crude” comparisons that do not take age into account. Comparison of prevalence of high blood pressure between two populations is preferably based on rates for age strata of 10 years or less, such as ages 30–39, 40–49, and so on. A summary measure of overall population prevalence is often useful for comparing two or more populations. In that case, these age-specific data can be used to adjust for differences in the age distribution of each population, resulting in a single “ageadjusted” or “age-standardized” value for each population to be compared. This value, for example, for prevalence of high blood pressure, is not likely to be the true prevalence in any of the populations, but it can be compared with values calculated for other populations with the same standard population, with knowledge that the comparison is not distorted by differences in age composition among the populations. Ahlbom and Novell present a simple demonstration of this important principle, which can be applied to other characteristics as well as age.8 This is the implied underpinning of data presentations identified as “agestandardized” or “age-adjusted,” as already seen for the example in Figure 2-2. Another example of different representations of age and the meanings they convey is shown in Figure 2-3.1 Deaths for major cardiovascular diseases each year from 1979–2004 are represented in two ways: The bar graph or histogram shows the number of deaths each year (in 1000s) in relation to the scale on the left; the curve represents the age-adjusted death rate each year (per 100,000 population), in relation to the scale on the right. The numbers of deaths were roughly constant, nearly 1 million per year, from 1979 to 2004. The death rate declined substantially, from more than 500 to just under 300 deaths/100,000 per year. Did cardiovascular deaths decline or not? The answers are yes for the rates, calculated as though the age composition of the population were constant over these years, and no for the actual numbers of deaths because the increasing proportion of older persons in the population (with their very high cardiovascular mortality—see Figure 2-1) offset the decline in age-adjusted rates.
46894_CH02_019_038.qxd
3:23 PM
Page 24
1,200
600
1,000
500
800
400
600
300
400
200
200
100
0
79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 00 01 02 03 04 Years
Deaths/100,000 Population (Line)
CHAPTER 2 DISTRIBUTIONS AND DISPARITIES
Deaths in Thousands (Bar)
24
2/25/10
0
Figure 2-3 Deaths and Age-Adjusted Death Rates for Cardiovascular Diseases, US, 1979–2004. Source: Morbidity & Mortality: 2007 Chart Book on Cardiovascular, Lung and Blood Diseases. National Institutes of Health, National Heart, Lung and Blood Institute. June 2007.
Interpretation of Health Patterns by Age The Fetal and Neonatal Period The fetal and neonatal periods have been the focus of research suggesting that conditions of life in this period may be critical for future risks of adult cardiovascular diseases.9 For example, in a cohort of British men born in the early 1900s, associations were demonstrated between low birth weight, or limited weight gain to age 1 year, and adult cardiovascular mortality. This and related observations have been interpreted as reflecting adverse social conditions that influence fetal and neonatal growth and development. Under this theory, the developing metabolic and physiologic systems of the fetus become “programmed” in response to these conditions in a manner that unfavorably affects risk factor development in later life. This theory is addressed further in the context of social and environmental conditions (Chapter 16). Childhood and Youth Childhood, adolescence, and “youth” (defined by the World Health Organization as spanning the ages from birth to 24 years10) are conventional categories of preadult ages. For this period of life, chronologic age is only an approximate indicator of biological age
because of wide variation between individuals in growth tempo, or rates of growth and maturation.11 Much attention has been devoted to cardiovascular risk factors in these age groups in the past three decades or more. As a result, the view is increasingly accepted that prevention of the risk factors and early manifestations of atherosclerotic and hypertensive cardiovascular diseases requires intervention before adulthood. The basis of this view includes several links between factors measured in youth and observations in adults, as follows: familial aggregation—the greater similarity of risk factor levels of parents and offspring within families than between them; familial concordance—the relatively adverse risk factor levels of offspring of parents with cardiovascular disease when compared with those of unaffected parents; tracking—the tendency for risk factor levels in an individual at a given age in childhood or adolescence to be predictive of levels at later ages; correlations with vascular pathology—associations between risk factor levels and the extent of atherosclerosis in the aorta and coronary arteries, in childhood, adolescence, and youth. The current view of the potential for preventive measures in this period of life is illustrated by exten-
46894_CH02_019_038.qxd
2/25/10
3:23 PM
Page 25
AGE AND LIFE STAGES
sive reviews and recommendations from the World Health Organization (WHO) and the American Heart Association, among others.10,12 Young and Middle Adulthood Studies of the atherosclerotic and hypertensive diseases have most often focused on middle adulthood. This emphasis was based primarily on a sufficient frequency of detectable disease events in initially healthy subjects, the feasibility of long-term followup from first observation through a period of years at moderately high risk, and other design considerations. Older people experience much higher event rates but competing risks of other diseases and death may confuse the picture; younger people offer the greatest potential for true prevention of advanced atherosclerosis but experience few clinically detectable events. Methods for detection of subclinical disease are especially applicable in youth and early adulthood and include, for example, measures of endothelial function, arterial stiffness and calcification, and echocardiography. These methods provide intermediate measures that can be useful both as outcomes of earlier influences and as predictors of later clinical disease. The period of young and middle adulthood remains important for continued study, although ideally in continuity with observations and development of preventive strategies throughout life. The Elderly The view that chronological age is only a surrogate for modifiable underlying characteristics implies that many characteristics of older persons are not inevitable concomitants of aging. Consistent with this view, current concepts of aging focus not on an inexorable progression of disease but on the potential for preserving maximum functional capacity, independence of living, and quality of life.13 Greatly increasing numbers of persons in the United States and many other countries survive myocardial infarction or stroke only then to experience chronic ischemic heart disease, congestive heart failure, or vascular dementia, with significant disability and dependence. A major question is the extent to which these common occurrences can be prevented, resulting in better health and quality of life in the later years. Better understanding of how to preserve low risk throughout life is therefore important. In many populations, life expectancy is increasing and proportions of persons attaining advanced ages are growing, with profound impacts on society and on health and disease. Recognizing this, the World
25
Health Organization convened a Study Group on Epidemiology and Prevention of Cardiovascular Diseases in Elderly People.14 The Study Group Report is a valuable resource for assessing the cardiovascular disease situation of the “elderly” (those aged 65 years and older) with respect to the occurrence of cardiovascular diseases and their prevention, rehabilitation, and related health policy. In addition, the report demonstrates the projected growth of older age groups as proportions of the population of each World Health Organization Region (Table 2-1). The projected aging of Europe is striking, with more than 20% of the population over age 65 by 2025. North America is not far behind. Substantial increases are projected everywhere but in sub-Saharan Africa. The health of the elderly, not least their cardiovascular health, becomes an ever greater concern. It should also be noted that, in addition to usually higher rates, morbidity and mortality data in the elderly differ from those for younger adult age groups in two other respects. First, data on specific diagnoses and causes of death may be less reliable owing to multiple coexisting health conditions in the elderly. Second, the high prevalence of many cardiovascular disease risk factors in older persons tends to result in reduced estimates of relative risk, or the apparent force of a factor such as blood pressure on the risk of an event. This is because even those who remain clinically free of these events have, on average, higher risk than younger persons. The absolute risks attributable to a given factor, on the contrary, are greater than in younger persons owing to the much higher disease rates at older ages. Interpretation of measures of risk among older adults therefore requires special caution. This will be illustrated subsequently in connection with the relation of blood pressure to risk of stroke in different age groups (Chapter 5). A recurring issue regarding the progression from earliest onset to latest outcomes of the atherosclerotic and hypertensive diseases is the inability to study the full process in one population continuously through the life span. For example, only a few studies have followed this process from childhood into early adulthood. The exceptions are highly informative; but at the same time, the factors in childhood and adolescence that were influential for a particular adult population may have changed over time and may have limited relevance to a subsequent generation. Such potential “cohort effects” must be considered in interpreting long-term trends across age groups.
46894_CH02_019_038.qxd
26
2/25/10
3:23 PM
Page 26
CHAPTER 2 DISTRIBUTIONS AND DISPARITIES
Table 2-1 Region
Projected Proportion of the Population Aged 65 Years and Over, 1990–2025 Percentage of the Population Year ⬎ 65 Years ⬎ 75 Years ⬎ 80 Years
Europea
1990 2010 2025
13.7 17.5 22.4
6.1 8.4 10.8
3.2 4.9 6.4
North America
1990 2010 2025
12.6 14.0 20.1
5.3 6.5 8.5
2.8 4.0 4.6
Oceania
1990 2010 2025
9.3 11.0 15.0
3.6 4.8 6.6
1.8 2.8 3.6
Asiaa
1990 2010 2025
4.8 6.8 10.0
1.5 2.5 3.6
0.6 1.2 1.8
Latin America/Caribbean
1990 2010 2025
4.6 6.4 9.4
1.6 2.4 3.6
0.8 1.2 1.8
Eastern Mediterranean/North Africa
1990 2010 2025
3.8 4.6 6.4
1.2 1.6 2.2
0.5 0.8 1.1
Sub-Saharan Africa
1990 2010 2025
2.7 2.9 3.4
0.7 0.8 1.0
0.3 0.3 0.4
a
Data exclude countries of the former USSR.
Source: Kinsella K, Tauber CM. An aging world II. Washington, DC, US Government Printing Office, 1992 (US Bureau of the Census, International Population Reports, P-95, No. 79).
Finally, it is important to note a set of populationbased studies in the United States, supported by the National Heart, Lung and Blood Institute, spanning the adult years—in sequence by age, the Coronary Artery Risk Development in Young Adults Study (CARDIA), the Atherosclerosis Risk in Communities Study (ARIC) of middle-aged adults, and the Cardiovascular Health Study (CHS) in the elderly— all discussed in subsequent chapters.
SEX OR GENDER Definition, Classification, and Ascertainment In the cardiovascular arena, there is little ambiguity about classification as male or female, and both the terms “sex” and “gender” are used. “Gender,” the grammatical term, and “sex,” the biological term,
are used here interchangeably, although “sex” is more frequently found in presentations of epidemiologic data. Sex-Specific Observations Epidemiologic observations are nearly always reported on a sex-specific basis, when data for both males and females are available. However, when the numbers of observations are insufficient for sexspecific analysis, combined rates may be presented instead. In such instances, comparisons between populations or groups that differ in composition by sex may, like those for age, be distorted unless adjustment is made for this difference. Interpretation of Differences in Health by Sex A widely recognized example of sex differences in the occurrence of cardiovascular disease is the “lag” of 10 years or more in coronary mortality by age in women
46894_CH02_019_038.qxd
2/25/10
3:23 PM
Page 27
RACE OR ETHNICITY
relative to men. This lag was noted, for example, in describing Figure 2-1. Such sex differences are interpreted variously as evidence of biological, social, cultural, or behavioral effects. Plausible bases for such effects could include chromosomal, hormonal, or reproductive factors; sex-specific social roles leading to different patterns of activity, occupation, and interpersonal relationships, and therefore different exposures; or sex-related perceptions or practices that determine access to education, health information, or health services. These factors can be viewed as either protective for women or hazardous for men. Sex differences in distributions of blood lipids and blood pressure as well as body fatness are apparent beginning in adolescence, and these and other risk factors progress differently between women and men throughout adulthood. The relative lack of information about several aspects of cardiovascular diseases in women—including evaluations of preventive strategies, diagnostic test performance, and responses to medical and surgical treatments—has been addressed with increasing frequency. Reviews include the report of the National Heart, Lung and Blood Institute conference, Cardiovascular Health and Disease in Women,15 a text of the same title,16 and the American Heart Association’s Report of the Special Writing Group on Cardiovascular Disease in Women and guidelines for cardiovascular disease prevention in women, first published in 2002 and updated in 2007.17,18 An essay by Barrett-Connor is particularly thought-provoking in bringing attention to long-held and untested theories of sex differences in coronary heart disease.19 This emphasis reflects in large part a sense of relative neglect of the problem of coronary heart disease in women. The cumulative lifelong coronary mortality of women is actually no less than that for men in the United States and several other populations. Misapprehension of this fact results from a longstanding failure to recognize that the greater life expectancy of women than men places more women than men at the highest risks of coronary heart disease death. A critical review of sex differences in coronary heart disease posed the question of what role sex hormones—either exogenous or endogenous—could actually be playing, especially at older ages where coronary heart disease incidence and mortality are greatest.20 Gradients in coronary mortality across countries are closely parallel for women and men, and rates are higher among women in the highestrate countries than among men in the lowest-rate countries (see, for example, Figure 2-6, as follows, for the 45- to 64-year age stratum). Thus, coronary mor-
27
tality for women is not universally low but varies widely in different environments. This view suggests that environmental factors are paramount determinants of these population differences and that intrinsic hormonal differences between the sexes may have a lesser role than has often been presumed. The continuing uncertainties about the meaning of these and other sex differences in atherosclerotic and hypertensive diseases add to the importance of further study of women as a population of special concern. Risks and Rates in Successive Periods of Life Sex differences in risk factors are recognized in the first months of life but become clearer with adolescence and the many changes associated with puberty. Age-specific mean values for systolic and diastolic blood pressure and for each component of the blood lipid profile differ by sex from puberty onward reflecting distinct time patterns of development. Tobacco use progresses rapidly in this period and tends to remain somewhat less prevalent among females than among males at successive ages. In the age interval from 15 to 34 years, atherosclerosis of the coronary arteries at the fatty streak stage is about equally frequent between females and males, but raised lesions are notably less frequent among females.21 Through early and middle adulthood, women tend to have lower age-specific mean values for blood pressure and total cholesterol concentration, but greater prevalence of physical inactivity and being overweight.16 In middle age, although rates for coronary heart disease events increase for women more gradually than for men, rates of stroke for women closely parallel those for men. In later adulthood, both total cholesterol concentration and blood pressure (especially systolic pressure) increase to values greater than those for men of the same age. Rates of coronary heart disease and stroke both increase sharply among women and men with further increase in age, as do rates of other complications of advanced atherosclerosis and hypertension.
RACE OR ETHNICITY Definition and Classification If “race” once denoted a specific biologic concept and “ethnicity” a cultural/anthropological one, the earlier distinction between terms has become less clear. A common convention, “race/ethnicity” or “race and ethnicity,” is generally used here. Because isolated populations with very limited genetic admixture and
46894_CH02_019_038.qxd
28
2/25/10
3:23 PM
Page 28
CHAPTER 2 DISTRIBUTIONS AND DISPARITIES
little contact with other cultures are rare, the concept of genetically or culturally homogeneous groups may no longer be meaningful. It is nonetheless true that when groups of persons identify themselves or are classified by others as members of particular racial or ethnic groups, they often differ in disease rates or risk factor distributions. In the United States, classifications of racial or ethnic minority populations have changed in recent practice. One such system distinguishes Black Americans, Hispanics, Native Americans, and Asians/ Pacific Islanders.22 Other classifications include “color,” which differentiates among non-Hispanic Blacks, non-Hispanic Whites, and Mexican Americans. “Underprivileged minorities” is intended to distinguish some racial/ethnic groups, characterized as being of low socioeconomic status by income, education, or occupation, from others. Country by country, and from one situation to another, group distinctions may be made on various bases, such as religion, national origin, or others. Group membership defined by race/ethnicity may indicate genetic, cultural, socioeconomic, or other differences, and interpretation of related differences in health and disease may require some depth of investigation. Ascertainment In routine recording of health data or in population studies, assignment of race or ethnicity is often based on individual self-report. Classification by simple observation may be unsatisfactory and unreliable, given that parental origins, language preference, and other considerations are difficult to take into account in this approach. In studies of children, the reported race or ethnicity of parents may be used to classify offspring. Inconsistencies in methods between data sources can introduce potential error. For example, when the numerator of a rate is based on one data system, such as death certificates, and the denominator is based on another, such as census data, distortions in the resulting rates may occur. For example, this is at issue in interpreting mortality data for Mexican Americans, for whom individuals may be “Mexican American” by self-identification in census data but simply “White” as often described on death certificates. Mortality for Mexican Americans would be underestimated in this case. State-level mortality statistics have been cited as indicating comparatively low coronary mortality among Mexican Americans (largest of the Hispanic subgroups). However, results of community surveillance for coronary heart disease “on the ground” in Corpus Christi/Nueces County, Texas, have shown higher rates of hospitalization, higher
case-fatality, and higher long-term mortality for coronary disease among Mexican Americans than among non-Hispanic Whites in the same community.23 The discrepancy may result from distortion of the vital statistics data in this way. Interpretation of Differences in Health by Race or Ethnicity If such classifications identify groups at special risk, it is desirable from a public health perspective that this information be obtained and applied appropriately. This implies looking beyond a genetic or cultural basis for the observed differences. Just as age and sex may be markers for any of a variety of group differences, so may race and ethnicity have underlying bases—some known, others unknown. Potentially important cofactors of race or ethnicity may remain to be recognized. Others that are presumed may not be valid for a particular group in one specific time or place. There is some risk of misinterpreting group differences when potential confounding factors are not considered. For example, if a study finds higher blood pressure in Black American adolescents than in Whites, this could be due to earlier occurrence, on average, of the adolescent skeletal growth spurt in Blacks. Because blood pressure levels are strongly correlated with height in adolescence, Blacks should be expected to exhibit higher age-specific levels of systolic blood pressure at these ages. This underlying variation in growth tempo suggests the explanation that, in Black as in White adolescents, “growth” in blood pressure parallels skeletal growth.11 Data for Specific Population Groups in the United States Data for multiple racial/ethnic minority groups in the United States are increasingly available, although many gaps remain. Data from the National Health Interview Survey of 2004, reported in a special issue of Health Affairs, illustrated the groups often represented in such data and the marked variation—or disparities— among them.24 The proportions of survey participants who reported having had a stroke, other heart disease, coronary heart disease (CHD), and hypertension were demonstrated for groups identified as White, non-Hispanic; Black, non-Hispanic; Hispanic or Latino; Asian; Native Hawaiian or Pacific Islander; and American Indian or Alaska Native. Several observations stood out, including marked dominance of hypertension for all groups and its striking excess among non-Hispanic Blacks; relatively high prevalence of both coronary heart disease and stroke among Native Hawaiians or other Pacific Islanders; and ex-
46894_CH02_019_038.qxd
2/25/10
3:23 PM
Page 29
GEOGRAPHY OR PLACE
ceptionally low frequency of all four conditions among Asians. These patterns are important from both epidemiologic and public health perspectives. Concerning cardiovascular diseases and other aspects of the health of Blacks and other minority groups in the United States, there remains a relative lack of data for these groups. As a result, special reviews of available data, identification of priorities for acquiring needed health data, and adoption of policies for participation in research by certain minorities have taken place. Similar developments address concerns about women and the elderly. Cardiovascular diseases have been a prominent part of this concern. These were addressed, for example, in the 1985 Report of the Secretary’s Task Force on Black and Minority Health of the US Secretary of Health and Human Services,22 and in a subsequent special report from the American Heart Association, “Cardiovascular Diseases and Stroke in African-Americans and Other Racial Minorities in the United States, A Statement for Health Professionals.”25 Corresponding reports on cardiovascular diseases in women were noted earlier. The first of these reports provided summary information on cardiovascular diseases for Black Americans, Hispanics, Native Americans, and Asian/ Pacific Islanders.22 The paucity of information for all but Black Americans was emphasized. Data for other groups were considered less reliable because, for example, of underreporting of race on death certificates. Data on morbidity are even more limited. Heterogeneity within groups is a further consideration. “Hispanic,” for example, includes Mexican Americans, Puerto Ricans, Cuban Hispanics, and Central American immigrants to the United States, or subsequent generations. These considerations emphasize the importance both socially and scientifically of increased representation of diverse groups within the US population in both official health surveillance activities and in research on patterns in cardiovascular disease risks, morbidity, and mortality by race/ethnicity. Whether such expanded information will reveal new insights into causation or point to distinct requirements for intervention remains to be seen. It is often recommended that community- or individual-level interventions be tailored in culturally sensitive ways to meet the needs of specific groups. This is to acknowledge that interventions that are effective in one group may not be so in others. However, it is also important to recognize interventions that are effective across multiple population groups, so that their benefits are not withheld inappropriately.
29
Other Populations Migrants to the United States illustrate a special type of population comparison involving race and ethnicity. Kelleher and others examined the effect of chiefly Northern European immigration to the United States from 1850 to 1930 on coronary heart disease mortality in this country from 1900 to 1980.26 They concluded that this immigration contributed to both the rise and fall of coronary heart disease mortality rates in the United States. They estimated that between the time of immigration and rise of the epidemic, there were lags of about 50 years for men and 38 years for women. They attributed the experience of the immigrant population to deprived socioeconomic conditions on arrival in the United States and to changes in behavior that followed. Another analysis of migrant populations used the National Longitudinal Mortality Study, 1979–1989, to evaluate the mortality experience of native-born compared with foreign-born Americans.27 It was necessary first to adjust for differences between groups in distributions of age, sex, marital status, rural/urban residence, education, and family income. Several groups, including foreign-born minorities, were found to have experienced lower mortality than their USborn counterparts. Foreign-born non-Hispanic Whites, Blacks, Asian and Pacific Islanders, or Hispanics experienced significantly lower cardiovascular disease mortality than those of the same racial/ethnic group who were born in the United States. As an explanation, the authors noted that immigration policy has resulted generally in selection of “a much healthier, more driven, physically-fitter group than those who remain in their countries of origin.”27, p 103 These two studies of migration show the potential error in assumptions about group characteristics. US immigrants of many decades ago are thought to explain part of the mid-20th century peak of coronary mortality, on the basis of poorer general health. Persons able to immigrate to the United States more recently may enjoy relatively good health. Interpreting group identity requires some depth of knowledge beyond the group classification itself.
GEOGRAPHY OR PLACE Cartography of Cardiovascular Diseases The idea of place raises images of maps, such as those shown in Figure 2-1. Such geographic distributions of mortality typically represent officially defined reporting areas—especially cities, counties, states, or
46894_CH02_019_038.qxd
30
2/25/10
3:23 PM
Page 30
CHAPTER 2 DISTRIBUTIONS AND DISPARITIES
the nation as a whole. Mapping of this kind requires sufficient population size in a reporting unit to provide enough events, on an annual basis or aggregated over several years, for reliable estimation of rates. Accordingly, mapping of annual data on cardiovascular mortality by state in the United States or by country on a regional or global basis is convenient and commonplace. More detailed and refined mapping can be done for more local areas, with smaller units requiring multiyear aggregation of events. Techniques of geographic information systems (GIS) are now widely available that permit incorporation of other data of interest for public health, such as social or environmental characteristics or healthcare resources, thereby stimulating a renewed interest in cartography of disease.28 Using current mapping and graphics techniques, Mackay and Mensah produced a world atlas of heart disease and stroke.29 The atlas, with a multicolor layout and multiple text overlays, effectively communicates information about cardiovascular diseases, risk factors, and related social and economic conditions on a regional and global scale. It presents visually what is imagined in studying the tabular data from the Global Burden of Disease Study, discussed in Chapter 1.30 (Unfortunately it is not feasible to reproduce it here.) For nearly 200 countries, the atlas presents tabular data on mortality and disability from heart disease and stroke; rheumatic heart disease deaths; smoking prevalence, policies, and legislation; and prevalence of diabetes among adults. Place can mean something other than geographic location. An organizational or institutional entity whose membership constitutes a population of employees, military personnel, students, or other social groupings can also be seen as a “place” of sorts. Places of these kinds constitute settings in which customs, behaviors, and exposures may contribute to risk or promote good health. Such settings— workplaces, schools, religious institutions, or community centers—can also offer opportunities for interventions such as improving food service, providing health information, screening to detect risk factors or cardiovascular conditions, or providing health services. The United States and Appalachia Racial and Ethnic Disparities in Heart Disease among Women is one of several recently published atlases from a team led by Casper at the Centers for Disease Control and Prevention (CDC). These atlases present GIS mapping of heart disease and stroke mortality for
the United States at a much finer level than the statelevel map in Figure 2-2.31–34 This can be compared with Figure 2-4, in which smoothed county-level heart disease death rates for the period 1991–1995 have been mapped for women dying at ages 35 years and older.31 The publication and corresponding Web site present county-level data for the nation and state by state. The concentration of cardiovascular mortality in the southeastern United States seen in Figure 2-2 is seen here in greater detail. Person (age, sex, and race) and place are taken into account jointly. National maps are shown for all women and for American Indians and Alaska Natives, Asian and Pacific Islanders, Blacks, Hispanics, and Whites; state maps are provided for each group having sufficient data. Corresponding data on local economic resources, indicators of social isolation among elderly women, and medical care resources add dimensions that are potentially valuable in understanding the underlying determinants of the geographic patterns that are revealed. (The data for this series of atlases are publicly available and permit interactive analysis and mapping for selected counties—see http://www.cdc.gov/ dhdsp.) Within the area of especially high coronary mortality in the southeast, 399 counties define the Appalachian region. GIS mapping of cardiovascular mortality in this region was reported in 1998 and updated in 2004.35,36 The whole of West Virginia and parts of 12 other states from Mississippi to New York comprise the region, as shown in Figure 2-5. Ageadjusted coronary heart disease death rates for African American men aged 35–64 are represented for the period 1980–1993. Similar displays are given for four sex-race groups, male/female and African American/ White, and for two age strata, 35–64 years and 65 years and older. Among counties within the region with sufficient data, a wide range of rates was observed. As described in the 2004 report, “Many Appalachian counties with the most adverse health outcomes correlate geographically with socioeconomic characteristics, behavioral risk profiles, and available medical care resources. However, it appears that reasons for disparities in health outcomes are highly variable and localized. Identifying the causes of inconsistencies may help in developing interventions and policy at the local level.”36, p iii “Eight Americas” Another approach has made use of county-level characteristics to “map” disparities in cardiovascular and other causes of death, in accordance with geogra-
46894_CH02_019_038.qxd
2/25/10
3:23 PM
Page 31
GEOGRAPHY OR PLACE
Smoothed County Heart Disease Death Rates 1991–1995
31
All Women Ages 35 Years and Older
New York City
Washington, D.C.
Age-Adjusted Average Annual (Number of Deaths per 100,000 Counties) 212–337 338–379 380–415 416–449 450–670 Insufficient Data
(620) (626) (612) (624) (618) (3)
Figure 2-4 Smoothed County Heart Disease Death Rates, All Women Ages 35 Years and Older, 1991–1995. Source: From Women and Heart Disease: An Atlas of Racial and Ethnic Disparities in Mortality, 2nd Edition. © 2000. 1999 Office for Social Environment and Health Research, West Virginia University.
phy coupled with racial/ethnic and socioeconomic indicators. In this way, Murray and others described “eight Americas,” defined in Table 2-2.37,38 The characteristics used to distinguish these groups included location, population density, race-specific countylevel per capita income, and cumulative homicide rate. The total of 3141 counties in 1980 was condensed into 2072 county-based units aggregated to have stable boundaries through 2001. The populations of the eight Americas ranged in size from 1 million to 214 million; median per capita income varied twofold; percentage completing high school ranged from 84% to 61%; and each group constituted a relatively homogeneous cluster by race, location, and income. In general, the mortality gradient greatly favored “America 1” over “America 8” for cardiovascular
as well as other causes. Each of the “Americas” was compared with several other countries or regions with wide-ranging mortality. Relative to the experience of other countries as disparate in health as Japan and the highest-mortality regions of Africa, “America 1” had 3 years greater life expectancy than the population of Japan. For some other “Americas,” however, mortality was as great as in low- and middle-income countries. Murray and coauthors concluded: “The biggest problem is young and middle-aged male and female mortality from chronic diseases. The challenge to reduce disparities in these groups will require a major re-orientation of US public health over the coming years. While the challenge is great, opportunities exist to tackle the major risks such as tobacco, blood pressure, cholesterol, obesity, and alcohol with innovative strategies.”37, pp 9–10
46894_CH02_019_038.qxd
32
2/25/10
3:23 PM
Page 32
CHAPTER 2 DISTRIBUTIONS AND DISPARITIES
Age-Adjusted Rate per 100,000 Insufficient Data Low Outliers (152–161) First Quartile (152–278) Second Quartile (279–310) Third Quartile (311–356) Fourth Quartile (357–567) High Outliers (473–567)
Figure 2-5 Coronary Heart Disease Deaths: African American Men Aged 35–64, 1980–1993. Source: From Heart Disease in Appalachia: An Atlas of County Economic Conditions, Mortality and Medical Care Resources. © 1998. West Virginia University Prevention Research Center.
PERSON, PLACE, AND TIME Scales of Time The theory of epidemiologic transition (Chapter 1) raised the concept of decades- or century-long processes of change in disease patterns of populations. Cardiovascular epidemiology can encompass such widely ranging temporal aspects as changes in human nutrition on an evolutionary scale; fluctuation in mortality rates over only a few years; and momentary alterations in physiology or behavior that trigger acute cardiovascular events. Intermediate-term variations in mortality for coronary heart disease and stroke demonstrate linkage of time (nearly four decades, from 1950 through 1987) with person (both men and women at specified ages) and place (27 countries).39
Twenty-Seven Countries Secular trends in coronary heart disease mortality over four decades differed strikingly by country (Figure 2-6).39 The figure was compiled from data reported to the World Health Organization by each country throughout this period. This figure represents the average annual mortality for coronary heart disease at ages 45–64 years in each of eight time periods, from 1950–1954 to 1984–1987, for each of 27 countries. The countries are ordered from left to right in descending rank for male coronary heart disease mortality in 1950–1954. The initial rates for males differed widely, about sixfold, among these populations, from about 600/100,000 per year in the United States (USA) to less than 100/100,000 in France (FRA), with the remaining 25 countries arrayed continuously between them. The rates for fe-
Northland low-income rural white
Middle America
Low-income whites in Appalachia and the Mississippi Valley
Western Native American
Black middle America
Southern low-income rural black
High-risk urban black
2
3
4
5
6
7
8
7.5
5.8
23.4
1.0
16.6
$14,800
$10,463
$15,412
$10,029
$16,390
$24,640
$17,758
$21,566
Definition
Urban populations of more than 150,000 blacks living in counties with cumulative probability of homicide death between 15 and 74 y greater than 1.0%
Blacks living in counties in the Mississippi Valley and the Deep South with population density below 100 persons/km2, 1990 county level per capita income below $7,500, and total population size above 1000 persons (to avoid small numbers)
All other black populations living in countries not included in Americas 7 and 8
Native American populations in the mountain and plains areas, predominantly on reservations
Whites in Appalachia and the Mississippi Valley with 1990 county-level per capita income below $11,775
All other whites not included in Americas 2 and 4. Asians not in America 1, and Native Americans not in America 5
Whites in northern plains and Dakotas with 1990 county-level per capita income below $11,775 and population density less than 100 persons/km2
Asians living in counties where Pacific Islanders make up less than 40% of total Asian population
Source: From Eight Americas: Investigating Mortality Disparities Across Races, Counties and Race-Counties in the United States. PLoS Medicine. © 2006, Murray et al.
DOI: 10.1371/journal.pmed.0030260.t001
72%
61%
75%
69%
72%
84%
83%
80%
3:23 PM
214.0
3.6
10.4
2/25/10
Population, income per capita, and education were calculated for race-county combinations from the 2000 US census.
Asian
Definitions and Basic Sociodemographic Characteristics of the Eight Americas Percent Population Average Income Completing General Description (Millions) per Capita High School
1
America
Table 2-2
46894_CH02_019_038.qxd Page 33
PERSON, PLACE, AND TIME 33
46894_CH02_019_038.qxd
34
2/25/10
3:23 PM
Page 34
CHAPTER 2 DISTRIBUTIONS AND DISPARITIES
USA FIN CAN SCO AUL NZE NIR ISR EW IRE CZE AUS SWI SWE FRG DEN HUN BEL ITA
NET NOR YU POL POR SPA JA
FRA
800
Rate per 100,000 Population at Ages 45–64 Years (Age Adjusted)
700
600
500
400
300
200
100
0 Men Women
Figure 2-6 Secular Trend of Mortality from Coronary Heart Disease in 27 Countries, Age 45–64 Years, by Sex, 1950–1987. Source: From TJ Thom, FH Epstein, JJ Feldman, PE Leaverton, and M Wolz, 1992, National Institutes of Health, Pub No 92-3088.
males in every country were typically half or less the rates for males, the only exception being Japan (JA) with its very low rates for both sex groups. The relative gradient in initial rates for females was similar to that for males, about sixfold from highest (Ireland [IRE]) to lowest (France). Over the eight time periods, dramatic changes in coronary heart disease mortality occurred in most of these countries for both males and females. Most striking for males are the marked decreases in countries such as the United States, Canada (CAN), and Australia (AUL) and the increases in Czechoslovakia (CZE), Hungary (HUN), Norway (NOR), and Poland (POL). Intermediate are several countries whose rates for males peaked near the midpoint and later declined toward the initial levels. Also notable is the consistently downward trend for Japan, whose clustering generally with Spain (SPA), Portugal (POR), and France evokes imaginative interpretation. By contrast, however, Japan attained the lowest coronary heart disease mortality rate of any of these countries by the early- to mid-1960s and has since remained in
that rank. Trends for females were in general parallel to those for males, except in those countries where they changed little despite sharp increases in rates for males—Czechoslovakia, Hungary, Norway, and Poland. These data indicate that marked variation in coronary heart disease mortality can occur within populations over short historical intervals. The fact that changes in population genetics require periods of generations, indicates that only environmental factors, including social and behavioral changes, can explain these marked short-term variations. The corresponding picture for stroke is strikingly different from that for coronary heart disease. For example, the United States ranked at midrange instead of first, and Japan had the highest rates rather than the lowest. Stroke mortality in the United States continued to decline through the late 1980s, as shown in the context of such changes in 26 other countries in Figure 2-7, which is analogous to that for secular trends in coronary heart disease above.39 Here, data for deaths at ages 65–74 years are presented because of the typically later ages of stroke deaths in com-
46894_CH02_019_038.qxd
2/25/10
3:23 PM
Page 35
CONCLUSION
JA
35
SCO ITA POR FRG HUN FIN ISR EW NIR AUS SPA FRA AUL SWI USA SWE DEN CZE CAN NZE YU NOR NET IRE BEL POL
1,800
1,600
Rate per 100,000 Population
1,400
1,200
1,000
800
600
400
200
0 Men Women
Figure 2-7 Secular Trend of Mortality from Stroke in 27 Countries, Age 65–74 Years, by Sex, 1950–1987. Source: From TJ Thom, FH Epstein, JJ Feldman, PE Leaverton, and M Wolz, 1992, National Institutes of Health, Pub No 92-3088.
parison with coronary heart disease deaths. The trend for the United States (the 16th in descending rank of stroke mortality in 1950–1954) reflects a decrease through the 1950s and a sharp decline through 1984–1987. In this age group, stroke mortality for the United States declined for men from 600 to about 200 and for women from about 500 to 150 deaths per 100,000 population per year. In the other countries, by far the dominant trend was declining stroke mortality as in the United States, with very close parallels for men and women. Striking for their exceptional patterns of increasing stroke mortality were Czechoslovakia, Hungary, Portugal, and Yugoslavia (YU), which had higher rates at the end of this four-decade period than at the beginning. In Poland, rates were also exceptional in increasing over the latest intervals. A Race Against Time Finally, the projections of future burdens of coronary heart disease, stroke, and other cardiovascular conditions presented in Chapter 1 approach time from yet another perspective. The forecasts represented by the Global Burden of Disease Study40 and A Race Against Time41 forecast a growing disparity in mortality and
disability from cardiovascular and other chronic diseases between the high-income and the low-andmiddle-income regions of the world. Serious consequences were predicted for economic and social development. From this global perspective, Braveman’s concept of disparities, health equity, and social justice arise. Following discussion of the epidemiology of atherosclerotic and hypertensive diseases and their determinants in Parts II and III, these considerations are addressed further in connection with strategies of prevention and planning of public health action.
CONCLUSION The simple concepts of age, sex, and race are fundamental to epidemiologic characterization of individuals and are useful dimensions for examining distributions of disease. Together with place and time, these personal traits provide a basic description of variation in disease occurrence that gives rise to ideas about causation and prevention. Age, sex, and race have further meaning with respect to relative advantage or disadvantage of individuals or groups in terms of social and economic position, education, employment, access to
46894_CH02_019_038.qxd
36
2/25/10
3:23 PM
Page 36
CHAPTER 2 DISTRIBUTIONS AND DISPARITIES
the benefits of society, and potential victimization as by racism or other injustice. It is these considerations that give certain distributions of epidemiologic interest added importance as measures of disparities, or indications of inequities in health that are also of public health concern. Each specific cardiovascular condition and each of their determinants has its own epidemiology, population distributions, and potential disparities. It is useful to bear in mind the rudiments of definition, ascertainment, and interpretation of age, sex, and race. Adding place and time to the picture of cardiovascular diseases underscores the global reach they have assumed in very recent history and may suggest alternative visions for the future. REFERENCES 1. National Heart, Lung and Blood Institute. Morbidity & Mortality: 2007 Chartbook on Cardiovascular, Lung, and Blood Diseases. Bethesda, MD: US Department of Health and Human Services. Public Health Service, National Institutes of Health; June, 2007. 2. Braveman P. Health disparities and health equity: concepts and measurement. Ann Rev Public Health. 2006;27:167–194. 3. Addressing health disparities: the NIH Program of Action. What are health disparities? http:// healthdisparities.nih.gov/whatare.html. Accessed May 15, 2007. 4. US Department of Health and Human Services. Healthy People 2010. 2nd ed. With Understanding and Improving Health and Objectives for Improving Health. 2 vols. Washington, DC: US Government Printing Office; 2000. 5. Kumanyika SK, Morssink CB. Bridging domains in efforts to reduce disparities in health and health care. Health Education & Behavior. 2006;33:440–458. 6. Pooling Project Research Group. Relationship of blood pressure, serum cholesterol, smoking habit, relative weight and ECG abnormalities to incidence of major coronary events: final report of the pooling project. J Chronic Dis. 1978;31: 201–306. 7. Bennet JC, Board of Health Sciences Policy, Organization of the Institute of Medicine. Inclusion of women in clinical trials—policies
for population subgroups. N Engl J Med. 1993; 329:288–292. 8. Ahlbom A, Novell S. Introduction to Modern Epidemiology. Chestnut Hill, MA: Epidemiology Resources, Inc; 1990. 9. Barker DJP. The developmental origins of cardiovascular disease. In: Marmot M, Elliott P, eds. Coronary Heart Disease Epidemiology: From Aetiology to Public Health. 2nd ed. Oxford (England): Oxford University Press; 2005:547–567. 10. World Health Organization Study Group. Prevention in Childhood and Youth of Adult Cardiovascular Diseases: Time for Action. WHO Technical Report Series 792. Geneva (Switzerland): World Health Organization; 1990. 11. Tanner JM. Fetus into Man: Physical Growth from Conception to Maturity. 2nd ed. Ware (England): Castlemead Publication; 1989. 12. Williams CL, Hayman LL, Daniels SR, et al. Cardiovascular health in childhood. A statement for health professionals from the Committee on Atherosclerosis, Hypertension, and Obesity in the Young (AHOY) of the Council on Cardiovascular Disease in the Young, American Heart Association. Circulation 2002;106:143–160. 13. Harlan WR, Manolio TA. Coronary heart disease in the elderly. In: Marmot M, Elliott P, eds. Coronary heart disease epidemiology: from aetiology to public health. Oxford (England): Oxford University Press; 1992:114–126. 14. World Health Organization Study Group. Epidemiology and Prevention of Cardiovascular Diseases in Elderly People. WHO Technical Report Series 853. Geneva (Switzerland): World Health Organization; 1995. 15. Wenger NK, Speroff L, Packard B. Cardiovascular health and disease in women. N Eng J Med. 1993;329:247–256. 16 . Douglas PS. Cardiovascular health and disease in women. Philadelphia, PA: WB Saunders Co; 1993.
46894_CH02_019_038.qxd
2/25/10
3:23 PM
Page 37
REFERENCES
17. Mosca L, Appel, LJ, Benjamin EJ, et al. American Heart Association. Evidence-based guidelines for cardiovascular disease prevention in women. Circulation. 2004;109: 672–692. 18. Mosca L, Banka CL, Benjamin EJ, et al. Evidence-based guidelines for cardiovascular disease prevention in women: 2007 update. Circulation. 2007;115:1481–1501. 19. Barrett-Connor E. Sex differences in coronary heart disease: why are women so superior? The 1995 Ancel Keys Lecture. Circulation. 1997;95:252–264. 20. Khaw K-T, Barrett-Connor E. Sex differences, hormones, and coronary heart disease. In: Marmot M, Elliott P, eds. Coronary heart disease epidemiology: from aetiology to public health. Oxford (England): Oxford University Press; 1992:274–286. 21. Strong JP, Oalmann MC, Malcolm GT. Pathobiological Determinants of Atherosclerosis in Youth (PDAY) Research Group. Atherosclerosis in youth: relationships of risk factors to arterial lesions. In: Filer LJ Jr, Lauer RM, Luepker RV, eds. Prevention of Atherosclerosis and Hypertension Beginning in Youth. Philadelphia, PA: Lea & Febiger; 1994: 13–20. 22. US Department of Health and Human Services. Report of the Secretary’s Task Force on Black and Minority Health, 1: Executive Summary. Bethesda, MD: US Department of Health and Human Services; 1985. 23. Goff DC Jr, Varas C, Ramsey DJ, Wear ML, Labarthe DR, Nichaman MZ. Mortality after hospitalization for myocardial infarction among Mexican-Americans and non-Hispanic Whites: the Corpus Christi Heart Project. Ethnicity Dis. 1993; 3:55–63. 24. Mensah GA, Brown DW. An overview of cardiovascular disease burden in the United States. Health Affairs. 2007;26:38–48. 25. Cardiovascular diseases and stroke in AfricanAmericans and other racial minorities in the United States: a statement for health professionals. Circulation. 1991;83:1463–1480.
37
26. Kelleher CC, Lynch JW, Daly L, et al. The “Americanisation” of migrants: evidence for the contribution of ethnicity, social deprivation, lifestyle and life-course processes to the mid-20th century coronary heart disease epidemic in the US. Social Science & Medicine 2006;63:465–484. 27. Singh G, Siahpush M. Ethnic-immigrant differentials in health behaviors, morbidity, and cause-specific mortality in the United States: an analysis of two national databases. Human Biology. 2002;74:83–109. 28. Koch T. Cartographies of Disease: Maps, Mapping, and Medicine. Redlands, CA: ESRI Press; 2005. 29. Mackay J, Mensah GA. The Atlas of Heart Disease and Stroke. Geneva (Switzerland): World Health Organization; 2004. 30. Mazzati E, Vander Hoorn S, Lopez AD, et al. Comparative quantification of mortality and burden of disease attributable to selected risk factors. In: Lopez AD, Mathers CD, Ezzati M, Jamison DT, Murray CJL, eds. Global Burden of Disease and Risk Factors. Washington, DC: The International Bank for Reconstruction and Development/The World Bank; 2006:241–396. 31. Casper ML, Barnett E, Halverson JA, et al. Women and Heart Disease: An Atlas of Racial and Ethnic Disparities in Mortality. 2nd ed. Morgantown, WV: Office for Social Environment and Health Research, West Virginia University; 2000. 32. Barnett E, Casper ML, Halverson JA, et al. Men and Heart Disease: An Atlas of Racial and Ethnic Disparities in Mortality. 1st ed. Morgantown, WV: Office for Social Environment and Health Research, West Virginia University; 2001. 33. Casper ML, Barnett E, Williams GI Jr, Halverson JA, Braham VE, Greenlund KJ. Atlas of Stroke Mortality: Racial, Ethnic, and Geographic Disparities in the United States. Atlanta, GA: Department of Health and Human Services, Centers for Disease Control and Prevention; 2003.
46894_CH02_019_038.qxd
38
2/25/10
3:23 PM
Page 38
CHAPTER 2 DISTRIBUTIONS AND DISPARITIES
34. Casper ML, Denny CH, Coolidge JN, et al. Atlas of Heart Disease and Stroke Among American Indians and Alaska Natives. Atlanta, GA: Department of Health and Human Services, Centers for Disease Control and Prevention and Indian Health Service; 2005. 35. Barnett E, Elmes GA, Braham VE, Halverson JA, Lee JY, Loftus S. Heart Disease in Appalachia: An Atlas of County Economic Conditions, Mortality, and Medical Care Resources. Morgantown, WV: Prevention Research Center, West Virginia University; 1998. 36. Halverson JA, Ma L, Harner EJ. An Analysis of Disparities in Health Status and Access to Health Care in the Appalachian Region. Executive Summary. Prepared for the Appalachian Regional Commission, Washington, DC: West Virginia University; 2004. 37. Murray CJL, Kulkarni S, Ezzati M. Eight Americas: new perspectives on U.S. health disparities. Am J Prev Med. 2005;29:4–10. 38. Murray CJL, Kulkarni SC, Michaud C, et al. Eight Americas: investigating mortality disparities across races, counties, and race-counties in the United States. PLoS Med. 2006;3: 1513–1524.
39. Thom TJ, Epstein FH, Feldman JJ, Leaverton PE, Wolz M. Total Mortality and Mortality from Heart Disease, Cancer and Stroke from 1950 to 1987 in 27 Countries: Highlights of Trends and Their Interrelationships Among Causes of Death. NIH publication 92-3088. Bethesda, MD: National Heart, Lung and Blood Institute, National Institutes of Health; 1992. 40. Murray CJL, Lopez AD. Alternative visions of the future: projecting mortality and disability, 1990–2020. In: Murray CJL, Lopez AD, eds. The Global Burden of Disease: A Comprehensive Assessment of Mortality and Disability from Diseases, Injuries, and Risk Factors in 1990 and Projected to 2020. Boston, MA: The Harvard School of Public Health; 1996: 325–395. 41. Leeder S, Raymond S, Greenberg H. A Race Against Time: The Challenge of Cardiovascular Disease in Developing Countries. New York: The Trustees of Columbia University in the City of New York; 2004.
46894_CH03_039_058.qxd
2/25/10
3:24 PM
Page 39
The Major Cardiovascular Diseases
39
P A R T
2
46894_CH03_039_058.qxd
2/25/10
3:24 PM
Page 40
46894_CH03_039_058.qxd
2/25/10
3:24 PM
Page 41
C H A P T E R
3 Atherosclerosis with behavioral interventions, drugs, or surgery to lower cholesterol concentrations has been shown by serial angiography to delay progression and even produce regression of atherosclerotic lesions. How is the epidemiology of atherosclerosis changing on a national and global scale? Continuing development of noninvasive examination methods and wider application of existing standardized protocols for pathological examination of coronary arteries of young decedents could contribute to answering this question.
SUMMARY Atherosclerosis is a pathological condition that occurs in medium and large arteries throughout the body. Its clinical manifestations appear especially in the heart, brain, lower extremities, and aorta. It is the underlying condition in the occurrence of myocardial infarction (heart attack), ischemic cerebrovascular accident (occlusive stroke), peripheral arterial disease of the lower extremities, and aortic aneurysm. Postmortem study of tissues obtained at autopsy has been a major component of atherosclerosis research since long before the 20th century. Recent technology has permitted study of patients and populations through invasive and noninvasive methods, respectively—for example, by angiographic examination of patients and by ultrasonography to measure carotid artery lesions in large population studies. Through autopsy studies, atherosclerosis has been found in populations throughout the world. Early-stage lesions occur ubiquitously, beginning in childhood. Autopsies of US military casualties dying from combat-related or other injuries in Korea and Vietnam indicated unexpectedly severe coronary atherosclerosis in some young soldiers with no known symptoms of coronary heart disease. Recent evidence indicates strong correlations between the extent and severity of atherosclerosis in adolescents and young adults and such characteristics as adverse blood lipid profile, high blood pressure, and cigarette smoking. Singly or in combination, these factors, measured in childhood and adolescence, predict the extent and severity of coronary atherosclerosis not only in early adulthood but at least to the mid-50s. Public health recommendations promote desirable behavior patterns beginning in childhood to prevent development and progression of atherosclerosis. Treatment of adults
INTRODUCTION “Atherosclerosis” refers to the consistency of material typically found on the inner, or luminal, surface and in the wall of large- to medium-diameter arteries throughout the body. The areas of the circulation, or vascular beds, where this development is most critical are those supplying the heart and brain. Atherosclerosis is a descriptive term for the pathological mushy areas (atheromata), often encrusted or hardened (sclerosed) by deposition of calcium, which weaken the arterial wall and intrude into the lumen or channel of the vessel, restricting or obstructing blood flow. The atherosclerotic plaque, whether fully matured or in intermediate stages of development, is now regarded as the key to precipitating blood clot formation (thrombosis) with sudden interruption of blood flow. A variety of outcomes may follow, depending on the location, severity, and duration of the interruption. Most important among these are heart attack, stroke, and related conditions. Figure 3-1 presents a microscopic cross section of a coronary artery from a fatal case of myocardial infarction.1 The clear area at the upper right within the
41
46894_CH03_039_058.qxd
42
2/25/10
3:24 PM
Page 42
CHAPTER 3 ATHEROSCLEROSIS
Figure 3-1 Light Micrograph of a Histological Section of a Complicated Atherosclerotic Plaque from Human Coronary Artery with Attached Mural Thrombus. Source: Reproduced with Permission from RW St. Clair, Biology of Atherosclerosis, in TA Pearson et al., Primer in Preventive Cardiology, © 1994, American Heart Association.
artery (from 12 to 3 o’clock) is all that remains of its lumen or channel of blood flow. This area has been greatly reduced by the dark-staining thrombus that shares this quadrant of the artery and rests on top of a large accumulation of material that occupies three quarters of the vessel area, from 3 to 12 o’clock. This latter region of the wall contains not only dead tissue but also crystals of cholesterol, often described as “shale-like” in microscopic appearance, which are an essential component of the classical atherosclerotic plaque. Details of classification of atherosclerotic lesions as observed at the histologic, or tissue, level by light microscopy are found in two successive reports from the Committee on Vascular Lesions of the Council on Arteriosclerosis of the American Heart Association.2,3 The first category of abnormality, not yet an atherosclerotic condition, is adaptive thickening of the coronary artery. Grades 1 and 2 apply to low-grade lesions, grossly visible yellow fatty streaks in the innermost, or intimal, layer of the artery wall. Grade 3 lesions are intermediate in their development. Grades 4 and 5 apply to advanced lesions, potentially most significant for development of clinical manifestations. These latter lesions include the “vulnerable plaques,” which are most susceptible to rupture and
initiation of a process leading to thrombosis, with obstruction of blood flow as seen in Figure 3-1. Chronic inflammation, which tends to soften the material within the fibrous cap of the plaque, predisposes to rupture or cracking (fissuring) of the plaque covering.4 The resulting thrombosis may heal, adding to growth of the plaque, or it may expand to obstruct blood flow. Whether the inflammatory process within the plaque depends on systemic infection or another chronic inflammatory condition is unresolved. It has recently been proposed that the vulnerable plaque should include three types of conditions: plaque rupture, as described above, with a preceding stage called a “thin-cap fibroatheroma”; plaque erosion, not associated with inflammation; and calcified nodules.5 These differ in their patterns of occurrence in specific clinical situations. The concept of progression from the lowest- to highest-grade lesions is basic to understanding development of atherosclerosis beginning in childhood and adolescence.
METHODS OF MEASUREMENT— INVASIVE AND NONINVASIVE The foregoing perspective on atherosclerosis is the pathologist’s view, based mainly on examination of
46894_CH03_039_058.qxd
2/25/10
3:24 PM
Page 43
MANIFESTATIONS—ABOVE AND BELOW THE “CLINICAL HORIZON”
material obtained at autopsy. Study of atherosclerosis in living persons utilizes measurement methods broadly categorized as “invasive” and “noninvasive.” Location and extent of atherosclerotic lesions can be investigated clinically, in patients suspected of having atherosclerosis, by injection of radiopaque dye into specific regions of the arterial circulation, as in coronary angiography. This technique permits repeated measurement in individual patients to detect change in size of atherosclerotic lesions—such as reduced rates of progression or actual regression of lesions in response to treatment.2,6 By reason of risk, feasibility, and cost, less direct, or “noninvasive,” methods of examination are needed for populationbased studies of atherosclerosis in living persons. Two methods for rendering images of atherosclerosis are ultrasound examination of the carotid artery, which supplies blood flow in the neck to connect the thoracic aorta with the cerebral circulation, and computed tomography of the coronary arteries, which supply the heart muscle or myocardium. Ultrasound reflectance by B-mode ultrasonography has been validated for estimating carotid arterial wall thickness as an indicator of underlying atherosclerosis. The carotid artery is readily accessible, is relatively immobile, is relevant to circulation to the brain and implicated in risk of stroke, and correlates well with pathology in the coronary arteries. The measure obtained is the intimal-medial thickness (IMT), sometimes referred to as the carotid IMT (CIMT). This technique uses an ultrasound transducer and sensor applied over the carotid artery and measures variations in wall thickness or lumen diameter to the scale of tenths of a millimeter. By this means, epidemiologic assessment of atherosclerosis is no longer limited to postmortem samples or highly selected cases of known or clinically suspected arterial disease but can be extended to the general, living population. This technique is being used, for example, in a population study of some 15,800 healthy American adults in the Atherosclerosis Risk in Communities (ARIC) Study. A recent review provides details of the technique, representative images in normal and diseased arteries, findings of the ARIC Study linking CIMT with other characteristics of participants, and summary results from a number of clinical trials showing reduction in CIMT after blood lipid-lowering therapy.7 A second noninvasive method now being investigated in a number of studies, most prominently the Multi-Ethnic Study of Atherosclerosis (MESA) Study, is ultra-fast computed tomography (UFCT), or CT scanning.8 As applied to examination of the heart, UFCT detects and quantifies calcified atherosclerotic
43
lesions by measuring the intensity of the coronary artery image due to calcium deposits in plaques. “Ultrafast” refers to rapid imaging that permits examining coronary arteries even though they are in constant motion throughout the cardiac cycle. The resulting measure is the coronary artery calcium (CAC) score. The primary purpose of the MESA Study, a major multicenter prospective epidemiologic study, is to determine the value of the CAC score in estimating the risk of future coronary events and its potential role in clinical risk assessment. Parallel development of CT angiography offers another mode of examination for assessing the diameter of the arterial lumen. However, it is important to recognize that these techniques involve exposure to ionizing radiation, a concern when mass screening for coronary artery calcium, for example, is contemplated. Discussion of the CAC as a screening tool has stimulated development of a scientific consensus statement reviewing available evidence and providing recommendations on its role.9 Current judgment, as reflected in this review, is that neither high- nor lowrisk individuals stand to benefit from this examination, but persons at intermediate risk might be reclassified to a higher-risk category on the basis of the CAC score. Among several reservations expressed about utility of the score, the report notes that presently available data are largely limited to nonHispanic Caucasian men and cautions against extrapolation to women and ethnic minorities. (The issues of sex and race/ethnicity raised in Chapter 2 have clearly not yet been resolved in this area but are being addressed in the MESA Study.) Other approaches to assessing atherosclerosis without clinical angiography or autopsy include several clinical or subclinical examinations that can provide indirect evidence of disease. Magnetic resonance imaging (MRI) may become applicable for population use in the future, to assess plaque characteristics and as a form of angiography, avoiding radiation exposure.
MANIFESTATIONS—ABOVE AND BELOW THE “CLINICAL HORIZON” Atherosclerosis is a systemic disorder, in that it occurs in large- and medium-diameter vessels throughout the arterial tree. For this reason its manifestations above the “clinical horizon,” as termed by McGill and others and illustrated in Figure 3-2, may relate to any of several different regions of the circulation, for example, the heart, brain, lower extremities, or aorta.10 This schematic view of the development of atherosclerosis from below to above the clinical horizon has
46894_CH03_039_058.qxd
44
2/25/10
3:24 PM
Page 44
CHAPTER 3 ATHEROSCLEROSIS
Figure 3-2 Schematic View of the Development of Atherosclerosis. Source: Reprinted with Permission from HC McGill Jr., JC Geer, and JP Strong. The Natural History of Human Atherosclerotic Lesions, in Atherosclerosis and Its Origin, M Sandler and G Bourne, eds, p 42, © 1963, Academic Press.
been published in several versions since the 1950s. In the intervening years, atherosclerotic lesions at several stages of development below the clinical horizon have become detectable, for example, by CIMT and UFCT. There are two especially important implications of this schematic representation. First is that the earliestappearing lesions, fatty streaks, are truly precursors of fibrous plaques, and they in turn lead to complicated lesions. Today, refinements in categorization of lesion types and the concept of the vulnerable plaque would be added to the scheme. Second is that the atherosclerosis begins in the first decades of life even though its clinical manifestations most often occur in the fifth decade and beyond. Today, the occurrence of clinical manifestations in the 30s would be noted. With respect to clinical manifestations, no single cardiovascular condition, such as coronary heart disease alone, reflects the full contribution of atherosclerosis to disease, disability, and death in a population. Accordingly, among the conditions identified in Table 1-1 by ICD 10 codes, all those from I10 through I79 and some in the category I95 through I99
are potential consequences of atherosclerosis. The epidemiology of atherosclerosis has been approached mainly in studies of coronary heart disease and stroke. A more complete assessment includes peripheral arterial disease and aortic atherosclerosis. The epidemiology of these latter conditions is based on their occurrence either as causes of death or as conditions detectable in life through clinical data sources or special ascertainment in epidemiologic studies. These conditions are discussed separately in Chapters 4–6. The extent to which atherosclerotic conditions occur below the clinical horizon, and with what manifestations, is a question addressed, for example, in the Cardiovascular Health Study (CHS) involving adults aged 65 years and older.11 CHS investigated a composite index of subclinical atherosclerosis through carotid ultrasound, echocardiography, electrocardiography, the ankle-brachial index (ABI, a measure of blood pressure differences between arm and ankle reflecting obstruction in lower extremity arterial supply), and responses to questionnaires about symptoms of impaired coronary or lower extremity arterial blood flow, also known as ischemia. More than onethird of participants had subclinical (i.e., clinically inapparent) signs or symptoms of atherosclerosis. In the same study, only one-quarter to one-third of men and women had clinical disease. Detection of subclinical disease among the rest of the population showed atherosclerosis to be more than twice as prevalent as indicated by clinical evidence alone. In a further report on the findings of CHS, presence of several indicators of subclinical disease was described for Black and White men (Figure 3-3a) and women (Figure 3-3b).12 The two pictures appear generally consistent. They show somewhat higher prevalence of these conditions for men than for women, more frequent occurrence among Blacks than Whites for most conditions for both men and women, and— taking all conditions together—more than two-thirds of men and more than one-half of women being affected by one or more conditions. Additional data in this report indicate the importance of these observations in relation to major clinical events. Among those with subclinical disease, compared with those without, total mortality was more than twice as frequent (5.5% versus 1.9%); total coronary heart disease was nearly twice as frequent (8.2% versus 4.3%); and incident myocardial infarction was 1.2 times as frequent (3.1% versus 2.6%). CHS demonstrates the high frequency of subclinical disease that can be identified when sought by special-purpose examination among Black and White men and women in this age group.
46894_CH03_039_058.qxd
2/25/10
3:24 PM
Page 45
MANIFESTATIONS—ABOVE AND BELOW THE “CLINICAL HORIZON”
70
66.8
69.5
Prevalence (%)
60 49.4
50 40
16.4
24.7
22.1
21.9 20
33.8
31.9
30
30 17.8
8.1
10
1.3 0.5
0
s ll e ll e % tie air air wa wa 25 ali nn tive nn tive tid 0% rotid 80% sis > o m o i i r o t i t i r 8 s s s s a a no ue po ue po l c ss > on c ess > steno ab e q gina e q tion G n rna kne m s s d k i e C o t t o a c m E R R dic an In thic Co thi ro jor u Ca cla Ma Condition m
Hg
m 0.9
/<
AA
1.8 2.3 tal
To
Figure 3-3a Prevalence of Subclinical Disease Among Black (Shaded Bar) and White (Open Bar) Men with at Least One Subclinical Disease at Their Baseline Examination in the Cardiovascular Health Study. Source: Reprinted with permission from Cardiology Clinics, Vol 17, Kuller L, Sutton-Tyrrell K, Aging and cardiovascular disease. Use of subclinical measurements, pp 51–65, © Elsevier 1999.
70 61.4
Prevalence (%)
60
55.9
50 41.4 40 31.1
27.9
30 19.3
20 10
14.2 8.5
12.4 14.2
16.2
11.1 1
1.1
3
2.4
0 s all ire all ire 5% tie na ve na ve ali d w 0% otid w 0% is >2 n n i t o m i o i i 8 s 8 or . st sit sti sit ar aro bn <0 ue a po ue n po l c ess > on c ess > steno a / q q a G in m kn AA se se tio id ern kn EC Ro ang Ro dica Int thic Com thic arot jor u C a a l c M Condition m 9m
Hg
tal
To
Figure 3-3b Prevalence of Subclinical Disease Among Black (Shaded Bar) and White (Open Bar) Men with at Least One Subclinical Disease at Their Baseline Examination in the Cardiovascular Health Study. Source: Reprinted with permission from Cardiology Clinics, Vol 17, Kuller L, Sutton-Tyrrell K, Aging and cardiovascular disease. Use of subclinical measurements, pp 51–65, © Elsevier 1999.
45
46894_CH03_039_058.qxd
46
2/25/10
3:24 PM
Page 46
CHAPTER 3 ATHEROSCLEROSIS
MECHANISMS OF ATHEROGENESIS The question of how atherosclerosis develops has been investigated at least since the mid-19th century. An overview by Stamler is the basis for the following highlights of the early research.13 The work by Virchow in the latter half of the 19th century established atherosclerosis as a distinct pathological entity in which the “mush” was identified as fatty (lipid) material, specifically including cholesterol. Clinical research that followed showed that diseases characterized by prolonged high blood cholesterol concentrations produced severe premature atherosclerosis, establishing a link between atherosclerosis and blood lipid concentrations. Early in the 20th century, the accidental discovery was made that experimental manipulation of rabbits’ diets by feeding them animal products produced atherosclerosis. Subsequent animal experiments indicated that even small cholesterol supplements as part of long-term dietary changes would also produce this condition. Further animal studies suggested cellular processes in which migrating cells especially high in cholesterol content invaded the arterial wall. The laboratory research data on diet were reinforced in the mid-1930s on an altogether different scale by reports on “geographical pathology.” These were studies that linked differences among populations in frequency of occurrence of atherosclerosis with differences in their typical diets. In the period immediately following the Great Depression and Second World War, laboratory and clinical research on atherosclerosis expanded rapidly and included biochemistry, biophysics, pathology, and cell biology. Mechanisms of lipid metabolism and transport were investigated intensively. By the mid-1970s, studies in vascular biology led to refined concepts of atherogenesis. Three complementary mechanisms were outlined by Ross in the “response-to-injury” hypothesis, attributed in its original formulation to Virchow.14 An early step in atherogenesis appeared to be adhesion of blood monocytes to the arterial endothelium, followed by migration of these cells into the intimal layer, and subsequent concentration of lipid in these cells. This process was facilitated by the oxidation of lipids carried by lowdensity lipoproteins. In a second phase of development, fatty streaks beneath the endothelium may be altered by migration of smooth muscle cells from the medial into the intimal layer of the arterial wall. Here, factors that stimulate cellular proliferation may produce a marked multiplication of these cells to produce raised lesions. Then, in later progression of these lesions, a long-hypothesized step of endothelial cell
dysfunction and damage may occur from elevated blood lipid concentrations, viral infection, or other plausible causes. Platelet adherence to damaged endothelial cells then leads to local thrombosis, with occlusion of the artery and the resulting signs and symptoms of impaired or obstructed blood flow to the tissues beyond.1 Further laboratory studies continued to elaborate on this concept of atherogenesis, some aspects of which still require confirmation, including the relation of this process to clinical manifestations of atherosclerosis. By the late 1990s, Ross declared that the fundamental process underlying the response-to-injury hypothesis was inflammation, a series of specific cellular and molecular changes in the arterial wall:15, p 115 The most recent version of this hypothesis emphasizes endothelial dysfunction . . . each characteristic lesion of atherosclerosis represents a different stage in a chronic inflammatory process in the artery; if unabated and excessive, this process will result in an advanced, complicated lesion. Possible causes of endothelial dysfunction leading to atherosclerosis include elevated and modified LDL [low-density lipoprotein cholesterol— see Chapter 11]; free radicals caused by cigarette smoking, hypertension, and diabetes mellitus; genetic alterations; elevated plasma homocysteine concentrations; infectious microorganisms such as herpesviruses or Chlamydia pneumoniae; and combinations of these or other factors. A separate and more limited line of investigation lends some support to the concept that atherosclerosis originates with viral infection.16 Members of the herpesvirus group (herpes simplex virus, Epstein-Barr virus, cytomegalovirus, and herpes zoster virus) have been of particular interest on the basis of both animal and human studies. It has been suggested that coronary reocclusion after angioplastic surgery to relieve stenosis may be associated with viral infection.17 Widespread occurrence of these viruses and high incidence of infection early in life add to epidemiologic interest in this concept, which also deserves further investigation. How this might be linked with evidence for other processes involved in atherosclerosis remains to be established. Bacterial infection has also been invoked as a factor in the progression of atherosclerosis, with growing interest in apparent associations between infection with Chlamydia pneumoniae or Helicobacter pylori, for example, and presence of coronary heart disease.17,18 Weighing against this hypothesis is failure of many trials of antibiotic therapy
46894_CH03_039_058.qxd
2/25/10
3:24 PM
Page 47
PERSON, PLACE, AND TIME
and some prospective epidemiologic studies to provide supporting evidence.
PERSON, PLACE, AND TIME Published accounts indicate the presence of atherosclerosis even in antiquity. Well-preserved human remains from ancient China and Egypt have revealed atherosclerosis in the coronary arteries.19,20 As to how widespread its occurrence may have been in early times, it can be presumed that those whose remains were so carefully prepared for burial were persons of exceptional affluence and not typical of their societies. In more recent times, occurrence of atherosclerosis has been well recognized since the mid-19th century. A detailed account of work in the pathology of atherosclerosis up to the mid-1950s was presented by Holman and colleagues in 1958, the source of the original version of Figure 3-2.21 The concept advanced at that time was based on a large series of autopsies in persons 1–40 years of age, at Charity Hospital of Louisiana in New Orleans, as well as previous reports from as early as 1911. The essential observation was that the earliest lesions of atherosclerosis, the fatty streaks, appear commonly in the first decade of life. These can progress with age to fibrous plaques and more advanced lesions. Rates of progression with age differ among population groups, being more gradual in Blacks than in Whites in New Orleans, for example, according to Holman’s review. Holman suggested that factors initiating fatty streaks differ from the factors leading to their gradual conversion, over 15 years or more, to advanced lesions. In addition, a rapid increase during puberty in the percentage of aortic surface involved with fatty streaks was taken by Holman as evidence that hormonal changes, rather than diet, were the major determinant of progressive atherogenesis. In 1953, a report of postmortem examinations among US military casualties in Korea revealed atherosclerosis of the coronary arteries in the majority of young men examined, typically in their early 20s.22 Lesions were visible at autopsy in the coronary arteries of 77.3% of the 300 cases studied; complete occlusion of one or more coronary arteries was found in 3.0% of cases. Presumably, these occlusive lesions had evolved very gradually, and therefore from much earlier ages, with concurrent development of collateral circulation to protect against symptoms (usually acute chest pain) of myocardial ischemia. An analogous investigation was conducted 18 years later among US military casualties in Vietnam,
47
on the basis of postmortem coronary angiography to detect atherosclerosis.23 The resulting estimate of the frequency of “some degree” of coronary atherosclerosis was 45% among the 105 examinations performed and of “apparently severe atherosclerosis” was 5%. This was thought to represent a decrease in prevalence relative to the observations in Korea two decades earlier, but subsequent reinvestigation of materials from the Korean War casualties indicated that the two results were consistent, at least in the frequency of severe lesions.24 These studies reinforce the view presented in Figure 3-2 that atherosclerosis has already progressed before age 20 in a substantial proportion of US population groups, as found in young military personnel and both Black and White residents of New Orleans. Further studies of atherosclerosis in populations living under widely varying conditions were undertaken more than four decades ago, after development of carefully standardized techniques for postmortem collection and examination of specimens of the aorta and coronary arteries. A landmark investigation in the epidemiology of atherosclerosis was the International Atherosclerosis Project (IAP) in which 23,207 sets of coronary arteries and aortas were collected in 14 countries, from 1960 to 1965.25 The materials were from autopsy examinations of males and females from 10 to 69 years of age at death from conditions unrelated to atherosclerosis. Dissected and prepared under a common protocol locally, the materials were examined grossly and microscopically by teams of pathologists in a central laboratory. One measure of the extent of atherosclerosis was the percentage of the intimal surface of the artery covered by raised atherosclerotic lesions, that is, by fibrous plaques and complicated or calcified lesions. This measure of the extent of atherosclerosis in the aorta (abdominal and thoracic aorta combined) and coronary arteries (three coronary arteries combined) was presented for each of 19 race-location groups in the IAP. Subjects in these comparisons died at 25 to 64 years of age. For New Orleans Whites, for example, the percentage surface involvement of the coronary arteries was greatest for males (more than 20%), more than twice the percentage for females. For Bantus in Durban, South Africa, by contrast, the corresponding percentages of surface involvement for both males and females were much less, only about 8%. In general, the IAP showed that atherosclerosis in the aorta was not consistently more extensive among males than among females; for the coronary arteries, the extent was almost without exception greater among males. Large population differences can thus occur in the extent of atherosclerosis. It
46894_CH03_039_058.qxd
48
2/25/10
3:24 PM
Page 48
CHAPTER 3 ATHEROSCLEROSIS
differs in different regions of the circulation, such as aorta versus coronary arteries, even between males and females in the same population. In addition, when nine populations of men aged 45–54 and 55–64 years were ranked by extent of raised lesions in the coronary arteries, there was close correspondence with their ranks by mortality rates for coronary heart disease. Marked progression of atherosclerotic plaques by age is indicated by selected results from the IAP and several populations studied under a protocol of the World Health Organization (WHO).26 A composite
curve for Malmo, Sweden, and four Eastern European cities studied under the WHO protocol shows increases in the proportion of persons with atherosclerotic plaques from less than 15% at age 20 to 100% at ages 50 and older (vessel and lesion type not specified) (Figure 3-4). From selected sites in the IAP, percentages of males and females combined with fibrous plaques in the left anterior descending coronary artery are indicated by age. Wide variation in the extent of such lesions even as early as age 20 is remarkable, as is the steep increase in all populations to peaks from
100 Oslo
WHO study in 5 towns: Malmo, Prague, Ryazan, Yalta, and Tallin
90
New Orleans
Jamaica
80
70
Puerto Rico (White population)
60
Lima
50 Mexico 40
X
30
X 20
X X X
10
X 0 0
10
20
30
40
50
60
70
80
Years
Figure 3-4 The Extent of Atherosclerosis by Age, from the WHO Project and the International Atherosclerosis Project. Derived from autopsy studies; prevalence rates at age 20 (X). Source: Reprinted with permission from TRS 678, p 35, World Health Organization.
46894_CH03_039_058.qxd
2/25/10
3:24 PM
Page 49
ATHEROSCLEROSIS IN CHILDHOOD, YOUTH, AND EARLY ADULTHOOD
50% to nearly 100% of persons being affected. That 25% of the New Orleans population is affected by age 20 indicates that atherosclerosis has commonly begun well before the earliest adult years. By inference, whatever factors determine the rate of progression of atherosclerosis must operate well before age 20 in some populations.
ATHEROSCLEROSIS IN CHILDHOOD, YOUTH, AND EARLY ADULTHOOD Methods and Findings The preceding work provided a strong impetus to further investigation of the onset and progression of atherosclerosis from childhood through early adulthood. In particular, interest developed in determining whether factors predictive of the later clinical manifestations of atherosclerosis are related to its early progression. A challenge for epidemiologic study of this question was to obtain information for the same individuals about both the presence and extent of atherosclerosis, on the basis of postmortem examination, and the presence of factors before death that might be predictive of the nature and extent of the disease. Two distinct approaches have been used. First, a study in childhood of factors related to atherosclerotic and hypertensive diseases was conducted by Berenson and others in Bogalusa, Louisiana, beginning in the 1970s. Ultimately, several thousand school-age children were examined. In 1986, the investigators first reported that 35 study participants had died and undergone autopsy examinations including standardized examination of the heart, coronary arteries, and aorta.27 Updates in 1992 and 1998 included totals of 62 and 93 decedents, respectively, from ages 3 to 38 years.28,29 A second epidemiologic approach was to begin by identifying large numbers of individuals dying young from noncardiovascular causes. At the time of death it would be possible to obtain information about relevant characteristics and conduct highly standardized and detailed postmortem examination of coronary arteries and the aorta. This was the approach of a large, 15-center study in the United States, the Pathobiological Determinants of Atherosclerosis (PDAY) Study. A publication from 2002 reported on more than 3000 subjects dying at ages 15–34 years between 1987 and 1994.30 Methods and many of the findings are presented in that report. In the PDAY Study, the extent of atherosclerosis was measured by the percentage of surface area of the intimal lining of the vessel involved with fatty streaks
49
or raised lesions. Such data, separately by sex and 5-year age groups, are represented in Figure 3-5, taken from a 1997 report.31 The upper panels show the extent of surface area involved with fatty streaks in the abdominal aorta (A) and right coronary artery (C), for men (open bars) and women (solid bars). The lower panels (B and D) refer to raised lesions in the corresponding vessels. Relative to males, females had less extensive raised lesions in the coronary artery, at every age. They also had more extensive fatty streaks but a similar extent of raised lesions in the aorta and a similar extent of fatty streaks in the coronary artery. Sex differences in development of coronary atherosclerosis, favoring females, were apparent among these teens and young adults. Precursors and Predictors Beyond these observations on atherosclerosis itself between the mid-teens and mid-thirties, both studies addressed the possible association between factors thought to be relevant to development of atherosclerosis and the extent of pathology in the coronary artery, aorta, or both. Because the decedents in the Bogalusa Heart Study had been examined in previous school-based surveys, the earlier measurements of blood lipids, blood pressure, and body mass index (BMI, weight/height2 measured in kg and m) could be analyzed in relation to the extent of atherosclerosis present at autopsy (Table 3-1).29 Separately for Black and White males and females, correlations were calculated between various combinations of these measurements (systolic blood pressure and BMI with different blood lipid measures) and the extent of aortic and coronary artery lesions. Although nearly all of the correlations were statistically significant in the total pool of 93 cases, the sex-race-specific groups showed some variation. For example, for White females (only 19 in number) correlations with fibrous plaques in the coronary artery were exceptionally weak. For Black females, too, these correlations were weaker than among males though stronger than among females. These and other observations hint at some degree of relative protection of females, especially Whites, from the effects of blood pressure, BMI, and blood lipids. This is consistent with the impression from Figure 3-5 of young women being favored over young men in extent of coronary atherosclerosis. An early report of the PDAY Study presented results for 390 males with postmortem examinations at ages from 15 to 34 years.32 In addition to tissue samples, the study obtained extensive information at the time of death regarding blood lipid concentrations, blood thiocyanate levels (to indicate exposure
46894_CH03_039_058.qxd
50
2/25/10
3:24 PM
Page 50
CHAPTER 3 ATHEROSCLEROSIS
Figure 3-5 Extent of Fatty Streaks and Raised Lesions in the Abdominal Aorta (A and B) and Right Coronary Artery (C and D) in Men (white) and Women (black) Aged 15–34 Years from the PDAY Study. Source: Reprinted with permission from HC McGill Jr. et al., Atherosclerosis, Thrombosis and Vascular Biology, Vol 17, p 98, © 1997, American Heart Association.
to cigarette smoke), weight and height (for calculating BMI), and other characteristics. Figure 3-6, taken from that report, demonstrates the relation of percentage surface area of the abdominal aorta involved with any lesions to blood lipid concentrations and to smoking status. Taking these data into account allowed derivation of a statistical prediction of the extent of abdominal aortic atherosclerosis for males at each year of age in relation to these factors. For the group with unfavorable blood lipids who were smokers, the surface involvement was above 30% at age 15 and 50% or greater at age 34; by contrast, for the group with favorable blood lipids who were not smokers, the corresponding percentages of surface involvement were about 10% at age 15 and less than 25% at age 34. A close link was apparent between these factors and the extent of early aortic atherosclerosis.
It was of interest to investigate these relationships further, by assessing the ability to predict the extent of atherosclerosis in the coronary artery from composite scores reflecting multiple factors. By 2002, the Bogalusa Heart Study could report on examination of 517 individuals at ages 20 to 38 years with use of CIMT to indicate subclinical atherosclerosis.33 The combinations of measures considered in the analysis included up to three or more of the following: total cholesterol to high-density lipoprotein (HDL) cholesterol ratio, waist circumference, systolic blood pressure, insulin concentration, and smoking. For each of three anatomical locations in the carotid artery, IMT was consistently associated with increasing numbers of these factors. CIMT was recommended as a marker of the cumulative effects of these factors at early ages and as a stimulus to preventive measures.
0.47 0.09 0.50 0.16
Source: Am J Cardiol © 1998, Excerpta Medica, Inc.
Calculated as sum of the study-, age-, race-, and sex-specific z-scores of risk factor variable combinations shown in the table. Sample size varies based on variables used. BP = blood pressure; BMI = body mass index; HDL = high-density lipoprotein; LDL = low-density lipoprotein. *p 0.001. † p 0.01. ‡ p 0.0001. § p 0.05.
0.56§ 0.56§ 0.53§ 0.63†
0.93† 0.51 0.99‡ 0.35
0.53‡ 0.26§ 0.54‡ 0.49‡
0.61* 0.26 0.67‡ 0.65*
Systolic BP, BMI, triglycerides, and total/HDL-cholesterol combination Fatty streak, aorta Fibrous plaque, aorta Fatty streak, coronary Fibrous plaque, coronary
0.61 0.68 0.99‡ 0.35
0.41* 0.27§ 0.58‡ 0.51‡
0.53 0.09 0.51 0.17
0.53† 0.26 0.76‡ 0.66*
Systolic BP, BMI, and triglycerides combination Fatty streak, aorta Fibrous plaque, aorta Fatty streak, coronary Fibrous plaque, coronary
3:24 PM
0.35 0.40 0.37 0.58†
0.57‡ 0.23 0.54‡ 0.44*
Total (N ⴝ 93)
Spearman Correlation Coefficients Between Extent of Aortic and Coronary Artery Lesions and Multiple Risk Factor Indices White Black White Black Males Males Females Females (n ⴝ 41) (n ⴝ 23) (n ⴝ 19) (n ⴝ 10) Systolic BP, BMI, and LDL-cholesterol combination Fatty streak, aorta 0.61* 0.67† 0.43 0.52 Fibrous plaque, aorta 0.18 0.60§ 0.07 0.68 Fatty streak, coronary 0.63* 0.45 0.61§ 0.90 Fibrous plaque, coronary 0.56† 0.59§ 0.09 0.35
2/25/10
Table 3-1
46894_CH03_039_058.qxd Page 51
ATHEROSCLEROSIS IN CHILDHOOD, YOUTH, AND EARLY ADULTHOOD 51
46894_CH03_039_058.qxd
3:24 PM
Page 52
CHAPTER 3 ATHEROSCLEROSIS
15–24 y, Prevalence, %
52
2/25/10
100
100
80
80
60
60
40
40
20
20
25–34 y, Prevalence, %
0
#0
1–5
$11
6 – 10
0
100
100
80
80
60
60
40
40
20
20
0
#0
1–5
$11
6 – 10
Risk score due to modifiable risk factors
0
Grades 1 – 5 Grades 2 – 5 Grades 3 – 5 Grades 4 – 5 Grade 5
#0
1–5
$6
#0
1–5
$6
Risk score due to modifiable risk factors
Figure 3-6 Prevalence of AHA Lesion Grades by Category of PDAY Coronary Artery Risk Score Computed from the Modifiable Risk Factors for Men (Left) and Women (Right) and 10-Year Age Group. Source: From Pediatrics Vol 118, p 1450, © 2006, American Academy of Pediatrics.
The PDAY Study took the approach of deriving a risk score to predict the earliest and the most advanced atherosclerotic lesions.34 The score was based on non-HDL-cholesterol, HDL-cholesterol, smoking (present or absent), high blood pressure (present or absent, from anatomical indicators), BMI, and hyperglycemia. All were considered “modifiable” factors, distinguished from age and sex, which were regarded as “immutable” factors. Specifically for younger and older subjects and for men and women, prevalence especially of intermediate and more advanced lesions increased significantly with increasing risk score (Figure 3-6). Notably, women in the highest risk stratum were too few for separate analysis and were therefore combined into the third level of risk; also, prevalence of lesions at each level of severity was less in women than in men at every level of the risk score. The PDAY investigators took one further step. They applied the score derived from their 15- to 34year-old subjects to earlier data from the Community Pathology Study (CPS) in Orleans Parish, LA, to which many of the same investigators had contributed in the 1970s and 1980s.35 The PDAY score was used in two separate age strata from CPS: 15 to 34-year-olds and 35 to 54-year-olds. The overall result was strong association of the PDAY risk score with extent of coro-
nary atherosclerosis in both the same age range as PDAY and the later ages observed in CPS alone. This is depicted in Figure 3-7, by sex, for each of three coronary arteries (right, left circumflex, left anterior descending; see Figure 4-1). Deaths in the CPS data are categorized as “basal” (external, noncardiovascular causes), “related” (deaths with cardiovascular conditions present but not as cause of death), or “CHD” (death thought to be due to coronary heart disease). The extent of surface area involved was substantially greater for “CHD” or “related” than for “basal” deaths, greater for men than for women in each category, and greater in every instance for cases scored above the median than at or below the median value. The authors concluded that “The results from combined PDAY and CPS data suggest a seamless progression of the effects of the modifiable risk factors on atherosclerosis from 15 to 54 years of age.”35, p 371 A previous comparative study of the pathology of atherosclerosis in the United States and Japan showed substantially less extensive involvement of the aorta and coronary arteries in men in Tokyo than Black or White men in New Orleans, all autopsied at ages 25 to 44 years.36 Notably, the differences between New Orleans Whites and Japanese men was not in fatty streaks but in raised lesions as a percentage of the surface area of either aorta or coronary arteries. This
46894_CH03_039_058.qxd
2/25/10
3:24 PM
Page 53
ATHEROSCLEROSIS IN CHILDHOOD, YOUTH, AND EARLY ADULTHOOD
Right Coronary Surface Area Involved (%)
Men 80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
Left Circumflex Surface Area Involved (%)
Basal
Related
CHD
0
50
50
40
40
30
30
20
20
10
10
0
Left Anterior Descending Surface Area Involved (%)
Women
80
0
Basal
Related
CHD
0
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
Basal
53
0 Related CHD Cause of Death Category
Basal
Related
Basal
Related
Basal
Related
Figure 3-7 Mean Extent of Intimal Surface Area Involved with Raised Lesions in CPS Cases 35–43 Years of Age in the Right (Upper Panels) and Left Circumflex (Middle Panels), and Left Anterior Descending (Lower Panels) Coronary Arteries for Men (Left Panels) and Women (Right Panels) by That Part of the PDAY Risk Score Computed from the Modifiable Risk Factors (gray bars 4 [median]; black bars 5) and Cause of Death Category, Adjusted for Age. Source: From Atherosclerosis Vol 190, p 374, © 2006, Elsevier Ireland Ltd.
finding accords with long-recognized differences in coronary heart disease death rates in Japan relative to those in the United States and other industrialized countries. When a study of pathology of the aorta and coronary and cerebral arteries in relation to premortem factors such as blood lipids and blood pressure was undertaken in Japan, medical records were used as the source of information on these factors.37 The extent of more advanced atherosclerotic lesions was quite limited in the aorta, but in the coronary and
especially the cerebral arteries, fibrous plaques constituted a much-increased proportion of the lesions found. The extent of lesions in the aorta was primarily related to the age at death; in the coronary arteries, age, cholesterol, and blood pressure were all significantly related; and in the cerebral arteries, blood pressure was strongly and singularly related to a statistically significant degree. This report adds to the other insights to the epidemiology of atherosclerosis, especially the contrasting appearance of this
46894_CH03_039_058.qxd
54
2/25/10
3:24 PM
Page 54
CHAPTER 3 ATHEROSCLEROSIS
disease in different regions of the circulation. Further, it indicates that factors thought to accelerate the development of atherosclerosis even in its early stages relate differently to the disease in these different anatomic locations. A further observation on the relation between factors measured in childhood and the development of atherosclerosis is based on the long-term follow-up of children examined in school and recontacted as young adults in their early 30s in the Muscatine Study organized by Lauer and colleagues.38 This project in Iowa began in the 1970s and later included examination by UFCT scan of the coronary arteries to detect calcified coronary lesions. The results supported those described above, indicating that factors such as weight/height index, blood lipids, and blood pressure, all measured in the school years, predict the finding of these lesions at ages as early as the 30s. More recently, prehypertension (intermediate ranges of both systolic and diastolic pressure) before age 35 years was found to predict degrees of coronary calcium at a mean age of 44 years during follow-up in the CARDIA Study.39
PREVENTION AND TREATMENT OF ATHEROSCLEROSIS Occurrence of atherosclerosis in childhood, youth, and early adulthood and its relation to multiple modifiable factors has been emphasized in this discussion in part because these are the life stages when it first appears and because this period of life is not directly germane to coronary heart disease, stroke, and related conditions, as described in Chapters 4–6. Rather, the present discussion sets the stage for appearance of those conditions later, especially in middle and older adult years. Demonstration through the Bogalusa Heart Study, PDAY, and others that identifiable and modifiable factors are strongly and consistently associated with even the earliest lesions of atherosclerosis supports the view that preventive measures applicable from childhood could in principle be effective in reducing or eliminating the risks due to atherosclerosis. Despite the lack of direct evidence of effects on early development of atherosclerosis, preventive recommendations have been made with the expectation that their implementation in childhood would have the benefit of avoiding or delaying progressive change in atherosclerotic lesions that is the basis for later clinical manifestations. In an expert committee report from the World Health Organization in 1990, for example, encouragement was given to development of national policies in support of improvements in
diet and physical activity, as well as elimination of tobacco use, in childhood and youth.40 These and other such recommendations have to date undergone limited evaluations for their effect on blood lipids, blood pressure, and weight/height indices but none with respect to atherosclerosis itself. Much could be gained in prevention of atherosclerosis if increasingly practical methods for its early detection were applied to assess and monitor changes over time, whether as natural history or in response to interventions. Treatment of atherosclerosis is essentially treatment of its presenting manifestations in a particular patient at a given stage in its development and progression. Atherosclerosis in any individual patient is regarded mainly as a localized disease where it first appears clinically—chiefly in the coronary, cerebral, or peripheral arteries or the aorta. The question of whether treatment reduces the risk of complications, recurrence, or death due to atherosclerosis becomes a question related to one or another of these specific conditions. During late stages of the disease, however, regression of lesions has been clearly demonstrated in clinical trials of behavioral, medical, or surgical intervention. By 1992, for example, 10 studies of the effect of cholesterol reduction with one or another method of angiographic evaluation had already indicated that reducing cholesterol concentrations by approximately 40% resulted in significant benefit in terms of reduced progression, or even regression, of atherosclerotic lesions.41 More recent trials have shown similar outcomes.7 Treatment of the specific conditions is addressed to only a limited extent, in Chapters 4–6, in which the primary emphasis is on their prevention. Treatment guidelines that include these conditions are discussed further in Chapter 20. It should be noted that the “hypertensive” component of the atherosclerotic and hypertensive diseases is discussed not here in Part II but in Part III, Chapter 12, among determinants of cardiovascular diseases rather than the major conditions. This is because it occupies a different place in the cardiovascular disease spectrum than does atherosclerosis. The term “high blood pressure” is often used in place of “hypertension” to recognize that these terms refer not to a discrete disease state but to the upper levels in a continuous distribution of the physiologic trait, blood pressure. As discussed later, high blood pressure is both a significant public health problem in its own right—posing challenges for prevention and control— and a major factor in occurrence of coronary heart disease, stroke, and heart failure. Its impact is largely independent of atherosclerosis (as in causing hemorrhagic stroke) or superimposed on atherosclerosis (as in coronary heart disease), as will be discussed.
46894_CH03_039_058.qxd
2/25/10
3:24 PM
Page 55
REFERENCES
CURRENT ISSUES Among issues in atherosclerosis today, one is the potential for prevention of the progressive advance from earliest (levels 1–2) to intermediate (level 3) to advanced (levels 4–5) atherosclerotic lesions. Given the ubiquitous occurrence of the earliest lesions, preventing the condition altogether may be impractical, whereas slowing or preventing progression—at least to the threatening advanced stages—may be feasible. How to detect intermediate or early-advanced lesions by noninvasive means with sufficient sensitivity remains a challenge, but technological advances may make this possible. A second issue is to deepen understanding of the relative advantage of females over males in early development of atherosclerosis—an advantage that extends well into the adult years. Is this a universal circumstance? If there are exceptions, where and why do they occur? Can the explanation be applied to increase early protection for males as well? Third, what is the global distribution of atherosclerosis at the earliest stage detectable by noninvasive measures? Can populations be monitored, and if so, at how early an age, to assess whether the determinants of its progression are operating in the preadult years to produce a continuing, or accelerating, epidemic of atherosclerosis in one or another region of the world? Implementing existing standardized pathology protocols may make postmortem examination of samples of decedents the method of choice for monitoring the status of epidemic atherosclerosis in population samples around the world. REFERENCES 1. St. Clair RW. Biology of atherosclerosis. In: Pearson TA, Criqui MH, Luepker RV, Oberman A, Winston M, eds. Primer in Preventive Cardiology. Dallas, TX: American Heart Association; 1994:11–24. 2. Stary HC, Chandler AB, Dinsmore RE, et al. A definition of initial, fatty streak, and intermediate lesions of atherosclerosis: a report from the Committee on Vascular Lesions of the Council on Arteriosclerosis, American Heart Association. Circulation. 1994;89:2462–2478. 3. Stary HC, Chandler AB, Dinsmore RE, Fuster V, et al. A definition of advanced types of atherosclerotic lesions and a histological classification of atherosclerosis. A report from the Committee on Vascular Lesions of the Council on Arteriosclerosis, American Heart Association.
55
Arteriosclerosis Thromb Vasc Biol. 1995;15: 1512–1531. 4. Moreno PR, Shah PK, Falk E. Determinants of rupture of atherosclerotic coronary lesions. In: Willich SN, Muller JE, eds. Triggering of Acute Coronary Syndromes: Implications for Prevention. Dordrecht (The Netherlands): Kluwer Academic Publishers; 1996:267–283. 5. Virmani R, Burke AP, Farb A, Kolodgie FD. Pathology of the vulnerable plaque. J Am Coll Cardiol. 2006;47:C13–C18. 6. Blankenhorn DH. Prevention or reversal of atherosclerosis: review of current evidence. Am J Cardiol. 1989;63:38H–41H. 7. Mukherjee M, Yadav JS. Carotid artery intimalmedial thickness: indicator of atherosclerotic burden and response to risk factor modification. Am Heart J. 2002;144:753–759. 8. Bild DE, Detrano R, Peterson D, et al. Ethnic differences in coronary calcification: the MultiEthnic Study of Atherosclerosis (MESA). Circulation. 2005;111:1313–1320. 9. Greenland P, Bonow RO, Brundage BH, et al. ACCF/AHA 2007 clinical expert consensus document on coronary artery calcium scoring by computed tomography in global cardiovascular risk assessment and in evaluation of patients with chest pain: a report of the American College of Cardiology Foundation Clinical Expert Consensus Task Force (ACCF/AHA Writing Committee to Update the 2000 Expert Consensus Document on Electron Beam Computed Tomography) developed in collaboration with the Society of Atherosclerosis Imaging and Prevention and the Society of Cardiovascular Computed Tomography. J Am Coll Cardiol. 2007;49(3):378–402. 10. Strong JP. The natural history of atherosclerosis in childhood. In: Williams CL, Wynder EL, eds. Hyperlipidemia in Childhood and the Development of Atherosclerosis. New York: Annals of the New York Academy of Sciences; 1991:9–15. 11. Kuller L, Borhani N, Furberg C, et al. Prevalence of subclinical atherosclerosis and cardiovascular disease and association with
46894_CH03_039_058.qxd
56
2/25/10
CHAPTER 3
3:24 PM
Page 56
ATHEROSCLEROSIS
risk factors in the Cardiovascular Health Study. Am J Epidemiol. 1994;139:1164–1179. 12. Kuller LH, Sutton-Tyrrell K. Aging and cardiovascular disease: use of subclinical measurements. Cardiol Clinics. 1999;17:51–65. 13. Stamler J. Established major coronary risk factors. In: Marmot M, Elliott P, eds. Coronary Heart Disease: From Aetiology to Public Health. Oxford (England): Oxford Medical Press; 1992:35–66. 14. Ross R, Glomset JA. The pathogenesis of atherosclerosis. Second of two parts. N Engl J Med. 1976;295:420–425. 15. Ross R. Atherosclerosis—an inflammatory disease. N Engl J Med. 1999;340:115–126. 16. Benditt EP, Barrett T, McDougall JK. Viruses in the etiology of atherosclerosis. Proc Natl Acad Sci. 1983;80:6386–6389. 17. Epstein SE, Speir E, Zhou YF, et al. The role of infection in restenosis and atherosclerosis: focus on cytomegalovirus. Lancet. 1996;348: S13–S17. 18. Patel P, Mendall MA, Carrington D, et al. Association of Helicobacter pylori and Chlamydia pneumoniae infections with coronary heart disease and cardiovascular risk factors. Br Med J. 1995;311:711–714. 19. Hall AJ. A lady from China’s past. National Geographic. 1974;45:661–681. 20. Sandison AT. Degenerative vascular disease in the Egyptian mummy. Med Hist. 1962;6:77–81. 21. Holman RL, McGill HC, Strong JP, Geer JC. The natural history of atherosclerosis. Am J Pathol. 1958;34:209–235. 22. Enos WF, Holmes RH, Beyer J. Coronary disease among United States soldiers killed in action in Korea. JAMA. 1953;152:1090–1093. 23. McNamara JJ, Molot MA, Stremple JF, Cutting RT. Coronary artery disease in combat casualties in Vietnam. JAMA. 1971;216: 1185–1187.
24. Virmani R, Robinowitz M, Geer JC, et al. Coronary artery atherosclerosis revisited in Korean War combat casualties. Arch Pathol Lab Med. 1987;111:972–976. 25. Tejada C, Strong JP, Montenegro MR, et al. Distribution of coronary and aortic atherosclerosis by geographic location, race, and sex. Lab Invest. 1968;18:509–526. 26. Report of a WHO Expert Committee. Prevention of Coronary Heart Disease. WHO Technical Report Series 678. Geneva (Switzerland): World Health Organization; 1982. 27. Newman WP III, Freedman DS, Voors AW, et al. Relation of serum lipoprotein levels and systolic blood pressure to early atherosclerosis. N Engl J Med. 1986;314:138–144. 28. Berenson GS, Wattigney WA, Tracy RE, et al. Atherosclerosis of the aorta and coronary arteries and cardiovascular risk factors in persons ages 6 to 30 years and studied at necropsy (the Bogalusa Heart Study). Am J Cardiol. 1992;70:851–858. 29. Berenson GS, Srinivasan SR, Nicklas TA. Atherosclerosis: a nutritional disease of childhood. Am J Cardiol. 1998;82:22T–29T. 30. Zieske AW, Malcom GT, Strong JP. Natural history and risk factors of atherosclerosis in children and youth: the PDAY study. Pediatr Pathol Mol Med. 2002;21(2):213–237. 31. McGill HC Jr, McMahan CA, Malcom GT, Oalmann MC, Strong JP. PDAY Research Group. Effects of serum lipoproteins and smoking on atherosclerosis in young men and women. Pathobiological Determinants of Atherosclerosis in Youth. Throm Vasc Biol. 1997;17:95–106. 32. Pathobiological Determinants of Atherosclerosis in Youth (PDAY) Research Group. Relationship of atherosclerosis in young men to serum lipoprotein cholesterol concentrations and smoking. JAMA. 1990;264:3018–3024. 33. Berenson GS for the Bogalusa Heart Study Research Group. Childhood risk factors predict adult risk associated with subclinical car-
46894_CH03_039_058.qxd
2/25/10
3:24 PM
Page 57
REFERENCES
diovascular disease: the Bogalusa Heart Study. Am J Cardiol. 2002;90(suppl):3L–7L. 34. McMahan CA, Gidding SS, Malcom GT, Tracy RE, Strong JP, McGill HC, Jr. Pathobiological determinants of atherosclerosis in youth risk scores are associated with early and advanced atherosclerosis. Pediatrics. 2006;118(4): 1447–1455. 35. McMahan CA, McGill HC, Gidding SS, et al. PDAY risk score predicts advanced coronary artery atherosclerosis in middle-aged persons as well as youth. Atherosclerosis. 2007;190(2): 370–377. 36. Ishii T, Newman WP III, Guzman MA, et al. Coronary and aortic atherosclerosis in young men from Tokyo and New Orleans. Lab Invest. 1986;54:561–565. 37. Tanaka K, Masuda J, Imamura T, et al. A nation-wide study of atherosclerosis in infants, children and young adults in Japan. Atherosclerosis. 1988;72:143–156.
57
38. Mahoney LT, Burns TL, Stanford W, et al. Coronary risk factors measured in childhood and young adult life are associated with coronary artery calcification in young adults: the Muscatine Study. J Am Coll Cardiol. 1996; 27:277–284. 39. Pletcher MJ, Bibbins-Domingo K, Lewis CE, et al. Prehypertension during young adulthood and coronary calcium later in life. Ann Intern Med. 2008;149:91–99. 40. WHO Expert Committee on Prevention in Childhood and Youth of Adult Cardiovascular Diseases. Prevention in Childhood and Youth of Adult Cardiovascular Diseases: Time for Action. Technical Report Series 792. Geneva (Switzerland): World Health Organization; 1990. 41. Blankenhorn DH. Lipoproteins and the progression and regression of atherosclerosis. Cardiovasc Rev Rep. 1992;13:52–56.
46894_CH03_039_058.qxd
2/25/10
3:24 PM
Page 58
46894_CH04_059_088.qxd
2/25/10
3:25 PM
Page 59
C H A P T E R
4 Coronary Heart Disease nization (WHO) MONICA Project is the largest undertaking to date in cardiovascular epidemiology and has investigated determinants of changes in incidence, case-fatality, and overall mortality from coronary heart disease in 38 populations in 21 countries. From both the knowledge gained to date and the remaining gaps in that knowledge, major issues in coronary heart disease have been identified that require attention from public health professionals if further progress toward prevention of epidemic coronary heart disease is to be achieved.
SUMMARY Coronary heart disease has been the major component of cardiovascular morbidity and mortality in much of the western industrialized world in recent decades and now matches this distinction globally. Because the initial coronary event is often rapidly fatal, and because risks of recurrence and death among survivors are high, prevention of first coronary events through interventions at both the individual and the population-wide level is a high priority in the overall approach to cardiovascular disease prevention. Several long-term epidemiologic studies undertaken in the United States and other countries beginning in the late 1950s stimulated development of standard methods for diagnosis and classification of cases. These studies, including the Seven Countries Study, documented differences in population-level measures of coronary heart disease occurrence between populations whereas others, such as the Framingham Heart Study, focused on differences in risks among individuals within a population. Long-term trends based on vital statistics demonstrate the rise and fall of the vast epidemic curve of coronary heart disease mortality in the United States throughout the 20th century. Although the declining death rates from coronary heart disease represent important progress, the total burden and disparities due to this condition are little changed and remain of paramount public health concern. In addition, chronic heart failure, a late complication of coronary heart disease, has risen sharply in prevalence in consequence of improved survival after acute coronary events. Shorter-term but also striking trends in coronary mortality—with some increasing rates—have occurred in recent decades in many countries. The World Health Orga-
INTRODUCTION The Coronary Arteries “Coronary heart disease” is one of several terms referring to atherosclerosis of the arteries supplying the myocardium, or muscle of the heart. These arteries are illustrated in Figure 4-1.1 The arteries are named descriptively after their typical anatomic configuration. At the lower right of the figure (corresponding to the lower left of the subject’s chest) is the apex of the heart. At the top and left side of the figure is the base of the heart. From the base, the principal arteries extend distally toward the apex and can be imagined to form a crown (corona) around the main muscle mass comprising the left and right ventricles, the pumps to the systemic arterial circulation and to the lungs, respectively. Reduction of blood flow through one or more of these coronary arteries or their branches can result in insufficiency of myocardial blood supply (ischemia). Myocardial cells are dependent on a continuous or, during exertion, increased supply of oxygen and nutrients and removal of metabolic products through
59
46894_CH04_059_088.qxd
60
2/25/10
3:25 PM
Page 60
CHAPTER 4 CORONARY HEART DISEASE
Left Coronary Artery
Circumflex Artery
Right Coronary Artery
Marginal Branch
First Septal Perforator Branch AV Nodal Artery
Posterior Left Ventricular Branch Posterior Septal Branches Posterior Descending Artery
Left Anterior Descending Artery
Diagonal Branch
Anterior Septal Branches
Figure 4-1 The Coronary Arteries—A Schematic View. Source: Hutter A, Scientific American Medicine, Dale DC, Federman DD, eds. 5 Cardiovascular Medicine, Subsection IX, © 1996 Scientific American, Inc. All rights reserved.
the coronary circulation. Consequently, they undergo injury and death if blood flow is interrupted and not restored within minutes. Unless emergency treatment is instituted to minimize it, the extent of the injury often widens, with increased risk of complications or death. Typical symptoms experienced by the victim include pain described as pressing, stabbing, or crushing and located especially beneath the sternum or breastbone or in the jaw, arms, or midback, often accompanied by sweating, faintness, and a sense of impending death. It is now recognized that these classic symptoms may be absent among women experiencing myocardial ischemia. This makes diagnosis in women more difficult and more easily overlooked than in men. Portions of the electrophysiologic conducting system that controls the rate and rhythm of cardiac contraction pass through the septum or wall between the right and left atria and ventricles.
Especially if the area of ischemia includes the septum, an abrupt disturbance of cardiac rhythm may result with loss of effective pumping action of the heart, resulting in immediate collapse and sudden death unless resuscitation is successfully applied. Course of the Individual Case Some common features of the course of the individual case of an acute coronary event are shown schematically in Figure 4-2. Four phases of the process are depicted from the perspective of its biological and clinical progression. For each phase the status of disease and time frame are noted. First, against a background of progressive atherogenesis over many years or several decades, described in the preceding chapter, advanced atherosclerotic lesions develop in the coronary arteries. Second, under the current concept of the acute coronary event, one or more of several po-
46894_CH04_059_088.qxd
2/25/10
3:25 PM
Page 61
INTRODUCTION
61
Figure 4-2 Common Features of the Course of Coronary Heart Disease.
tential precipitating factors leads to disruption of an advanced atherosclerotic lesion or plaque, either at its endothelial surface or by fissuring into deeper levels of the lesion.2 Stimuli to thrombosis at the site of the plaque may produce only minor aggregation of blood platelets, with spontaneous resolution of the resulting small clot and some enlargement of the plaque. But alternatively, a large occlusive clot (thrombus) may form and persist for several hours or longer, producing acute symptoms (unstable angina), heart attack (myocardial infarction), or sudden death. If the acute event is not rapidly fatal, any of several short- or long-term outcomes is possible: recovery, with or without symptoms or residual cardiac dysfunction; short-term fatality (often defined as occurring within 28 days of clinical onset); later recurrence as a new episode (defined as symptoms present after a 28-day period from onset of an earlier event); or late coronary death, occurring more than 28 days (and up to many years) following the onset of the first event. Asymptomatic disease, or “silent” infarction, is another late outcome shown in Figure 4-2. This represents a circumstance in which evidence of myocardial infarction may be found on electrocardiographic (ECG) or echocardiographic examination and, for reasons not well understood, no history of chest pain can be elicited. Characteristics of the “vulnerable plaque” that predispose it to disruption, the precipitating factors triggering this process, and the consequent “acute coronary syndromes” including “sudden cardiac
death” are discussed extensively in published reports.2–8 Although sudden cardiac death is a common outcome of this process, definition has been controversial. One example is the following: “a natural death due to cardiac causes, heralded by the abrupt loss of consciousness within 1 hour of the onset of acute symptoms. Preexisting heart disease may or may not have been known to be present, but the time and mode of death are unexpected.”5, p 742 Previously, a 24-hour criterion was often used, and evidence of prior cardiac disease was sometimes taken to exclude cases. As currently defined, sudden cardiac death may represent 50% of all cardiovascular deaths in the United States. The other acute manifestations include additional deaths occurring within 28 days of onset (conventionally the limit for inclusion in casefatality), later-occurring deaths, or long-term nonfatal outcomes. Especially for unstable angina, there is a need for standardized definition and criteria to render reports of this condition comparable. Several aspects of the acute-stage process are particularly noteworthy from an epidemiologic perspective. First, this phase of the disease, which evolves in a time frame of seconds to hours, is a very late development in relation to the long-standing process of atherogenesis. Prevention of atherosclerosis may be opportune throughout much of the life span, whereas effective intervention within seconds, minutes, or even hours of the onset of an acute coronary event poses often insurmountable obstacles. Prevention of either irreversible myocardial damage or sudden death
46894_CH04_059_088.qxd
62
2/25/10
3:25 PM
Page 62
CHAPTER 4 CORONARY HEART DISEASE
requires effective action long in advance for those who are at risk. Second, investigation of precipitating or triggering factors is in progress at multiple levels, including both extrinsic factors (e.g., physical activity, psychological factors and emotional states, and physiological aspects of circadian variation) and intrinsic factors (especially physical and biochemical aspects of the plaque, blood coagulation, and blood vessel contractility or spasm). Third, in order for acute-phase interventions to be useful, they must be rapidly accessible. Studies and well-evaluated programs are needed to improve recognition of symptoms at their onset, quick action to initiate emergency care and transport, and rapid access to a qualified hospital emergency department for diagnosis and treatment.
BACKGROUND In a detailed historical account of coronary heart disease, Liebowitz traced to Egyptian and Greek antiquity reference to apparent cases of this disease.9 (Chapter 3 also includes references to cases in Chinese and Egyptian mummies.) Only gradual progress was made in understanding the clinical and pathological characteristics of coronary heart disease until the 19th century. Publication in 1896 of Osler’s Lectures on Angina Pectoris and Allied States provided numerous case descriptions and a wide spectrum of manifestations of coronary ischemia, including accounts of cases in persons of note from the 18th century and others compiled from his own practice.10 Osler commented on the rarity of this condition in hospital practice but noted that it was much more commonly observed in private consultation and especially in wealthy patients. As of 1900, accumulated knowledge of coronary atherosclerosis and angina pectoris did not yet include full recognition of the link between the acute thrombosis of a coronary artery and the occurrence of myocardial infarction. This understanding awaited Herrick, whose publications in 1912 and 1919 are classic descriptions of pathologic and electrocardiographic findings in acute myocardial infarction.9 The first half of the 20th century saw developments enabling later emergence of epidemiologic studies: clear recognition of the clinical and pathological entity and commonly applicable diagnostic procedures, which included electrocardiography. “Geographical pathology,” comprising somewhat fortuitous observations in populations around the world, pointed to variations in dietary patterns in relation to the frequency of coronary atherosclerosis. Keys’s account of
the genesis of the Seven Countries Study (discussed as follows) indicates the contribution of such observations from China, Java, the Netherlands, and wartime Scandinavia and Germany.11 By the 1950s and early 1960s, a remarkable series of epidemiologic studies had begun, of which some continue to the present. Examples include the Seven Countries Study, in which 16 groups of men, more than 12,000 in all, were examined in one of the seven participating countries under the leadership of Keys and his local colleagues;12 the Framingham Study and other community- or employment-based studies of more than 8000 men, collectively;13 the Ni-Hon-San Study of three cohorts of men of Japanese ancestry in Japan, Hawaii, and the San Francisco Bay Area, organized after discovery of striking gradients across the three populations in mortality from coronary heart disease and (oppositely) stroke;14,15 and other studies in the United States and elsewhere. The goal of these studies was to identify factors that could explain differences in rates of coronary heart disease between populations or in risks of coronary events among individual members of a particular population. Accounts of the organization and implementation of these early studies are rich sources of insight into the formative period of modern cardiovascular epidemiology. For example, in 1951 Dawber, Meadors, and Moore described the background of the Framingham Study.16 Sample size estimation was a necessarily inventive exercise. The authors’ concern about the potential for successful follow-up of participants for as long as 20 years was revolutionary but underestimated the study’s longevity by more than 45 years (thus far). The history of the Framingham Heart Study has recently been recounted in the context of the coronary epidemic.17 Three reports, from anecdotes of the key field study organizer (Blackburn) to a scientific update after 35 years’ experience, offer unique perspectives on the Seven Countries Study, a pioneering study of international contrasts in coronary heart disease rates and their determinants.18–20 Results of these early studies led to clinical trials and community interventions to test the ability to modify one or more of the factors identified and thereby to control or prevent coronary heart disease. Some major examples, addressed in subsequent chapters, are studies of diet (e.g., the Diet-Heart Feasibility Study), high blood pressure (Hypertension Detection and Follow-Up Program), blood cholesterol concentrations (Lipid Research Clinics—Coronary Primary Prevention Trial), or multiple factors in individual participants (World Health Organization European Collaborative Trial of Multifactorial Prevention of
46894_CH04_059_088.qxd
2/25/10
3:25 PM
Page 63
POPULATION STUDIES: DEFINITION AND CLASSIFICATION, DIAGNOSTIC ALGORITHMS, AND CRITERIA
Coronary Heart Disease and Multiple Risk Factor Intervention Trial). Whole communities became subjects of intervention to modify risk-related behavior (diet, physical activity, cigarette smoking, and others) in North Karelia, Finland; in California, Minnesota, and Rhode Island; and in numerous studies elsewhere (see Chapter 21). Trials on prevention of coronary heart disease and observational studies, all in adult populations, continued. Meanwhile, other investigations were undertaken as early as the 1970s to clarify factors in the early onset and progression of atherosclerosis. As reviewed in Chapter 3, these studies reconfirmed evidence for extensive atherosclerosis in the coronary arteries of some individuals in adolescence and early adulthood. Factors related to atherosclerosis and coronary heart disease in adults (e.g., adverse blood lipid profile, high blood pressure, and smoking) were also shown to predict the extent and severity of atherosclerosis in childhood and adolescence. This research has been progressive, developing from the earliest observational to the most recent experimental studies. Multicenter trials and demonstration programs are needed to continue this development. But observational epidemiology remains essential to investigate new questions, in new populations, and under circumstances changed dramatically from those studied earlier in this halfcentury of investigation.
POPULATION STUDIES: DEFINITION AND CLASSIFICATION, DIAGNOSTIC ALGORITHMS, AND CRITERIA Undertaking population studies of coronary heart disease in the 1950s and 1960s pointed to the need to standardize definitions and classification for improved comparability of data across studies. For determining the presence or absence of coronary heart disease among participants in general population surveys, information about participants’ personal health history was needed. In addition, electrocardiographic examination was needed both to supplement the history and, because myocardial infarction can occur without symptoms, to detect previous silent infarction in cases where abnormalities persisted. Methods were developed for standardized history taking by interview or questionnaire (the London School of Hygiene or Rose questionnaire), and an objective procedure was devised for coding and classifying electrocardiographic findings (the Minnesota code). These essential tools for epidemiologic studies were incorporated
63
in a more comprehensive guide, Cardiovascular Survey Methods, first published in 1968 through the auspices of the Cardiovascular Disease Unit of the World Health Organization.21 These methods were suitable for discriminating between already-established (and surviving) cases and noncases of coronary heart disease. They served for estimating its prevalence (i.e., the affected proportion of the population at large) and identifying those free of the disease. The latter would constitute a “population at risk” or cohort of persons to follow long-term and identify predictors of first coronary events. Diagnostic standardization was needed for hospital-based myocardial infarction case registers, community surveillance of newly occurring cases, and trials with new events as end points for evaluating the benefits of interventions. Acute changes in the electrocardiogram and abrupt appearance in the blood of increasing concentrations of myocardial cell enzymes were added to the diagnostic criteria for an evolving, or ongoing, myocardial infarction. In addition, a more detailed classification was developed to recognize particular characteristics of the acute event and distinguish among levels of confidence in the diagnosis. An algorithm for diagnosis and classification of acute coronary heart disease was formulated by Gillum and colleagues, under the auspices of the Criteria and Methods Committee of the Council on Epidemiology and Prevention of the American Heart Association.22 It was adopted by the WHO MONICA (MONItoring Trends and Determinants in CArdiovascular Disease) Project, in which 38 centers in 21 countries conducted a decade-long study of coronary events. The diagnostic elements for coronary heart disease in MONICA are summarized in Table 4-1.23 Both fatal and nonfatal events are classified as “definite” or “possible” on the basis of completeness of the findings; the possibility of a case representing successful resuscitation from cardiac arrest is recognized; and provision is made for the fatal case in which the rapid time course or other factors preclude the collection of the data needed to classify the event as definite or possible infarction.24 Recent development of study methods in cardiovascular epidemiology is reflected in the updated Cardiovascular Survey Methods published by the World Health Organization in 2004.25 In addition, detailed consideration of changing diagnostic criteria and classification of coronary events has led to a report on case definitions for acute coronary events with special attention to needs and opportunities for investigation in epidemiologic and clinical research.26
46894_CH04_059_088.qxd
64
2/25/10
3:25 PM
Page 64
CHAPTER 4 CORONARY HEART DISEASE
Table 4-1
Summary of Diagnostic Elements and Classification Scheme for Fatal and Nonfatal Coronary Heart Disease Events in Population Studies
Diagnostic elements • electrocardiogram (up to four per acute attack) • myocardial enzymes • history of chest pain (including time of onset) • necropsy evidence Classification of events • definite infarction • possible infarction • ischemic cardiac arrest (resuscitated) • fatal cases with insufficient data • no myocardial infarction Source: Data from World Health Organization MONICA Project Investigators, The World Health Organization MONICA Project (Monitoring Trends and Determinants in Cardiovascular Disease): A Major International Collaboration, Journal of Clinical Epidemiology, Vol 41, pp 105–114, © 1988.
Anticipated use of diagnostic imaging techniques and wider application of new indicators of myocardial cell damage, such as cardiac troponins, increase diagnostic sensitivity and greatly complicate standardization of disease estimates over time or between areas with differential uptake of these newer methods. As that report notes, citing an earlier contribution on this issue from Yusuf and colleagues: “The combination of new diagnostic tests, changing disease presentation, increasing numbers of survivors, and the predicted incidence increases in developing countries all argue for better surveillance to establish valid rates and trends. Such improvement depends on consistent, reliable, and valid case definitions.”26, p 2544 The case is made by investigators in the Minnesota Heart Survey for reliance on Minnesota coding of electrocardiograms as the most consistent diagnostic tool for monitoring trends that include previous years’ experience as in the first 25-year period of that program, 1970–1995.27 Clinical categories are also changing, as summarized for example in reports from the American College of Cardiology and American Heart Association.28 Within the broader classification of coronary heart disease, a category of “acute coronary syndrome,” or ACS, is now recognized that includes acute myocardial infarction and unstable angina, a condition with unexpected onset of anginal chest pain, usually at rest. Depending on the presence at initial examination of particular electrocardiographic findings (ST-segment elevation) and presence of abnormal levels of myocardial biomarkers (such as tro-
ponins), cases of ACS may be classified as ST-elevation MI (STEMI), non-STEMI, or unstable angina. Perhaps 20 or 30% of cases with ACS have STEMI, but changing practices make this estimate uncertain for the reasons noted previously. Already, however, these changes in case identification have led to new management guidelines recognizing the revised classification.29 Comparison of coronary heart disease experience between populations and over extended periods has mainly been based on mortality data from national vital statistics. These data are collected under the system of the International Classification of Diseases (ICD) codes. Currently, the classification of coronary heart disease events and conditions is organized in the Tenth Revision (ICD 10) as indicated in Table 4-2.30 Angina pectoris (I20) is further subclassified under the ICD code into four categories: unstable angina, angina pectoris with documented spasm of the coronary arteries, other forms, and unspecified. Acute myocardial infarction (I21) is subclassified in accordance with the site of infarction and its extent, whether transmural (affecting the full thickness of the myocardial wall) or only subendocardial (limited to a few millimeters from the interior surface of the heart chamber). Subsequent myocardial infarction (I22) refers to recurrent events and is subclassified only by site. Certain current complications (I23) are conditions such as cardiac rupture or development of a defect in the interatrial or interventricular septum—not concurrent with the acute infarction but developing as a late complication. Other acute ischemic heart disease (I24) includes conditions such as coronary insufficiency or coronary thrombosis not developing into myocardial infarction. Chronic ischemic heart disease (I25) refers not only to healed or past myocardial infarction (without current symptoms) but also to coronary events with survival beyond 28 days. Widespread use of these methods has greatly facilitated collection and reporting of comparable data from population studies. Nevertheless, issues of com-
Table 4-2 I20 I21 I22 I23
Categories of Coronary Heart Disease (Ischemic Heart Disease)
Angina pectoris Acute myocardial infarction Subsequent myocardial infarction Certain current complications following acute myocardial infarction I24 Other acute ischemic heart disease I25 Chronic ischemic heart disease
46894_CH04_059_088.qxd
2/25/10
3:25 PM
Page 65
RATES
parability still require consideration in interpreting data from different populations or time periods.
RATES Occurrence of coronary heart disease in populations is measured in several ways, each contributing some insight into the epidemic process. Mortality data, usually expressed as the number of coronary heart disease deaths per 100,000 population per year, are readily available in the vital statistics for many countries and other geopolitical units, often for population subgroups by age, race, and sex, and in some cases over periods of several decades. These data are ordinarily based on the cause of death as recorded by the party completing the death certificate and subsequently coded by a nosologist in accordance with the current ICD procedures. Cause-specific mortality is the measure of the rate of loss of life due to the disease as coded. Limitations of death registration and assignment of cause of death are widespread in much of the developing world and are discussed extensively in the Global Burden of Disease Study documents.31 Incidence data indicate the number of newly occurring cases within a given period. Incidence of coronary heart disease is usually expressed per 1000 to 100,000 population per year (the smaller value of the denominator being used for groups with very high rates, such as those at older ages), or per 1000 personyears, and includes both fatal and nonfatal cases. Due to the special requirements for diagnosis and classification of first events, availability of incidence data is limited. The methods used are principally two. In one instance, community surveillance, or long-term monitoring of defined populations, is undertaken to detect the occurrence of new coronary events, for example, through hospital admissions or death notices. The other is cohort studies, in which members of a population have been examined individually and those free of coronary disease at the starting point (baseline) are followed up by surveillance methods, periodic reexamination, or both to detect new coronary events. Both methods can also provide mortality data, distinct from ordinary vital statistics in having the potential for diagnostic validation through methods specific to the individual study. It is especially difficult, except in a cohort study, to determine whether a particular event was in fact the first occurrence of coronary heart disease in that individual and therefore strictly an incident, not a recurrent, event. Incidence rates, so defined, are taken to reflect the operation of factors that are causally related to the
65
occurrence of disease; inclusion of recurrent cases would cloud the interpretation due to factors influencing survival from previous events. Case-fatality is another useful measure. It indicates the proportion of all cases in the population, new and recurrent unless otherwise specified, that are fatal in outcome. Its determination requires knowledge of both the occurrence and the outcome of the event, obtained through surveillance or cohort studies. Case-fatality data are also therefore limited in availability. As already noted, the arbitrary interval of 28 days or less from the onset of symptoms is the convention used to attribute death to a given coronary event. A special subset of case-fatality is sudden death, variously defined. Case-fatality (i.e., within 28 days) is often interpreted as a measure of effectiveness (more properly, ineffectiveness) of medical treatment of acute coronary events. Sudden death has been considered distinct in being less amenable to medical care and is a particularly forceful indicator of the need to prevent acute coronary events altogether because a large proportion of these events are rapidly fatal. Prevalence, or the proportion of the population surviving with recognized coronary heart disease, is typically expressed in cases per 1000 population, usually for specific age groups, by sex, and often by race. Estimates of prevalence depend on knowledge of individual histories obtained through interview and examination surveys, as described previously. Surveys typically provide for only a single contact with participants and are much more readily conducted than surveillance or cohort studies. They are commonly undertaken as a first step in study of the coronary heart disease situation of a population. Prevalence does not include those who have died and may reflect factors influencing survival, so its interpretation warrants some caution. Nevertheless, it indicates something of the magnitude of risk and is often useful in estimating healthcare needs in a population. These several measures of disease occurrence in the population are interrelated in ways that bear on their interpretation. For example, incidence and casefatality can be considered as components of mortality because they indicate, respectively, the rate at which new cases occur and the proportion of cases with fatal outcomes. However, a death rate is usually calculated for the events in a 12-month interval, and case-fatality is restricted by definition to 28 days from the date of symptom onset. Therefore, the overall coronary death rate cannot be calculated from incidence and case-fatality alone, which does not account for all coronary deaths in 12 months among incident
46894_CH04_059_088.qxd
66
2/25/10
3:25 PM
Page 66
CHAPTER 4 CORONARY HEART DISEASE
cases and excludes recurrent cases. Also, prevalence is related to incidence and mortality because it identifies previously incident cases that were nonfatal. But differences in prevalence between populations, or changes in prevalence over time, cannot be attributed simply to differences in incidence or to differences in mortality without additional information. With this understanding of measures of disease occurrence, it is useful to examine data on coronary heart disease from these several perspectives, in the United States and elsewhere. Mortality The United States Statistical information on cardiovascular diseases in the United States is provided by the National Center for Health Statistics, on the basis of vital records and sample surveys of the US population, by other components of the Centers for Disease Control and Prevention (CDC), for example, from the Behavioral Risk Factor Surveillance System, and by special studies in selected communities such as the multicenter epidemiologic studies supported by the National Heart, Lung and Blood Institute—the Framingham Heart Study and others. Data from these sources are compiled and published by the American Heart Association in its encyclopedic annual statistical update28 (accessible at http://www.americanheart.org). Detailed tabulations of health data for the US population are also published annually by the National Center for Health Statistics. The most recent tabulation of mortality specific to ischemic heart disease was Health, United States, 2006, with mortality data current through 200332 (accessible at http://
www.cdc.gov/nchs/hus/htm). From Health, United States, 2006, the overall age-adjusted death rate per 100,000 population of all ages from ischemic heart disease in 2003 was 162.9–209.9 for men, 127.2 for women, 195.0 for all Blacks or African Americans, 114.1 for all American Indians or Alaska Natives, 92.8 for Asians and Pacific Islanders, 130.0 for Hispanics or Latinos, and 164.3 for non-Hispanic Whites. These values can be compared with the national objective for coronary heart disease mortality for the year 2010, 162/100,000, a level that had very nearly been reached early in the decade.33 However, wide disparities remain, as, for example, among Blacks or African Americans whose coronary mortality exceeds that of Whites by nearly 20%. Further insight to the burden of coronary heart disease in the United States, as of 2004, is presented in Table 4-3, based on data compiled in the AHA statistical update for 2009.28 Numbers of deaths are indicated for coronary heart disease (CHD) and specifically for myocardial infarction (MI), for ages 20 years. Of the nearly half-million CHD deaths, only slightly more occurred among men than women; onethird of these deaths were due to MI, and these too were only slightly more frequent among men. Numbers of deaths from these causes are shown for non-Hispanic Whites and Blacks but not for other population groups. The geographic distribution of CHD deaths by county throughout the United States was illustrated for all women aged 35 years and older in Figure 2-6. Corresponding county-level data and United States and state maps of CHD mortality for 1991–1995 for women and men by race/ethnicity are published and internet accessible34–36 (accessible at http://www.cdc.gov/cvh/maps).
Deaths from Coronary Heart Disease and Myocardial Infarction, Age 20 Years, United States, 2005 Coronary Heart Disease Myocardial Infarction Total 445,687 151,004 All males 232,115 80,079 % of total 52.1 53.0 All females 213,572 70,925 % of total 47.9 47.0 Non-Hispanic white males 203,924 70,791 % of subgroup 52.2 53.5 Non-Hispanic white females 186,497 61,573 % of subgroup 47.8 46.5 Non-Hispanic black males 22,933 7527 % of subgroup 49.8 48.4 Non-Hispanic black females 23,094 8009 % of subgroup 50.2 51.6 Table 4-3
Source: Data from Heart Disease and Stroke Statistics—2009 Update. A Report from the American Heart Association Statistics Committee and Stroke Statistics Committee. D Lloyd-Jones et al. © 2009, Courtesy of the American Heart Association/American Stroke Association.
46894_CH04_059_088.qxd
2/25/10
3:25 PM
Page 67
RATES
Europe and Elsewhere A major source of data on CHD morbidity and mortality in Europe is the WHO MONICA Project.23 The goal was to resolve the question of the 1978 Decline Conference of whether trends in coronary heart disease mortality were explained by changes in risk factors or treatment of cases. This would require data on key aspects of the coronary epidemic, including incidence, case-fatality, and overall coronary mortality. Participating countries were primarily in Europe, but populations in Australia and New Zealand, China, and the United States (the Stanford Five-City Study) were also included. The basic design called for monitoring of coronary heart disease events, related personal characteristics of population samples, and medical care practices over the 10-year period from the mid-1980s to the mid-1990s. At the start of the MONICA Project, coronary event registration under the Project protocol was reported in comparison with official coronary heart disease mortality for 38 populations in 21 countries, as shown in Figure 4-3.24 The figure lists study populations alphabetically by country and center and gives age-standardized annual coronary heart disease mortality for men and women, respectively. (See Appendix 4-A for site codes in MONICA.) The heavy horizontal bars represent the combined definite and possible coronary deaths according to MONICA registration, and the extended lines show the additional, nonclassifiable deaths. Official coronary heart disease mortality is represented by the short crossing vertical lines. Rates based on registered definite and possible coronary events and the official rates generally corresponded closely, although in some exceptional cases (e.g., Canada-Halifax, for men) the differences were quite large. (The scale for mortality rates is logarithmic and not arithmetic; this means that equal distances along the scale are equal multiples of the rate.) For example, the official rate for men in Finland-North Karelia (first entry in the left panel) was nearly 500/100,000, or about 10 times that of men in ChinaBeijing at 50/100,000. Similarly, a greater than 10fold range characterized the difference in rates between women in United Kingdom-Glasgow (approximately 110/100,000) and in Spain-Catalonia (11/100,000). Clearly, the United States has not been unique in its 20th-century experience with coronary heart disease mortality and did not rank highest in rates in the mid1980s; Finland, the United Kingdom, and the former Soviet Union were at or near the head of the list for both men and women. Global dimensions of ischemic heart disease are indicated in Chapter 1 as measured by estimated per-
67
centages of deaths from this cause in 2001 by income group and region (Table 1-5). Table 1-6 shows the rank order of ischemic heart disease among causes of death, years of life lost, and DALYs, worldwide. Here, the estimated number of ischemic heart disease deaths for each region and mortality stratum for 2002 is presented in Table 4-4, based on data from the World Health Report, 2004.34 The 7.2 million deaths from ischemic heart disease in 2002 constituted nearly half of all 16.7 million cardiovascular deaths and 12.6% of all 57 million deaths worldwide. Variation in numbers of deaths by region reflects both death rates and population size. Sudden Death Sudden death is a large component of CHD deaths. Figure 4-4 is based on a 1993 report that combines information on sudden death from studies collected from a 20-year period.35 “Overall incidence” (in this case a measure of mortality because the incident events are all fatal) is reported to be 0.1 to 0.2% of the population or about 300,000 events per year. In subgroups of the population defined by various predictors (left panel), much higher proportions become victims of sudden death: approximately 2% of those with high CHD risk; 5% of those with a prior coronary event; 20% of those with heart failure; 25% of survivors of a previous out-of-hospital cardiac arrest; and more than 30% of those with specific disturbances of heart rhythm during recuperation from a myocardial infarction. Conversely, of the total number of sudden deaths each year (right panel), successively smaller numbers of events come from these very-high-risk groups, each of which comprises only a very small part of the population. Incidence The United States First-time instances of myocardial infarction, true incident cases, can only be distinguished from recurrent events with reliable information about prior history of cardiovascular disease. This requires special studies designed to identify those with a history of prior CHD. Accordingly, Table 4-3, which illustrates several recent measures of CHD for the United States, refers to undifferentiated numbers of “New and Recurrent MI and Fatal CHD” and “New and Recurrent MI” among persons aged 35 years and older.28 These data indicate the numbers of persons potentially requiring emergency medical services, transport, and hospital care for acute coronary events, given that
46894_CH04_059_088.qxd
68
2/25/10
3:25 PM
Page 68
CHAPTER 4 CORONARY HEART DISEASE
Men
Women
FIN-NKA UNK-GLA FIN-KUO RUS-MOC
UNK-GLA UNK-BEL AUS-NEW RUS-MOI
LTU-KAU RUS-MOI UNK-BEL FIN-TUL CZE-CZE YUG-NOS SWE-NSW NEZ-AUC AUS-NEW CAN-HAL DEN-GLO POL-TAR USA-STA ICE-ICE POL-WAR SWE-GOT GER-AUU BEL-CHA GER-BRE AUS-PER BEL-LUX GER-AUR GER-RHN GER-EGE ITA-FRI ITA-BRI SWI-TIC BEL-GHE
RUS-MOC FIN-NKA NEZ-AUC LTU-KAU
CZE-CZE USA-STA YUG-NOS POL-WAR FIN-KUO FIN-TUL
GER-AUU BEL-CHA AUS-PER SWE-GOT DEN-GLO CAN-HAL GER-RHN GER-EGE
SWE-NSW ICE-ICE POL-TAR BEL-GHE CHN-BEI ITA-FRI GER-BRE
ITA-BRI GER-AUR FRA-STR
SWI-VAF FRA-LIL FRA-STR FRA-TOU
BEL-LUX FRA-LIL
SPA-CAT CHN-BEI
FRA-TOU SPA-CAT 200
100
50
20
MONICA Project: Definite and possible coronary deaths
10
Official CHD mortality
5
500
200
100
50
20
10
5
Mortality rate per 100,000 men age 35–64 years (log. scale)
Mortality rate per 100,000 women age 35–64 years (log. scale)
Official CHD mortality
MONICA Project: Additional unclassifiable deaths
MONICA Project: Definite and possible coronary deaths
MONICA Project: Additional unclassifiable deaths
Figure 4-3 Coronary Heart Disease (CHD) Mortality in 38 Populations in 21 Countries, According to the WHO MONICA Project and Official Rates. See key for MONICA abbreviations in Appendix 4–A. Source: Reprinted with permission from Circulation, Special Report, Vol 90, No 1, p 599, © 1994, American Heart Association.
46894_CH04_059_088.qxd
2/25/10
3:25 PM
Page 69
RATES
Table 4-4 Africa 1 160 1. 2. 3. 4. 5. 6.
2 172
69
Numbers of Deaths (ⴛ1000) Due to Ischemic Heart Disease by Mortality Strata in Each WHO Region, 2002 South-East Eastern Western The Americas Asia Europe Mediterranean Pacific Total 3 574
4 319
5 28
4 265
5 1774
3 672
4 464
6 1237
4 142
5 396
3 129
4 864
7208
high child, high adult mortality high child, very high adult mortality very low child, very low adult mortality low child, low adult mortality high child, high adult mortality low child, high adult mortality
Source: Data from World Health Report 2004, Statistical Annex, Table 2, pp 122–123.
these needs do not differ between first and recurrent events. But they do not measure the separate impact of efforts to prevent first events (“primary prevention”) or subsequent ones among survivors (“secondary prevention”). These are important questions for evaluating preventive strategies, a topic of later discussion. Incidence of coronary heart disease can be estimated most reliably through one of two types of studies. One type is study of a cohort of persons determined initially to be free of coronary heart disease at the start of several years’ follow-up and judged to have developed coronary heart disease or not on the basis of examination at a subsequent time. The second type of study is community surveillance, by which a defined population is monitored for the occurrence of events chiefly through
review of hospital admissions and out-of-hospital deaths, in contrast to periodic examination of all members of a cohort. This was the design of the WHO MONICA Project, illustrated above. Six current studies in the United States involving a total of 23 communities are described in detail in Incidence & Prevalence: 2006 Chart Book on Cardiovascular and Lung Diseases, prepared by the National Heart, Lung and Blood Institute.36 These are the Atherosclerosis Risk in Communities (ARIC) Study, the Cardiovascular Health Study (CHS), the Coronary Artery Risk Development in Young Adults (CARDIA) Study, the Framingham Heart Study (FRS), the MultiEthnic Study of Atherosclerosis (MESA), and the Strong Heart Study (SHS). (The SHS is noteworthy as a unique source of data for the American Indian/ Alaska Native population.)
Overall Incidence in Adult Population High-Coronary-Risk Subgroup Any Previous Coronary Event Ejection Fraction ,30%; Heart Failure Out-of-Hospital Cardiac-Arrest Survivors Convalescent Phase VT/VF after Myocardial Infarction 0
1
2
5 10 20 30
Percent/Year
0
100
200
300
(⫻ 1000) Events/Year
Figure 4-4 Sudden Cardiac Deaths Among Population Subgroups. Source: Reprinted with permission from Annals of Internal Medicine, Vol 119, p 1188, © 1993, American College of Physicians.
46894_CH04_059_088.qxd
70
2/25/10
3:25 PM
Page 70
CHAPTER 4 CORONARY HEART DISEASE
One of the six studies, the Atherosclerosis Risk in Communities (ARIC) Study, conducted both cohort and surveillance studies and contributed data from both methods to the Chart Book. Data from the two methods are generally in good agreement. The six studies differ in the age range and race/ethnicity of their study populations, as well as in geographic location. Some differences among studies are noted in diagnostic methods or criteria and the specific cardiovascular conditions addressed. Results are not strictly comparable among studies by reason of these differences, but agreement is generally close. The data presented are estimated mainly from the overall period of the 1990s. Age-, sex-, and race/ethnic-specific incidence rates per 1000 person-years of observation are presented separately for each study in graphic and tabular format. No information on trends in incidence is given in this report. The ARIC Study, for example, indicated incidence of CHD (MI or CHD death) per 1000 person-years in cohort follow-up, for the specific age group 55–64 years, to be 7.0 for White men, 7.7 for Black men, 2.7 for White women, and 4.7 for Black women. Rates were higher for men than women and higher for Blacks than Whites. Rates were higher for each sex-race group at age 65–74 than at age 55–64 years, and the relative increase was greater for Blacks than for Whites. Sex and race/ethnic patterns of age-standardized results
Table 4-5
Cohort Dalmatia Slavonia Tanushimaru East Finland West Finland Crevalcore Montegiorgio Zutphen Ushibuka Crete Corfu Rome railroad Velika Krsna Zrenjanin Belgrade Total
for ages 45–84 years were similar in relative rates to those described, for both CHD and MI. For angina pectoris, however, incidence was equal (about 11 per 1000 person-years) for Whites and for Black men but notably greater (about 18 per 1000 person-years) for Black women. Data based on the ARIC Study were also presented in the American Heart Association statistical update for 2009, shown in Table 4-3.28 Extrapolated to the US population, they provided estimates of 700,000 first coronary events and 500,000 recurrent events annually in the United States. Of a total of 1.2 million acute coronary events in one year in the United States, about 60% were projected as incident events and 40% as recurrences. Other Countries The cohort follow-up approach has also been used as a study design for comparing multiple populations. Incidence data from the Seven Countries Study at 10year follow-up are shown in Table 4-5.10 A familiar distinction is made in these data between “hard” and “any” coronary heart disease. The former category includes coronary heart disease death and definite myocardial infarction only, whereas the latter also includes angina pectoris and other less reliably documented conditions. Incidence of newly occurring
Ten-Year Incidence of Coronary Heart Disease (CHD) Among Men Free of Cardiovascular Disease at Entry (Age-Standardized Rate per 10,000), Seven Countries Study, 1958–1964 to 1968–1974 Hard CHD Any CHD Total N N Rate SE N Rate SE 662 13 185 52 40 629 94 680 18 253 60 40 561 88 504 8 148 54 20 354 82 728 71 1074 115 201 2868 168 806 45 539 80 129 1582 129 956 43 450 67 105 1080 100 708 22 353 69 64 966 111 845 45 513 76 91 1066 106 496 11 204 63 23 458 94 655 2 26 20 13 210 56 525 17 337 79 37 686 110 736 25 357 68 57 786 99 487 6 132 52 21 452 94 476 12 239 70 37 715 118 516 13 317 77 35 794 119 9780 351 369.9a 19.1 913 943.8a 29.6
Note: SE, standard error. a Mean of the cohort rates weighted by the number at risk in each cohort. Source: Reprinted with permission of the publisher from Seven Countries by A Keys, Cambridge Mass; Harvard University Press, © 1980 by the President and Fellows of Harvard College.
46894_CH04_059_088.qxd
2/25/10
3:25 PM
Page 71
RATES
hard coronary heart disease events over the full 10year period ranged from 1074 per 10,000 in East Finland to 26 per 10,000 in Crete, a 40-fold difference. When all coronary heart disease events were included, the range was from 2868 per 10,000 in East Finland to 210 per 10,000 in Crete, more than a 10-fold difference. The proportions of events classified as hard coronary heart disease were compared in regional groupings of these cohorts and varied rather little, from 36.6% to 54.1%. This observation lends credence to the range of incidence rates reported, whereas wide variation in the proportion of hard coronary heart disease might suggest a systematic difference, or bias, in the identification or classification of cases between populations. The relation of baseline measures of factors thought to affect the risk of coronary events (e.g., diet, blood lipids, blood pressure, smoking) to event rates was analyzed at the population level. That is, for each of the 16 cohorts, an average value was estimated for each characteristic (e.g., baseline blood cholesterol concentration) and linked with the event rate for that population. For example, by regression analysis, the question of whether differences in average cholesterol concentration were related to differences in event rates was evaluated. With this ap-
71
proach, the contribution of these and other factors to population differences in event rates was assessed. Results of these analyses are addressed in subsequent chapters where the respective factors are discussed. More recent data for populations studied beyond the United States are provided by the WHO MONICA Project.24 Table 4-6 presents selected data on CHD events reported in MONICA, including event rates and 28-day case-fatality under each of two case definitions, as well as the proportion of fatal cases for which history of prior MI was unknown. Selected from the detailed report on all MONICA populations are the data for those with the highest and lowest event rates, the highest and lowest case-fatality rates, and the overall mean rates, all under definition 1, separately for men and women. Events were enumerated for this tabulation in four classifications that included different combinations of the following subsets: F1 definite fatal event; F2 possible fatal event; F9 unclassifiable fatal event; NF1 definite nonfatal event; and NF2 possible nonfatal event. “1st event” refers to those cases meeting definition 1 for which there was a known negative history of prior myocardial infarction. The final column indicates the percentage of participants
Table 4-6
Age-Standardized Annual Event Rates per 100,000 Population, 28-Day Case-Fatality, and Confidence Intervals for Different Definitions of Events in Men and Women Age 35 to 64 in Selected Study Populations, WHO MONICA Project Population—Men Definition 1 First Event Fatal Where F1F2F9NF1 Definition 1, History of MI No Previous MI Not Known, % Event Rate, Case-Fatality Event Rate Case-Fatality 95% CI 95% CI CHN-BEI 76 9 53 6 58 51 3 FIN-NKA 915 62 48 3 586 44 1 ICE-ICE 540 45 37 4 395 33 4 POL-TAR 465 26 81 2 45 12 93 Average (excludes 465 49 281 37 22 2 populations) Population—Women
CAN-HAL FRA-LIL SPA-CAT UNK-GLA Average
Definition 1 F1F2F9NF1 Event Rate, 95% CI 138 20 67 7 30 4 256 20 101
Case-Fatality 95% CI 31 6 68 5 46 8 49 4 54
First Event Definition 1, No Previous MI Event Rate
Fatal Where History of MI Not Known, % Case-Fatality
46 46 21 187 65
9 58 34 49 43
Source: Adapted from Circulation, Special Report, Vol 90, No 1, pp 598–599, © 1994, American Heart Association.
69 31 32 4 22
46894_CH04_059_088.qxd
72
2/25/10
3:25 PM
Page 72
CHAPTER 4 CORONARY HEART DISEASE
for whom history of myocardial infarction was unknown. Event rates were dependent, of course, on the inclusiveness of the definition. Definition 1 gave the largest numbers of events except when “possible nonfatal events” were added (not shown). First events were far fewer in number, especially when the percentage with unknown history was large. This striking demonstration of the dependence of rates on their definition underscores the need for explicit criteria when comparisons are made across studies. Population differences are apparent. For example, in reference to definition 1, rates ranged for men from 915/100,000 in Finland-North Karelia to 76/100,000 in China-Beijing, and for women from 256/100,000 in United Kingdom-Glasgow to 30/100,000 in Spain-Catalonia. Summary rates are given for comparison of overall male and female rates. Because two centers did not include women, they were excluded from the average values for men. Event rates were four to five times as great among men as among women. Case-Fatality On a national basis, case-fatality for acute myocardial infarction in the United States is unknown, although data are available from some of the community-based studies cited above. Data from the one US center in the WHO MONICA Project, the Stanford Five-City Study, gave the following estimates in parallel with those shown in Table 4-6: for definition 1 events, a rate of 508/100,000 and case-fatality of 50%; for first events under definition 1, a rate of 299/100,000 with case-fatality of 41% (history of MI unknown for 18% of fatal events). Again, the MONICA experience provides a unique source of information on this aspect of coronary heart disease in many populations, as seen in Table 4-6.24 Case-fatality differs in accordance with definitions as do event rates, for example, between the more inclusive definition 1 and first events with no previous MI, from 49% to 37% for men and from 54% to 43% for women. Under definition 1, overall case-fatality was 48–49% for men and 54% for women, with most centers reporting 40–55% casefatality for men and 45–70% for women. Overall, regardless of definition, case fatality was about 1.1 times as great for women as for men, after adjustment for differences in age at death. Prevalence Survivors beyond 28 days from onset of an acute coronary event constitute the known nonfatal cases. Survivors together with persons with silent infarc-
tion (detectable only through screening by electrocardiography, as conducted in surveys of the general population) constitute the true population alive with previous myocardial infarction. Surveys dependent on self-reported history necessarily underestimate this true total by missing the unknown cases. (See Table 4-3.28) Studies represented in the Incidence & Prevalence Chart Book overcome this limitation by direct examination that can reveal unrecognized CHD.36 Prevalence, like incidence, is reported for each of the six population studies by age, sex, and race/ethnicity as availability of data permits. Time periods vary among the studies, and trends in prevalence are not reported here. Prevalence was estimated in 2004 as 7.3% of the total US population aged 20 years or older, or 15.8 million persons (Table 4-3). Prevalence varies by sex and race/ethnicity, from 4.2% among Asians (18 years or older) to 9.4% among non-Hispanic White males. Between 1994 and 1995, the estimated prevalence of coronary heart disease in the United States nearly doubled, from 6.3 to 11.2 million. This abrupt increase was due to a change in classification that has since included persons with a self-reported history of angina pectoris (chest pain indicative of ischemic heart disease), or other evidence of coronary heart disease. Prevalent cases of CHD that are due to prior MI are at relatively high risk of cardiovascular events, including recurrent MI, sudden death, angina pectoris, heart failure, and stroke. Within 5 years after a first MI at age 40–69, 15% of White men, 22% of White women, 27% of Black men, and 32% of Black women will die; 16% of men and 22% of women will have a recurrent MI or fatal CHD; 7% of men and 22% of women will develop heart failure; 4% of men and 6% of women will have a stroke.28 Field surveys to estimate prevalence of coronary heart disease have been conducted in many populations since the 1950s. These surveys have usually been independent investigations without formal standardization, especially in many developing countries. One contrasting application of standardized methods for such surveys is illustrated in Table 4-7, which presents the electrocardiographic findings in surveys primarily of Pacific and Indian Ocean populations.37 When classified according to the strictest electrocardiographic criteria (Q-wave codes 1.1–1.2) by Minnesota code, “probable coronary heart disease” was found in 0–5.7% of men aged 35–59 years and in 0–2.8% of women; “possible coronary heart disease” was several times more common in most populations for both sexes but especially for women. Taking both probable and possible coronary heart disease into account, the combined prevalence was estimated to range from
46894_CH04_059_088.qxd
2/25/10
3:25 PM
Page 73
RATES
73
Table 4-7
Prevalence of Probable and Possible Coronary Heart Disease (CHD) in Developing Countries in Men and Women Aged 35 to 59 Years. By Ethnic Group, 1978–1987 Men Women Q and ST-T Q and ST-T Population Groups Qa ST-Tb Together Q ST-T Together Chinese Beijing—Chinese Mauritius—Chinese
0 2.5
1.5 15.0
1.5 (0–3.8) 16.6 (10.6–22.6)
0.2 0
4.1 27.5
3.6 (0.8–6.5) 27.2 (21.2–33.2)
Polynesian Cook Islands—Rarotonga Niue Western Samoa
3.5 0.9 1.1
3.1 2.6 2.2
6.3 (2.2–10.3) 3.3 (0–7.6) 2.9 (0–6.9)
0.7 0.7 0.3
18.2 9.9 12.0
19.3 (15.3–23.1) 10.8 (6.9–14.7) 11.9 (8.3–15.5)
Asian Indian Fiji Mauritius—Hindu Mauritius—Muslim
3.5 1.3 1.8
13.7 11.1 8.4
17.3 (13.2–21.4) 12.8 (10.2–15.3) 10.9 (5.8–15.9)
0.9 0.6 0.6
23.6 27.6 28.5
24.4 (20.7–28.2) 28.7 (26.1–31.2) 29.6 (24.7–34.5)
0 5.7 2.6
1.0 5.7 5.1
0.8 (0–7.2) 9.8 (0–20.9) 8.0 (4.7–11.4)
0 0 2.8
12.1 15.5 11.4
11.9 (5.3–18.4) 14.4 (5.8–23.0) 14.8 (11.1–18.4)
Melanesian—Fiji
2.5
6.4
8.7 (4.8–12.6)
1.0
17.5
18.7 (14.9–22.4)
Micronesian Kiribati Nauru
0 1.8
6.2 5.9
6.7 (4.0–9.3) 7.3 (2.9–11.7)
0.2 0.8
25.6 5.1
26.3 (23.6–28.9) 5.5 (1.3–9.5)
Creole—Mauritius
1.4
14.5
15.5 (11.7–19.3)
0.5
34.1
34.3 (30.9–37.6)
Melanesian/Polynesian Fiji (Lakeba) New Caledonia (Loyalty) New Caledonia (areas of Touho, Oundjo, Noumea, and Wallis Island)
a
Q: Probable CHD; Minnesota codes 1.1, 1.2. b ST-T: Possible CHD; Minnesota codes 1.3, 4.1, 5.1, 5.3, and 7.1.1. Source: Reprinted from Journal of Clinical Epidemiology, Vol 47, p 602. Copyright 1994 by Elsevier Science, Inc.
1.5% to 17.3% for men and from 4.1% to 34.1% for women. These estimates do not include selfreported angina pectoris, unlike the most recent data for the United States cited previously. Disability Among the consequences of CHD is substantial disability, leading to loss of employment and economic productivity, estimated to cost the US economy $10.6 billion in 2009 of $39.1 billion for all cardiovascular diseases. (This is additional to $62.0 billion in losses due to death and $92.8 billion in healthcare expenditures, from CHD alone.28) Quality of life among the population surviving with CHD is an additional concern. Beginning in 2000, the Medical Expenditure Panel Survey (MEPS) of the Agency for Healthcare Research and Quality (AHRQ) (accessible at http://www.meps.ahrq.gov/ mepsweb) has included a self-administered house-
hold questionnaire to assess health-related quality of life (HRQoL) among respondents, a national probability sample of noninstitutionalized adults 18 years of age or older.38 Four rating scales were used and addressed mental health, physical health, health utility, and self-rating of health. On every scale, persons with CHD reported reduced quality of life relative to those without CHD. Blacks were affected more than Whites on the majority of scores, and Hispanics were especially affected on the mental health score. Globally, the impact of ischemic heart disease on health-related disability has been estimated in terms of DALYs (disability-adjusted life years), as shown in Table 1-6. Ischemic heart disease was ranked fifth in contributions to total disability in 1990 and projected to rank first among all causes of disability by 2020.39 Further, the expected burden in terms of years of productive life lost due to ischemic heart disease in 2030 was projected for Brazil, South Africa, Russia,
46894_CH04_059_088.qxd
74
2/25/10
3:25 PM
Page 74
CHAPTER 4 CORONARY HEART DISEASE
China, and India, compared with Portugal and the United States, shown in Table 1-7.40 This measure included both death and disability. These measures underscore the personal and societal costs of CHD and aid in appreciating its full impact on health and well-being, beyond the more common reliance on more readily available but in some respects less informative death rates alone. Disparities Disparities in CHD add further to understanding its social impact, as discussed in detail in Chapter 2. Figures 2-1, 2-2, 2-6, and 2-7 illustrate differences in CHD mortality by age, sex, race/ethnicity, and geography in the United States. Figure 2-5 shows variation in prevalence of major cardiovascular conditions by race/ ethnicity among US adults, and Table 2-2 describes the “Eight Americas” that define race/ethnic subgroups of the US population with striking differences in mortality.41 Regional variation in magnitude of CHD mortality throughout the world is shown in Table 4-4. A fundamental issue regarding racial/ethnic variation in CHD and other health conditions concerns the concept of race itself and understanding the underlying characteristics that may distinguish one group, on average—or one individual—from another. The complexity of this issue, touched on in Chapter 2, is well illustrated in the case of persons of African origin, discussed with particular insight by Cooper, who suggests that molecular genetics may introduce, and implicitly better answer, questions of variation in health and disease within and between populations.42 Examination of distributions of determinants of atherosclerotic and cardiovascular diseases in the chapters that follow includes discussion of variation in their patterns by age, sex, race/ethnicity, and other factors as available data permit. That discussion will offer some insight into the observed disparities in CHD and the other major cardiovascular conditions.
RISKS Risk Factors Variation in individual risks of coronary heart disease events within a population indicates operation of factors at the personal level. A focus on individuals or subgroups with characteristics of special interest, in relation to other members of the same population, is the approach taken in several large cohort studies since the early 1950s. Although in design each such study is analogous to study of any one cohort in the Seven Countries Study, the objective was fundamentally different, that is, not to derive estimates of event
rates and relevant exposures for comparison between whole cohorts but to compare individual subjects as the units of observation and analysis. The conceptualization and implementation of the Framingham Heart Study in the United States at midcentury are addressed in detail by Dawber and coworkers.16 Similar investigations in other community or employment settings were undertaken at about the same time in the United States. Although the Framingham Heart Study has continued longest and is widely recognized on the basis of its exceptionally extensive collection and reporting of data, other studies of its type have also contributed importantly to current knowledge of the epidemiology of coronary heart disease. For example, it became apparent early in the course of these US studies that more definitive analysis of individual differences in risks of coronary heart disease might be achieved by combining data from those studies most alike in design and examination methods. Under the aegis of the Committee on Epidemiological Studies (now the Scientific Council on Epidemiology and Prevention) of the American Heart Association, and its Subcommittee on Criteria and Methods, discussions began in 1961 that culminated in formation of the US National Cooperative Pooling Project. With support from the American Heart Association and both the Heart Disease Control Program and the (then) National Heart Institute of the US Public Health Service, this project was the major source of data for research planning and study design in the area of cardiovascular disease epidemiology and prevention for many years subsequent to its initiation in 1964. In 1978, its landmark Final Report was published as a comprehensive presentation of the pooled analyses relating characteristics of 8422 men aged 40–64 years who were free of coronary heart disease at entry to each study to the subsequent occurrence of an initial coronary heart disease event.13 The five participating studies (Albany Civil Servants, Chicago Gas Company, Chicago Western Electric Company, Framingham, and Tecumseh) provided information based on 72,011 person-years of experience and the occurrence of 658 first major coronary events prior to age 65. Table 4-8 is reproduced from the Final Report and presents the results of multivariate analysis in which baseline diastolic blood pressure, serum cholesterol concentration, smoking status, and age were taken into account. This analysis was to assess the contribution of each of these characteristics to the probability of occurrence of a first major coronary event during the 8.6 years of observation for each man. The multivariate equation derived from the pooled experience yielded significant coefficients for all four factors. When the distribution of “expected risk” based on these results was
46894_CH04_059_088.qxd
2/25/10
3:25 PM
Page 75
RISKS
Table 4-8
Parameters SE t Standard No. of Men: 6875
75
Factors Related to the Incidence of Coronary Heart Disease, the US National Pooling Project, 1978 Diastolic Blood Serum Age Pressure Cholesterol Smoking 10.328022 0.068467* 0.030365* 0.006138* 0.304628* 0.008315 0.003584 0.000887 0.038270 6.234 8.473 6.920 10.050 0.349599 0.361013 0.293662 0.428591 No. of Events
Quintile of Expected Risk All I II III IV V Ratio: V/I % of events in V % of events in VI V Difference: V I
Risk/1,000 Men/ 8.6 Year
41.7 41.7–62.0 62.0–87.3 87.3–129.0 129.0
Expected 623.1 41.3 71.2 101.1 145.5 264.0 6.4 42.4 65.7 222.7
Observed 621 29 71 106 164 251 8.7 40.4 66.8 222
Rate/1,000 Men/8.6 Year Expected 90.6 30.0 51.8 73.5 105.8 192.0 6.4
Observed 90.3 21.1 51.6 77.1 119.3 182.5 8.7
162.0
161.4
*p 0.01 Source: Reprinted with permission from Journal of Chronic Diseases, Final Report of the Pooling Project, p 253, © 1978 Elsevier Science, Inc.
categorized into quintile groups (each comprising onefifth of the pooled study population), the expected and observed numbers of events and event rates could be examined in relation to the gradation of multivariate risk based on these four characteristics. Incidence of first major coronary events was 8.7 times as great in quintile V as in quintile I, and more than 40% of all events occurred in the group of men whose risk made up the highest 20% of the pooled population. This indicated the strength of the predictive relation between these four factors and risks of coronary heart disease for individuals. Notably, if the minimum risk in the pooled population were given by that of quintile I, the incidence of coronary heart disease was more than two times as great for the second quintile of risk (rate 51.6 versus 21.1), more than three times as great for the third quintile (rate 77.1 versus 21.1), nearly six times as great for the fourth quintile (119.3 versus 21.1), and nearly nine times as great for the fifth quintile (182.5 versus 21.1). These results for quintiles II–IV indicate that substantial excess risk occurred well below the highest-risk category. This indicates further that preventive measures limited to the highest-risk group could not effectively address all of the increased risk in the population. This principle is reinforced by studies of other populations. (See Chapter 18, Strategies of
Prevention.) It was clearly demonstrated 30 years ago by the Pooling Project. An example of more recent data for estimating risks of coronary events for individuals within a population is the exceptionally large cohort of men, also middle-aged Americans, who underwent risk-factor screening in the mid-1970s as potential candidates for entry to the Multiple Risk Factor Intervention Trial (MRFIT).43 Represented in Table 4-9 are the 342,815 men free of known prior heart attack or diabetes at the screening examination, which included measurements of blood pressure and serum cholesterol concentration and a questionnaire history of cigarette smoking. Death due to coronary heart disease over an average follow-up period of 11.6 years was ascertained through vital statistics sources. Because of the very large numbers of observations, it was possible to cross-classify the population according to quintile groups of systolic blood pressure (from 118 to 142 mm Hg), quintile groups of serum cholesterol concentration ( 182 to 245 mg/dl), and two smoking categories (nonsmokers versus smokers). For each of the resulting 50 groups, the death rate from coronary heart disease is presented. This follow-up experience through the 1980s indicates the same general relation of risk to these fac-
46894_CH04_059_088.qxd
76
2/25/10
3:25 PM
Page 76
CHAPTER 4 CORONARY HEART DISEASE
Table 4-9
Factors Related to the Incidence of Coronary Heart Disease Among Men Screened as Candidates for the Multiple Risk Factor Intervention Trial, 1992 Systolic Pressure (mm Hg)
Serum Total Cholesterol (mg/dl)
118
118–24
125–31
132–41
142
Q5/Q1
Nonsmokers 182 182–202 203–220 221–244 245 Q5/Q1
3.09 4.39 5.20 6.34 12.36 4.00
3.72 5.79 6.08 9.37 12.68 3.41
5.13 8.35 8.56 8.66 16.31 3.18
5.35 7.66 10.72 12.21 20.68 3.87
13.00 15.80 17.75 22.69 33.40 2.45
4.42 3.60 3.41 3.58 2.70 —
Smokers 182 182–202 203–220 221–244 245 Q5/Q1
10.37 10.03 14.90 19.83 25.24 2.43
10.69 11.76 16.09 22.69 30.50 2.85
13.21 19.05 21.07 23.61 35.26 2.67
13.99 20.67 28.87 31.98 41.47 2.96
21.04 33.69 42.91 55.50 62.11 2.30
2.61 3.36 2.88 2.80 2.46 —
Source: Reprinted with permission from J Stamler, Coronary Heart Disease Epidemiology: From Aetiology to Public Health, p 49, © 1992, by permission of Oxford University Press.
tors as found in the Pooling Project and other reports. In addition it provides data indicating a much greater relative risk between categories than is possible in studies that are on the order of one-hundredth the size of this one: The highest risk, that of smokers in the top quintile of both systolic blood pressure and serum cholesterol concentration (62.11 deaths per 10,000 person-years) is more than 20 times that of the lowest-risk stratum, the nonsmokers in the lowest quintile groups of both systolic blood pressure and serum cholesterol concentration. (This observation points to the relativity of relative risk: The lower the risk of the reference category, the greater the relative risk of the highest category. MRFIT is exceptional in defining 50 strata of risk.) Demonstration of the gradient of increasing risk beginning from the lowest levels of these characteristics is also important. Even among nonsmokers in the lowest quintile for systolic blood pressure, a marked gradient of increased risk is observed with increased cholesterol concentration; the same is true for those in the lowest quintile for cholesterol as systolic pressure increases. In the lowest quintile class for both of these factors, smoking alone increases the risk more than threefold, that is, from 3.09 to 10.37 per 10,000 person-years. The INTERHEART Study was an international collaboration involving 15,152 cases of acute myocardial infarction and 14,820 controls in 52 countries representing all inhabited continents.44 Yusuf and
colleagues reported on the association of nine risk factors—current smoking, diabetes, hypertension, abdominal obesity, psychosocial factors, fruit and vegetable consumption, exercise, alcohol, and ApoB/ ApoA1 ratio (a measure of blood lipid profile)—with the occurrence of acute MI. Figure 4-5 summarizes results for each risk factor, by sex, in the overall study population, indicating each odds ratio and population attributable risk (PAR) with its 99% confidence interval. Table 4-10 demonstrates, for men and women together, the results by region and overall. The four “lifestyle factors” and five “other risk factors” are distinguished, and a cumulative total PAR is shown first for the lifestyle factors and second for all factors together. The lifestyle factors—smoking, fruits and vegetables, exercise, and alcohol—accounted for the greater part of the PAR in nearly every region, from 47.6% in the Middle East to 69.9% in Southeast Asia and Japan. All nine risk factors together yielded an estimated PAR of 90.4%, ranging from 72.5% in central and eastern Europe to 98.7% in North America. Consistent with previous studies conducted mainly in Western industrialized countries, recognized risk factors for acute MI appear to account to a very large extent for the occurrence of this condition within diverse populations throughout the world. A more recent concept relates to the continuously graded relation between such factors as blood lipids and blood pressure and risk of cardiovascular dis-
46894_CH04_059_088.qxd
2/25/10
3:25 PM
Page 77
RISKS
Risk Factor Current Smoking
Diabetes
Hypertension
Abdominal Obesity Psychosocial Index Fruits/Veg
Exercise
Alcohol
ApoB/ApoA1 Ratio
Sex
Control (%) Case (%) Odds Ratio (99% CI)
77
PAR (99% CI)
F
9.3
20.1
2.86 (2.36 – 3.48)
15.8% (12.9 – 19.3)
M
33.0
53.1
3.05 (2.78 – 3.33)
44.0% (40.9 – 47.2)
F
7.9
25.5
4.26 (3.51 – 5.18)
19.1% (16.8 – 21.7)
M
7.4
16.2
2.67 (2.36 – 3.02)
10.1% (8.9 – 11.4)
F
28.3
53.0
2.95 (2.57 – 3.39)
35.8% (32.1 – 39.6)
M
19.7
34.6
2.32 (2.12 – 2.53)
19.5% (17.7 – 21.5)
F
33.3
45.6
2.26 (1.90 – 2.68)
35.9% (28.9 – 43.6)
M
33.3
46.5
2.24 (2.03 – 2.47)
32.1% (28.0 – 36.5)
F
–
–
3.49 (2.41 – 5.04)
40.0% (28.6 – 52.6)
M
–
–
2.58 (2.11 – 3.14)
25.3% (18.2 – 34.0)
F
50.3
39.4
0.58 (0.48 – 0.71)
17.8% (12.9 – 24.1)
M
39.6
34.7
0.74 (0.66 – 0.83)
10.3% (6.9 – 15.2)
F
16.5
9.3
0.48 (0.39 – 0.59)
37.3% (26.1 – 50.0)
M
20.3
15.8
0.77 (0.69 – 0.85)
22.9% (16.9 – 30.2)
F
11.2
6.3
0.41 (0.32 – 0.53)
46.9% (34.3 – 60.0)
M
29.1
29.6
0.88 (0.81 – 0.96)
10.5% (6.1 – 17.5)
F
14.1
27.0
4.42 (3.43 – 5.70)
52.1% (44.0 – 60.2)
M
21.9
35.5
3.76 (3.23 – 4.38)
53.8% (48.3 – 59.2) 0.25
0.5
1
2
4
8
16
Odds Ratio (99% CI)
Figure 4-5 Association of Risk Factors with Acute Myocardial Infarction in Men and Women After Adjustment for Age, Sex, and Geographic Region. Source: Reprinted with permission from The Lancet, Vol 364, p 944, © 2004.
eases. Rather than referring to the extreme category in the distribution of one or another factor, risk is becoming defined in relation to multiple factors, any or all of which may be only “borderline” and not “elevated” in value.45 On this basis, several factors may now be included in a “global risk assessment” to estimate the probability that an individual will experience a coronary event in some defined period, typically 10 years. From the Framingham Heart Study, Figure 4-6 shows the frequency distributions of less than 10%, 10–20%, and greater than 20% risk of “hard” coronary events in 10 years, separately for men and women, by age. These frequencies are derived from Framingham risk predictions and risk factor distributions in the US population determined in the Third National Health and Nutrition Examination Survey (NHANES III). On the basis of projected prevalence of optimal, borderline, and elevated values for five risk factors—blood pressure, serum LDL- and HDL-cholesterol concentrations, glucose tolerance, and smoking—less than 10% risk predominates for younger men and for every age level of women from 35–44 to 65–74 years. The upward shift in prevalence of higher risk is striking for men and notably less for women. For men older than age 45, risk of 10% or greater required at least one elevated risk
factor, with others at the borderline level, or two or more elevated factors. For women, 10% or greater risk occurred only at age 55 or older and required at least three elevated risk factors. Borderline risk factors alone were judged to account for only a small proportion of these events, which would occur within the next 10 years. Below “borderline” is “optimal” risk or, as discussed elsewhere in recent literature, “low risk.” Building on the data presented in Table 4-9 from the MRFIT screenees, Stamler and colleagues added the experience of nearly 40,000 employed persons who participated in the Chicago Heart Association Detection Project in Industry Study to determine the impact of “low risk” on CHD-CVD outcomes and allcause mortality.46 Low risk was defined in an early report on this work as serum cholesterol less than 200 mg/dl, systolic/diastolic blood pressure below 120/80 mm Hg without drugs, absence of cigarette smoking, no history of diabetes or myocardial infarction, and (in some but not all groups) absence of electrocardiographic abnormalities. Markedly reduced coronary, cardiovascular, and all-cause mortality was observed in low-risk men and women at all ages. Details of this topic and the concept of maintaining lifetime low risk are discussed in Chapter 21.
Region
7.3 4.8 18.3 18.0 11.2 11.1 6.6 19.8 12.9 13.7
45.5 38.9 37.4 35.9 36.2 44.8 38.3 26.1 36.4 35.7
27.6 25.6 25.5 12.2
4.2 10.1 27.1 20.3 31.4 23.8
38.4 11.3
67.6 49.6 47.6 63.4 56.6 62.3 69.9 66.0 56.6 59.9 62.9 54.6
1.0 26.6 5.5 5.7 27.9 18.6 3.7 25.5 13.9 6.7
All Lifestyles (%)
18.7 12.9
Alcohol (%)
32.7 19.0 23.4 17.9
9.2 29.6 19.3 22.1 38.4 22.6
21.9 24.5
12.7 8.0 12.3 9.9
15.5 16.7 11.8 10.0 21.0 7.2
15.0 9.1
Other Risk Factors Hypertension Diabetes (%) (%)
45.5 59.5 33.7 20.1
25.9 58.4 37.7 5.5 58.0 61.3
63.4 28.0
Abdominal Obesity (%)
35.6 51.4 28.8 32.5
41.6 40.0 15.9 35.4 26.7 28.9
38.9 4.9
All Psychosocial (%)
47.6 50.5 54.1 49.2
70.5 74.31 58.7 43.8 67.7 43.4
44.6 35.0
Lipids (%)
89.4 98.7 90.4* 90.4*
95.0 97.4 89.4 89.9 93.7 89.5
93.9 72.5
All Nine Risk Factors (%)
Source: Adapted from Yusuf S et al. Lancet Vol. 364, p. 945 © 2004.
PAR estimates in women in some countries are based on small numbers and so they are less reliable. Overall 1 adjusted for age, sex, and smoking; Overall 2 adjusted for risk factors. An extended version of this table with 99% Cls is shown in webtable 3 (http://image.thelancet.comm/extras/04art8001webtable3.pdf). *Saturated model, no difference between adjusted and unadjusted models. † Non-estimatable.
12.4 10.2
29.3 30.2
Exercise (%)
3:25 PM
Men and Women West Europe Central and eastern Europe Middle East Africa South Asia China Southeast Asia and Japan Australia and New Zealand South America North America Overall 1 Overall 2
Lifestyle Factors Smoking Fruits and (%) Vegetables (%)
Population Attributable Factors (PARs) Associated with Nine Risk Factors in Men and Women by Geographic Region
78
2/25/10
Table 4-10
46894_CH04_059_088.qxd Page 78
CHAPTER 4 CORONARY HEART DISEASE
46894_CH04_059_088.qxd
2/25/10
3:25 PM
Page 79
TRENDS AND EXPLANATIONS
10% 2 20%
99.6%
88.8%
,10%
.20%
10
93.5%
62.3%
12
75.2%
1.5%
23.3% 1.5%
5.0%
55–64 y
65–74 y
0.4%
45.2%
52.6% 26.8%
20.6%
2
19.0%
4
6.9%
6
35.8%
30.8%
8
11.2%
At-Risk Persons in Millions, n
14
100%
10-Year Absolute Event Rates
18 16
79
0 35–44 y
45–54 y
55–64 y
65–74 y
Men
35–44 y
45–54 y
Women
Figure 4-6 Estimated Numbers of US Individuals at Risk for Hard Coronary Heart Disease Events, According to Estimated 10-Year Absolute Risk. Source: Reprinted with permission from Annals of Internal Medicine, Vol 142, p 400, © 2005, American College of Physicians.
“Triggers” Factors that have immediate effects on risks of acute coronary events, usually within 1–2 hours before the onset of symptoms, are distinguished as “precipitating factors” (as in Figure 4-2) or “triggers.” Areas of interest noted in a recent review include behavioral and emotional factors such as anger, hostility, depression, physical exertion, and sexual activity; coffee and tea consumption; marijuana use; and exposure to particulate air pollution or environmental tobacco smoke.47 Study of these factors is made difficult by the time relation between exposure and effect. The “casecrossover” design permits characterizing exposures immediately prior to the event for cases and a matching time period for controls as well as a historical point of reference for each subject, such as six months or one year earlier. Differences in exposure histories for cases and controls permit calculation of an odds ratio as a measure of association between the potential trigger and events. Specific examples of these precipitating factors are discussed in subsequent chapters.
TRENDS AND EXPLANATIONS Marked changes in coronary mortality by country were shown in Figure 2-8. Understanding these changes in the United States and elsewhere would be aided
greatly by availability of data on incidence and casefatality, such as those collected by the WHO MONICA Project. However, in the United States no such data are available on a continuous basis nationally. This is due mainly to the special requirements noted above for ascertainment and standardized diagnostic validation of acute fatal and nonfatal coronary events in the conduct of both long-term surveillance and cohort studies. The few studies that have been carried out indicate the value of this approach. The United States At the time when the population studies described above were initiated, coronary mortality in the United States had been increasing every year for several decades. Unexpectedly, a decline in national coronary mortality began in the 1960s (although a sentinel report by Borhani and Hechter indicated a downturn in California in the late 1950s48). This major change was first doubted, then debated, and finally examined extensively in a conference of the National Heart, Lung and Blood Institute in 1978 (subsequently referred to as “the Decline Conference”).49 Two leading explanations were offered, either of which could in principle have reduced the death rate: first, that medical care, especially coronary intensive care units, had reduced inhospital mortality; second, that preventive efforts had resulted in more favorable risk-factor distributions and
46894_CH04_059_088.qxd
80
2/25/10
3:25 PM
Page 80
CHAPTER 4 CORONARY HEART DISEASE
thereby reduced the incidence of coronary events. Data were lacking to substantiate either explanation—an unwelcome admission of ignorance about the foremost cause of death in the United States. One response to the Decline Conference was an effort to reconstruct the history of the epidemic throughout the 20th century by compiling mortality data from published vital statistics. Stallones undertook this analysis, beginning from the premise that “whatever comes down must have gone up,” and sought a unified explanation of the rise and fall of the epidemic.50 The result is illustrated in Figure 4-7, in which each point represents the rate of death due to “diseases of the heart” in a given year from 1900 to 1978. This category is broader than “coronary heart disease” because this term was not in use early in the last century. Each year’s rate was adjusted to the age distribution of the US population in 1940, the midpoint of the period. This calculation removed any effect of increasing overall rates due to the upward shift in age composition of
the population. (This is in contrast to Figure 1-2, which shows the “crude” or absolute death rates for each category of disease each year, causing heart disease rates to appear to increase across the entire time period indicated.) The epidemic curve rises beginning in the 1920s, peaks at 1950, and declines through the 1970s. Stallones found no compelling unified explanation for both the rise and fall of the curve. However, the 20thcentury epidemic curve of coronary mortality in the United States was clearly demonstrated. Figure 2-3, from Morbidity & Mortality: 2007 Chart Book on Cardiovascular, Lung, and Blood Diseases, updates the course of the epidemic in the United States from 1979 to 2004.51 Age-adjusted coronary heart disease mortality for the United States declined by approximately 50%, from 350/100,000 to about 150/100,000. Numbers of deaths also declined, but by only about 20%. Efforts to explain the continuing decline in US coronary mortality have relied in part on analyses of
Figure 4-7 Heart Disease Mortality in the United States, 1900–1980. (The vertical scale is logarithmic.) Source: From The Rise and Fall of Ischemic Heart Disease, by RA Stallones, © 1980 by Scientific American Inc. All rights reserved.
46894_CH04_059_088.qxd
2/25/10
3:25 PM
Page 81
TRENDS AND EXPLANATIONS
trends in incidence in local communities, such as in Framingham, MA; Rochester, MN; Minneapolis/St. Paul, MN; and the four ARIC Study communities (Forsyth County, NC; Jackson, MS; suburbs of Minneapolis, MN; and Washington County, MD).52–55 The Framingham Heart Study reported that both sudden death and nonsudden death due to CHD had declined from the 1950s through the 1990s, as shown in Figure 4-8. Framingham data indicate that both first and recurrent-event-related deaths declined as a consequence of improvement in both primary and secondary prevention. The differential decrease in types of CHD death left sudden cardiac deaths equal in frequency to nonsudden deaths, or 50% of the total.52 In Rochester, declining incidence of coronary heart disease—despite increasing proportions of cases being identified through angiography—was attributed to more effective primary prevention.53 In Minneapolis/ St. Paul, incidence, case-fatality, and recurrent MI all decreased. Primary and secondary prevention were credited, as well as acute care during coronary attacks.54 The ARIC communities experienced declines overall in hospitalized recurrent MI, but the rate increased among Blacks; sudden deaths and post-MI survival both improved.55 Another approach was to determine outcomes in two cohorts of participants in the NHANES I and II surveys, one followed from 1971 to 1982 and the other from 1982 to 1992.56 From the first to the sec-
81
ond cohort, there were reductions in age-, sex-, and race-adjusted cardiovascular disease mortality (31%), incidence (21%), and 28-day case-fatality (28%). Both incidence and case-fatality contributed to the decline in mortality, leading again to the interpretation that primary and secondary prevention and treatment were responsible. Among other recently developed methods for estimating contributions of change in risk-factor distributions and in treatment practices, the IMPACT model developed by Capewell and colleagues has been applied to the US experience from 1980 to 2000 by Ford and others.57 The analysis incorporates available data on risk-factor trends in the population, efficacy and utilization of treatments for cases, and the observed difference in CHD mortality rates between the base and ending years. The difference between numbers of deaths actually observed in 2000 and expected deaths had the rates of 1980 still applied is then apportioned to the several contributing factors. Between ages 25 through 84 years, 341,745 fewer CHD deaths than expected occurred in 2000. Overall, 44% of the reduction appeared to be due to risk-factor change and 47% to improvement in use of effective treatments. The benefit from reduction in risk factors (24% for cholesterol, 20% for blood pressure, 12% for smoking, and 5% for physical activity) totaled 61% but was offset by an 18% negative effect of increased prevalence of body mass index (8%) and diabetes (10%).
600
Rate/100,000 Person-Years
CHD Death Nonsudden CHD Death
500
SCD
400 300 200 100 0 Referent
1970s
1980s
1990s
Trends in age- and gender-adjusted incidence rate per 100,000 person-years for overall CHD mortality, nonsudden CHD death, and SCD from 1950–1969 to 1990–1999. Bars represent upper and lower 95% Cls.
Figure 4-8 Trends in Age- and Gender-Adjusted Rate per 100,000 Person-Years for Overall CHD Mortality, Nonsudden CHD Death, and SCD from 1950–1969 to 1990–1999. Source: From Fox CS et al., Circulation, Vol 110, p 523, © 2004, American Heart Association, Inc.
46894_CH04_059_088.qxd
3:25 PM
Page 82
CHAPTER 4 CORONARY HEART DISEASE
Results from the IMPACT model in England and Wales, Scotland, Finland, and New Zealand and analyses by other methods and time periods in the United States, New Zealand, Holland, and Finland are remarkably consistent in showing 50% or more of the reduction in CHD mortality as due to risk-factor improvement. Finland indicated the greatest proportionate effect of risk-factor reduction, 76% during the early period from 1972 to 1992, reduced to 53% for the period from 1982 to 1997.57 This trend may reflect less frequent use of effective treatments in the 1970s in contrast to the latter period that includes the 1990s.
Three reports address specifically the contributions to the changing picture of coronary heart disease across MONICA Project populations to trends in coronary event rates and survival, classic risk factors, and coronary care.58–60 The observed trends for event rates and case-fatality are presented in Figure 4-9. MONICA CHD mortality declined by 2.7% in men and 2.1% in women during 371 “population-years” with 166,000 registered events. Overall event rates declined 2.1% in men and 0.8% in women; casefatality declined 0.6% and 0.8%, respectively. Main conclusions were: (1) the major determinant of the decline in CHD mortality was the change in event rates, although the declines in case-fatality were substantial; (2) the classic risk factors were mixed in direction of change—improving for blood pressure and cholesterol but opposite for body mass index—and, although a 4-year lag in the model improved the fit of risk factor and event rate trends, imprecision and ho-
Europe and Elsewhere The WHO MONICA Project aimed not only to monitor the components of change—event rates, incidence, case-fatality, and death rates—but also by identifying their major determinants to address the unanswered question from the Decline Conference.
5
0
0
25
25
210
210
FIN-NKA FIN-KUO SWE-NSW AUS-NEW NEZ-AUC ICE-ICE CAN-HAL UNK-BEL USA-STA FIN-TUL DEN-GLO SWE-GOT FRA-STR SWI-VAF GER-BRE GER-AUG BEL-GHE AUS-PER SWI-TIC ITA-BRI FRA-TOU UNK-GLA FRA-LIL RUS-NOC ITA-FRI GER-EGE CZE-CZE RUS-MOI BEL-CHA YUG-NOS POL-WAR RUS-NOC POL-TAR LTU-KAU SPA-CAT CHN-BEI RUS-NOI
5
Trend for Case Fatality in Men
RUS-MOC FRA-STR AUS-NEW FIN-NKA FIN-TUL FIN-KUO ICE-ICE SWE-GOT ITA-BRIA NEZ-AUC BEL-GHE RUS-MOI DEN-GLO USA-STA UNK-BEL SWE-NSW AUS-PER FRA-TOU FRA-LIL ITA-FRI CHN-BEI POL-TAR UNK-GLA CAN-HAL GER-BRE GER-AUG POL-WAR BEL-CHA RUS-NOI SPA-CAT CZE-CZE RUS-NOC GER-EGE LTU-KAU YUG-NOS
Trend for Coronary Events in Women 10
Trend for Case Fatality in Women
10
10
5
5
0
0
25
25
210
210
SWI-TIC FRA-TOU SWI-VAF SWE-NSW AUS-NEW ICE-ICE ITA-FRI BEL-CHA FRA-STR SPA-CAT USA-STA BEL-GHE AUS-PER CAN-HAL UNK-BEL UNK-GLA GER-BRE ITA-BRI NEZ-AUC FIN-NKA POL-WAR YUG-NOS FRA-LIL FIN-TUL RUS-NOC SWE-GOT CZE-CZE FIN-KUO LTU-KAU CHN-BEI POL-TAR GER-AUG RUS-NOC DEN-GLO GER-EGE RUS-MOC RUS-NOI
Average Annual Relative Trend (%)
Trend for Coronary Events in Men 10
ITA-BRI CAN-HAL FRA-TOU AUS-NEW GER-BRE AUS-PER FRA-STR GER-EGE UNK-GLA POL-WAR ITA-FRI FIN-TUL BEL-CHA BEL-GHE UNK-BEL LTU-KAU CZE-CZE ICE-ICE POL-TAR USA-STA GER-AUG FIN-NKA RUS-NOC SWE-NSW YUG-NOS NEZ-AUC FRA-LIL FIN-KUO SWE-GOT CHN-BEI RUS-MOC SPA-CAT RUS-MOI DEN-GLO RUS-NOI
82
2/25/10
Population
Figure 4-9 Population rankings, by sex, of trends in coronary-event rates and case fatality with 95% CIs. Source: Reprinted with permission from The Lancet, Vol. 353, p. 1554, copyright 1999 The Lancet.
46894_CH04_059_088.qxd
2/25/10
3:25 PM
Page 83
REFERENCES
mogeneity of trends across populations limited the explanatory power of the analyses; and (3) changes in coronary care and secondary prevention were closely associated with trends in event rates, casefatality, and CHD mortality, much more so than riskfactor changes. The authors speculate that treatment changes being greater in magnitude and more easily measured than risk-factor changes might explain the apparent dominance of treatment effects. On the basis of national vital statistics for 18 industrialized countries, declining coronary heart disease mortality was found for men and women in several Eastern European countries from 1999–2004, including the Czech Republic, Poland, Hungary, and Romania.51
FORECASTS Although trends in CHD mortality over the most recent two to three decades have been favorable in many Western countries, the actual burden as measured by numbers of events and survivors has diminished little if at all for cardiovascular diseases as a whole in the United States and perhaps elsewhere. Lower rates at all ages have been offset to a large degree by growing numbers of people attaining ages where the rates are still highest, despite their decrease. Coupled with increasing prevalence of obesity and diabetes, as in the United States, improvements in some risk-factor distributions are offset as well by rates that could increase once again. Global projections were discussed in Chapters 1 and 2 and are also unfavorable, especially in low- and middle-income countries. Worldwide projections indicate ischemic heart disease retaining the lead among all causes of death and taking the lead in years of life lost and disability-adjusted life years, by 2020. There is a wide gap between these forecasts and the present national and international efforts to monitor and favorably influence the course of change. Approaches to closing the gap will be discussed in Part IV.
CURRENT ISSUES Among many issues in the epidemiology and prevention of coronary heart disease, three are most central to its occurrence as a global public health problem: 1. Can the predicted increases in population rates of coronary heart disease occurrence— whether measured by mortality, incidence and case-fatality, prevalence of clinical or sub-
83
clinical disease, or extent of coronary atherosclerosis at autopsy of young decedents—be averted? 2. Can population differences in coronary heart disease occurrence and epidemic occurrence of coronary heart disease be eliminated, so that all populations share the most favorable experience? 3. Can coronary heart disease surveillance be maintained and expanded to monitor future changes in disease rates and the impact of preventive strategies effectively? REFERENCES 1. Hutter AM, IX. Ischemic heart disease: angina pectoris. In: Rubenstein E, Federman DD, eds. Scientific American Medicine. New York: Scientific American Inc.; 1995:1–19. 2. Fuster V, Fallon JT, Nemerson Y. Coronary thrombosis. Lancet. 1996;348:S7–S10. 3. Antman EM, Braunwald E. Acute myocardial infarction. In: Braunwald E, ed. Heart Disease. A Textbook of Cardiovascular Medicine. 5th ed. Philadelphia, PA: WB Saunders Co; 1997: 1184–1288. 4. Myerburg RJ, Castellanos A. Cardiac arrest and sudden death. In: Braunwald E, ed. Heart Disease: A Textbook of Cardiovascular Medicine. 5th ed. Philadelphia, PA: WB Saunders Co; 1997:742–779. 5. Willich SN, Muller JE. Triggering of Acute Coronary Syndromes: Implications for Prevention. Dutrecht (The Netherlands): Kluwer Academic Publishers; 1996. 6. Report of a WHO Scientific Group. Sudden Cardiac Death. Technical Report Series 726. Geneva (Switzerland): World Health Organization; 1985. 7. Fuster V, Moreno PR, Fayad ZA, Corti R, Badimon JJ. Atherothrombosis and high-risk plaque: part I: evolving concepts. J Am Coll Cardiol. 2005;46(6):937–954. 8. Fuster V, Fayad ZA, Moreno PR, Poon M, Corti R, Badimon JJ. Atherothrombosis and high-risk plaque: part II: approaches by noninvasive computed tomographic/magnetic
46894_CH04_059_088.qxd
84
2/25/10
3:25 PM
Page 84
CHAPTER 4 CORONARY HEART DISEASE
resonance imaging. J Am Coll Cardiol. 2005; 46(7):1209–1218. 9. Liebowitz JO. The History of Coronary Heart Disease. Berkeley, CA: University of California Press; 1970. 10. Frye WB. William Osler’s Collected Papers on the Cardiovascular System. Birmingham, AL: The Classics of Cardiology Library; 1985. 11. Keys A. From Naples to Seven Countries—a sentimental journey. Prog Biochem Pharmacol. 1983;19:1–30. 12. Keys A. Seven Countries: A Multivariate Analysis of Death and Coronary Heart Disease. Cambridge, MA: Harvard University Press; 1980. 13. Pooling Project Research Group. Relationship of blood pressure, serum cholesterol, smoking habit, relative weight and ECG abnormalities to incidence of major coronary events: Final Report of the Pooling Project. J Chronic Dis. 1978;31:201–306. 14. Gordon T. Mortality experience among the Japanese in the United States, Hawaii, and Japan. Public Health Rep. 1957;72:543–553. 15. Marmot M, Syme SL, Kagan A, et al. Epidemiologic studies of coronary heart disease and stroke in Japanese men living in Japan, Hawaii and California: prevalence of coronary and hypertensive heart disease and associated risk factors. Am J Epidemiol. 1975;102:514–525. 16. Dawber TR, Meadors GF, Moore FE Jr. Epidemiological approaches to heart disease: the Framingham Study. Am J Public Health. 1951;41:279–286. 17. Levy D, Brink S. A Change of Heart. Unraveling the Mysteries of Cardiovascular Disease. New York: Vintage Books; 2005. 18. Blackburn H. On the Trail of Heart Attacks in Seven Countries. Middleborough, MA: The Country Press, Inc; 1995. 19. Kromhout D, Menotti A, Blackburn H. The Seven Countries Study: A Scientific Adventure in Cardiovascular Disease Epidemiology.
Utrecht (The Netherlands): Brouwer Offset bv; 1993. 20. Toshima H, Koga Y, Blackburn H, eds. Keys A, honorary ed. Lessons for Science from the Seven Countries Study. Tokyo (Japan): Springer; 1994. 21. Rose GA, Blackburn H. Cardiovascular Survey Methods. Geneva (Switzerland): World Health Organization; 1968. 22. Gillum RF, Fortmann SP, Prineas RJ, Kottke TE. International diagnostic criteria for acute myocardial infarction and stroke. Am Heart J. 1984;108:150–158. 23. World Health Organization MONICA Project Principal Investigators. The World Health Organization MONICA Project (Monitoring Trends and Determinants in Cardiovascular Disease): a major international collaboration. J Clin Epidemiol. 1988;41:105–114. 24. Tunstall-Pedoe H, Kuulasmaa K, Amouyel P, Arveiler D, Rajakangas AM, Pajak A. Myocardial infarction and coronary deaths in the World Health Organization MONICA Project. Registration procedures, event rates, and case-fatality rates in 38 populations from 21 countries in four continents. Circulation. 1994;90:583–612. 25. Luepker RV, Evans A, McKeigue P, Reddy KS. Cardiovascular Survey Methods. Geneva: World Health Organization; 2004. 26. Luepker RV, Apple FS, Christenson RH, et al. Case definitions for acute coronary heart disease in epidemiology and clinical research studies. A statement from the AHA Council on Epidemiology and Prevention; AHA Statistics Committee; World Heart Federation Council on Epidemiology and Prevention; the European Society of Cardiology Working Group on Epidemiology and Prevention; Centers for Disease Control and Prevention; and the National Heart, Lung and Blood Institute. Circulation. 2003;108:2543–2549. 27. Crow RS, Hannan PJ, Jacobs DR Jr, Lee SM, Blackburn H, Luepker RV. Eliminating diagnostic drift in the validation of acute inhospital myocardial infarction—implication
46894_CH04_059_088.qxd
2/25/10
3:25 PM
Page 85
REFERENCES
for documenting trends across 25 years: the Minnesota Heart Survey. Am J Epidemiol. 2005;161(4):377–388. 28. Lloyd-Jones D, Adams R, Carnethon M, et al. Heart disease and stroke statistics—2009 update. A report from the American Heart Association Statistics Committee and Stroke Statistics Subcommittee. Circulation. 2009;119:e1–e161. 29. Antman EM, Anbe DT, Armstrong PW, et al. ACC/AHA guidelines for the management of patients with ST-elevation myocardial infarction––executive summary: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to Revise the 1999 Guidelines for the Management of Patients With Acute Myocardial Infarction). Circulation. 2004;110(5):588–636. 30. World Health Organization. International Statistical Classification of Diseases and Related Health Problems. 10th rev. Geneva (Switzerland): World Health Organization; 1992; 1. 31. Murray CJL, Lopez AD. Estimating causes of death: new methods and global and regional applications for 1990. In: Murray CJL, Lopez AD, eds. The Global Burden of Disease: A Comprehensive Assessment of Mortality and Disability from Diseases, Injuries, and Risk Factors in 1990 and Projected to 2020. Boston, MA: The Harvard School of Public Health; 1996. 32. National Center for Health Statistics. Health, United States, 2006. DHHS Publication No. 2006-1232. Hyattsville, MD: Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Health Statistics; 2006. 33. US Department of Health and Human Services. Healthy People 2010. 2nd ed. With Understanding and Improving Health and Objectives for Improving Health. 2 vols. Washington, DC: US Government Printing Office; November 2000. 34. World Health Organization. World Health Report 2004. Geneva (Switzerland): World Health Organization; 2004.
85
35. Myerburg RJ, Kessler KM, Castellanos A. Sudden cardiac death: epidemiology, transient risk, and intervention assessment. Ann Intern Med. 1993;119(12):1187–1197. 36. National Heart, Lung and Blood Institute. Incidence & Prevalence: 2006 Chart Book on Cardiovascular and Lung Diseases. Washington, DC: US Department of Health and Human Services, Public Health Service, National Institutes of Health; 2006. 37. Li N, Tuomilehto J, Dowse G, Virtala E, et al. Prevalence of coronary heart disease indicated by electrocardiogram abnormalities and risk factors in developing countries. J Clin Epidemiol. 1994;47:599–611. 38. Xie J, Wu EQ, Zheng ZJ, Sullivan PW, Zhan L, Labarthe DR. Patient-reported health status in coronary heart disease in the United States. Age, sex, racial, and ethnic differences. Circulation. 2008;118(5):491–497. 39. Murray CJL, Lopez AD. Alternative visions of the future: projecting mortality and disability, 1990–2020. In: Murray CJL, Lopez AD, eds. The Global Burden of Disease: A Comprehensive Assessment of Mortality and Disability from Diseases, Injuries, and Risk Factors in 1990 and Projected to 2020. Boston, MA: The Harvard School of Public Health; 1996. 40. Leeder S, Raymond S, Greenberg H. A Race Against Time: The Challenge of Cardiovascular Disease in Developing Countries. New York: The Trustees of Columbia University in the City of New York; 2004. 41. Murray CJL, Kulkarni SC, Michaud C, et al. Eight Americas: investigating mortality disparities across races, counties, and racecounties in the United States. PLoS Med. 2006;3:1513–1524. 42. Cooper RS. Coronary heart disease burden among persons of African origin. In: Marmot M, Elliott P, eds. Coronary Heart Disease Epidemiology: From Aetiology to Public Health. 2nd ed. Oxford (England): Oxford University Press; 2005:118–132. 43. Stamler J. Established major coronary risk factors. In: Marmot M, Elliott P, eds. Coronary
46894_CH04_059_088.qxd
86
2/25/10
3:25 PM
Page 86
CHAPTER 4 CORONARY HEART DISEASE
Heart Disease Epidemiology: From Aetiology to Public Health. Oxford (England): Oxford University Press; 1992:35–66. 44. Yusuf S, Hawken S, Ôunpuu S, et al. Effect of potentially modifiable risk factors associated with myocardial infarction in 52 countries (the INTERHEART study): case-control study. Lancet. 2004;364:937–952. 45. Vasan RS, Sullivan LM, Wilson PWF, et al. Relative importance of borderline and elevated levels of coronary heart disease risk factors. Ann Intern Med. 2005;142:393–402. 46. Stamler J, Stamler R, Neaton JD, et al. Low risk-factor profile and long-term cardiovascular and noncardiovascular mortality and life expectancy. Findings for 5 large cohorts of young adult and middle-aged men and women. JAMA. 1999;282:2012–2018. 47. Siscovick DS. Triggers of clinical coronary heart disease. Epidemiology. 2006;17(5): 495–497. 48. Borhani NO, Hechter HH. Recent changes in CVR disease mortality in California. Pub Health Rep. 1964;79:147–160. 49. Havlik RJ, Feinleib M, eds. Proceedings of the Conference on the Decline in Coronary Heart Disease Mortality. NIH publication 79-1610. Bethesda, MD: National Heart, Lung and Blood Institute, National Institutes of Health; 1978. 50. Stallones, RA. The rise and fall of ischemic heart disease. Scientific American. 1980;243: 53–59. 51. National Heart, Lung and Blood Institute. Morbidity & Mortality: 2007 Chartbook on Cardiovascular, Lung, and Blood Diseases. Bethesda MD: US Department of Health and Human Services. Public Health Service, National Institutes of Health; June 2007. 52. Fox CS, Evans JC, Larson MG, Kannel WB, Levy D. Temporal trends in coronary heart disease mortality and sudden cardiac death from 1950 to 1999. The Framingham Heart Study. Circulation. 2004;110:522–527.
53. Arciero TJ, Jacobsen SJ, Reeder GS, et al. Temporal trends in the incidence of coronary disease. Am J Med. 2002;117:228–233. 54. McGovern PG, Jacobs DR Jr, Shahar E, et al. Trends in acute coronary heart disease mortality, morbidity, and medical care from 1985 through 1997: the Minnesota heart survey. Circulation. 2001;104(1):19–24. 55. Rosamond WD, Folson AR, Chambless LE, Wang C-H for the ARIC Investigators. Coronary heart disease trends in four United States communities. The Atherosclerosis Risk in Communities (ARIC) Study 1987–1996. Int J Epid. 2001;30:S17–S22. 56. Ergin A, Muntner P, Sherwin R, He J. Secular trends in cardiovascular disease mortality, incidence, and case fatality rates in adults in the United States. Am J Med. 2004;117:219–227. 57. Ford ES, Ajani UA, Croft JB, et al. Explaining the decrease in U.S. deaths from coronary disease, 1980–2000. N Engl J Med. 2007;356: 33–43. 58. Tunstall-Pedoe H, Kuulasmaa K, Mahonen M, Tolonen H, Ruokokoski E, Amouyel P. Contribution of trends in survival and coronaryevent rates to changes in coronary heart disease mortality: 10-year results from 37 WHO MONICA project populations. Monitoring trends and determinants in cardiovascular disease. Lancet. 1999;353(9164): 1547–1557. 59. Kuulasmaa K, Tunstall-Pedoe H, Dobson A, et al. Estimation of contribution of changes in classic risk factors to trends in coronary-event rates across the WHO MONICA Project populations. Lancet. 2000;355(9205):675–687. 60. Tunstall-Pedoe H, Vanuzzo D, Hobbs M, et al. Estimation of contribution of changes in coronary care to improving survival, event rates, and coronary heart disease mortality across the WHO MONICA Project populations. Lancet. 2000;355(9205):688–700.
46894_CH04_059_088.qxd
2/25/10
3:25 PM
Page 87
APPENDIX 4-A
87
APPENDIX 4-A Key to Population Abbreviations Used by the WHO MONICA Project Population Abbreviation AUS-NEW AUS-PER BEL-CHA BEL-GHE BEL-LUX CAN-HAL CHN-BEI CZE-CZE DEN-GLO FIN-KUO FIN-NKA FIN-TUL FRA-LIL FRA-STR FRA-TOU GER-AUR
Population
Australia
Newcastle Perth Charleroi Ghent Luxembourg Halifax County Beijing Czech Republica Glostrup Kuopio Province North Karelia Turku/Loimaa Lille Strasbourg Toulouse Augsburg Rural
SPA-CAT SWE-GOT SWE-NSW
Spain Sweden
Augsburg Urban Bremen East Germanyb Rhein-Neckar Regionc Budapest
SWI-TIC SWI-VAF UNK-BEL UNK-GLA USA-STA YUG-NOS
Switzerland
Belgium
Canada China Czech Republic Denmark Finland
France
Germany
GER-AUU GER-BRE GER-EGE GER-RHN HUN-BUD
Population Abbreviation
Country
Hungary
HUN-PEC ICE-ICE ITA-BRI ITA-FRI LTU-KAU NEZ-AUC POL-TAR POL-WAR RUS-MOC
Country Iceland Italy Lithuania New Zealand Poland
Russia
RUS-MOI RUS-NOC RUS-NOI
UK USA Yugoslavia
a
Disagreement between local and national authorities on numbers of coronary deaths. Data for this center were incomplete as of 1994. c This register no longer exists. Data queries were answered up to January 1992. b
Source: Reprinted with permission from Circulation, Special Report, Vol 90, No 1, p 599, © 1994, American Heart Association.
Population Pecs Iceland Area Brianza Friuli Kaunas Auckland Tarnobrzeg Volvodship Warsaw Moscow Control Moscow Intervention Novosibirsk Control Novosibirsk Intervention Catalonia Gothenburg Northern Sweden Ticino Vaud/Fribourg Belfast Glasgow Stanford Novi-Sad
46894_CH04_059_088.qxd
2/25/10
3:25 PM
Page 88
46894_CH05_089_110.qxd
2/25/10
3:25 PM
Page 89
C H A P T E R
5 Stroke pressure. However, population differences in the effectiveness of public health interventions for high blood pressure control cannot wholly explain population differences in trends of stroke mortality. Considerations of the burden and disparities associated with stroke pose public health challenges for the United States that are shared with many other countries. There is a need for broad international collaboration in meeting this challenge if the greatest potential impact of stroke prevention is to be achieved for many diverse populations. Common approaches to population surveillance of stroke are advocated on a multinational basis to strengthen the potential for prevention through effective national policies and practices.
SUMMARY Stroke, or cerebrovascular accident (CVA)—also now termed “brain attack” by analogy to “heart attack”— is a second major class of “end-organ” outcomes of atherosclerotic and hypertensive diseases. Just as the heart is damaged by disturbance of flow in the coronary circulation, the brain is damaged by disturbance of flow in the cerebral circulation. From transient episodes of less than 24 hours to permanent brain dysfunction, disability, or death, stroke has a wide range of clinical expressions. It constitutes a large proportion of overall cardiovascular morbidity and mortality globally. In some regions, stroke predominates substantially over coronary heart disease in its frequency. In the United States, it is about one-fourth to one-third as common as coronary heart disease as measured by both death rate and prevalence. Stroke mortality ranges widely among different populations and has been observed to change significantly within only a few years. Stroke mortality has been decreasing in many countries in recent years, although several eastern European countries have experienced sharp increases over the same period. Environmental factors have unquestionably played a major role in the long-term trends, given that they antedated by many years any widespread effective treatments. Among stroke survivors, disability and dependency are common, as is risk of recurrent strokes. In the United States, racial differences in stroke deaths persist, with relative excesses over non-Hispanic Whites long recognized in Blacks and more recently among American Indians/Alaska Natives and Native Hawaiians/Pacific Islanders. Although particular factors contribute to different types of stroke, the most prominent controllable factor common to all types is high blood
INTRODUCTION The Cerebral Arteries Figure 5-1 illustrates the anatomical relations of the main intracranial arteries to the cerebral hemispheres.1 Like the coronary arteries for the heart, these vessels are the principal suppliers of blood to the brain. However, analogy to the coronary circulation is limited by several factors. First, intermediate vessels— the external carotid arteries on the right and left sides of the neck extend into the cranium as internal carotid arteries to feed the vessels pictured here. Their potential involvement with atherosclerosis poses risks of transient disturbance of circulation to the brain independent of the intracranial vessels themselves. Second, these arteries can convey to the brain small blood clots, or thrombi, that are formed within the heart and travel through vessels of diminishing diameter until becoming lodged in an artery and blocking circulation beyond. (A thrombus sent from one site
89
46894_CH05_089_110.qxd
90
2/25/10
3:25 PM
Page 90
CHAPTER 5 STROKE
Parietal Lobe LATERAL SURFACE Posterior Cerebral Artery
Anterior Cerebral Artery
Frontal Lobe
Middle Cerebral Artery
Occipital Lobe
Temporal Lobe
MEDIAL SURFACE Posterior Cerebral Artery Anterior Cerebral Artery
Figure 5-1 A Schematic View of the Posterior, Middle, and Anterior Cerebral Arteries, from the Outer (Lateral) and Inner (Medial) Aspects of the Right Hemisphere of the Brain. Source: Albers GP, Cutler RWP, “Cerebrovascular Diseases” from Scientific American Medicine, Dale DC, Federman DD, Eds., © 119. Scientific American, Inc. All rights reserved.
to another in this way is termed an “embolus.”) Third, there are smaller vertebrobasilar arteries reaching the posterior part of the brain whose obstruction can also cause disturbance of brain function. Fourth, another type of process disrupting brain circulation results from a diseased and dilated segment of an artery, an aneurysm, that may rupture causing hemorrhage into the brain. Fifth, either spontaneous rupture or traumatic head injury can produce hemorrhage beneath the surface lining of the brain, the arachnoid membrane, producing a subarachnoid hemorrhage. Finally, a thrombus may form within the venous circulation of the brain, thereby disturbing normal blood flow. These multiple pathways and processes present numerous possibilities for interruption of brain circulation. Compounded by the different effects depending on the area of the brain affected, a very wide range of manifestations can occur. “Stroke” can therefore mean many things. Clinical Course of the Individual Case The hallmark of a typical severe stroke is its abrupt onset, with sudden and dramatic loss of consciousness and motor and sensory function on one side of the
body. Warning signs of a stroke are described as sudden numbness or weakness of the face, arm, or leg; sudden confusion, trouble speaking, or trouble understanding; sudden trouble walking, dizziness, or loss of balance or coordination; sudden trouble seeing in one or both eyes; and sudden severe headache with no known cause. These are the chief clinical manifestations of acute interruption of arterial blood supply to one or more areas of the brain. In Western countries, the interruption is due most often to obstruction of a major artery by thrombosis (formation of a blood clot locally) and less often to hemorrhage or embolism. Figure 5-2 illustrates the time course and potential outcomes of the typical acute cerebrovascular event. Stroke occurs most often against a background of advanced atherosclerotic lesions in the cerebral arteries or longstanding high blood pressure, or both. Although it is plausible that precipitating factors may trigger plaque disruption and its consequences just as in the coronary arteries (see Chapter 4), this process has received less attention in connection with stroke. Whether due to occlusion or hemorrhage, suddenness of onset is a defining characteristic of stroke. Signs and symptoms may diminish and disappear in
46894_CH05_089_110.qxd
2/25/10
3:25 PM
Page 91
BACKGROUND
91
Figure 5-2 Common Features of the Course of Stroke.
minutes or hours without residual clinical abnormalities. Alternatively, they may persist and progress, ending in death or permanent disability. An episode that resolves completely within 24 hours (an arbitrary but widely adopted criterion) is designated a “transient ischemic attack” (TIA). One that persists longer than 24 hours is termed a “completed stroke.” A completed stroke that is followed by death within 28 days from the onset of the episode is called a “fatal stroke” and is included in enumeration of case-fatality. Many variations can occur in the time course, location, and clinical features of acute disturbances in cerebral circulation, for the reasons outlined above. In addition, cumulative effects of multiple unnoticed ischemic events, blocking blood flow to small areas of the brain, can culminate over a period of years in a condition described as “multi-infarct dementia,” or impairment of cognitive function. Among these many potential manifestations, fatal or nonfatal completed strokes are the most widely recognized form of stroke and the main public health concern to date. These events currently constitute the third most frequent cause of death in the United States, with agestandardized mortality of 46.6 per 100,000 per year, just under one-third the rate for ischemic heart disease (data for the total population, 2005).2
BACKGROUND Only in the late 1950s and early 1960s were geographic comparisons of stroke mortality first reported, and Stallones remarked in a 1965 review that little epidemiologic study had been undertaken of stroke, unlike ischemic heart disease and hypertension.3 He
noted several methodologic problems: In death certificate studies, it was difficult or impossible to distinguish between major types of stroke—mainly thrombotic or embolic occlusion, intracerebral hemorrhagic, and subarachnoid hemorrhage. There was believed to be considerable variation in death certification practices in different countries. In early studies of incidence or prevalence of stroke, study populations were often too young to generate sufficient numbers of cases for reliable estimation of rates. Despite these limitations, it was possible to discern in international comparative studies, based on death certificates, a threefold or greater range in mortality (from more than 150 to less than 50 per 100,000 population per year) from all classes of vascular lesions of the central nervous system. Finland and Japan had the highest rates. Especially noteworthy was the pioneering work of Gordon, who described mortality data among three groups of men of Japanese ancestry situated in Japan, Hawaii, or California. He found a sharp gradient of decreasing stroke mortality from Japan to Hawaii to California among these men of common genetic background who, through differences in migration histories, had come to live—and die—in distinct environments. The pattern of mortality from stroke was opposite that for coronary heart disease, which increased from Japan to Hawaii and California. This observation of diverging “natural histories” was to become the basis for a three-part collaborative study. On the presumption that atherosclerosis was a common underlying condition for both stroke and ischemic heart disease, Stallones expected to find a correlation in mortality rates for these two causes when rates for the two diseases were compared among the
46894_CH05_089_110.qxd
92
2/25/10
3:25 PM
Page 92
CHAPTER 5 STROKE
United States.3 However, there was no relation between the distributions of these causes of death among the states. Different factors seemed to influence the occurrence of stroke and ischemic heart disease within this country. A concentration of high stroke mortality rates in the southeastern states was noted, a pattern first termed “the stroke belt” by Borhani (personal communication, George Howard, 1997). A striking secular trend of decreasing stroke mortality in the United States from 1900 onward was also demonstrated. Little information was available about possible risk factors for stroke other than age and sex. A steep age gradient in risk was present. For Whites in the United States, rates were higher for males than for females, whereas for non-Whites, females had the higher rates. Gorelick and Alter presented an assessment of stroke and its epidemiology and prevention, The Prevention of Stroke, in 2002.4 Distribution, risk factors, and approaches to prevention including recurrent events are addressed by multiple contributors. Although important questions remain, substantial progress has been made in the decades since Stallones’s early review.
POPULATION STUDIES: DEFINITION, CLASSIFICATION, AND DIAGNOSTIC METHODS As was the case for coronary heart disease, definitions and criteria for epidemiologic studies of stroke were developed by the Criteria and Methods Committee of the Council on Epidemiology and Prevention of the American Heart Association.5 The methodologic report of the World Health Organization (WHO) MONICA Project, in which community surveillance for stroke was an optional addition to the central focus on coronary heart disease, demonstrates the resulting classification of strokes (Table 5-1).6 Surveillance methods require uniform criteria for case ascertainment and validation. These are described in the situation of the WHO MONICA Project as follows: Stroke was defined as rapidly developing signs of focal (or global) disturbance of cerebral function lasting more than 24 hours (unless interrupted by surgery or death), with no apparent nonvascular cause; the definition included patients presenting with clinical signs and symptoms suggestive of subarachnoid hemorrhage, intracerebral hemorrhage, or cerebral infarction. . . . Events were characterized as either definite stroke, not stroke, or unclassifiable. Criteria for definite stroke were fulfilled when the available
Table 5-1
A Diagnostic Classification for Fatal and Nonfatal Strokes in Population Studies
Diagnostic categories (classification based on the most severe findings obtained within 28 days of onset): • Definite stroke • Not stroke • Definite stroke associated with definite myocardial infarction • Insufficient data Subcategories of stroke (classification based on confirmatory findings from necropsy in fatal cases or computerized axial tomography [CT] scan in nonfatal cases; specific criteria are given for each subcategory): • • • • •
Subarachnoid hemorrhage Intracerebral hemorrhage Brain infarction due to occlusion of precerebral arteries Brain infarction due to cerebral thrombosis Embolic brain infarction
Source: Data from K Asplund, et al., Diagnostic Criteria and Quality Control of the Registration of Stroke Events in the MONICA Project, Acta Med Scand Suppl 728, pp 26–39, © 1988.
information permitted a clinical stroke diagnosis. Unclassifiable was used when no diagnosis other than stroke was present to explain the event but the available information was insufficient for determining whether symptoms and duration fully met the MONICA criteria for definite stroke. . . . Stroke events were subdivided into first or recurrent and into fatal or nonfatal. A period of 28 days was used to define case fatality and to distinguish one event from another.7, pp 500–501 The types of stroke are correspondingly distinguished in the classification system of the International Classification of Diseases and Related Health Problems (ICD), as summarized in Table 5-2.8 TIAs are not included with these classes of stroke in ICD 10. Because their inclusion or exclusion affects reported rates of both nonfatal and total stroke (being excluded from fatal strokes by definition), proper comparison among studies requires knowledge of the inclusion criteria used. Epidemiologic study of stroke has been hampered by the infrequent identification of stroke subtypes, especially lack of differentiation between occlusive and hemorrhagic strokes. Advent and wider use of diagnostic brain imaging procedures (e.g., computed tomography, or CT scans) offers a potential solution to this problem, and use of magnetic-resonance imaging, or MRI, appears to improve identification of stroke type and assessment of the extent of tissue damage significantly over the CT scan.9 But many fatal stroke
46894_CH05_089_110.qxd
2/25/10
3:25 PM
Page 93
RATES
Table 5-2 160 161 162 163 164 165 166 167 168 169
Categories of Cerebrovascular Diseases in ICD 10 (160–169)
Subarachnoid hemorrhage Intracerebral hemorrhage Other nontraumatic intracranial hemorrhage Cerebral infarction Stroke, not specified as hemorrhage or infarction Occlusion and stenosis of precerebral arteries, not resulting in cerebral infarction Occlusion and stenosis of cerebral arteries, not resulting in cerebral infarction Other cerebrovascular diseases Cerebrovascular disorders in diseases classified elsewhere Sequelae of cerebrovascular disease
Note: Transient ischemic attacks are not included.
cases are out-of-hospital deaths even in the highincome countries, and on a global basis a great many cases occur where no such technology is available. As a result, lack of data on specific stroke types is a continuing limitation of minor to major degree depending on the setting in which stroke data are collected. Recent advances in treatment of occlusive strokes with thrombolytic therapy have placed new emphasis on quality of acute stroke care, including differentiation among stroke types in the first minutes to hours after onset of a stroke. This is because use of thrombolytic therapy to reduce morbidity and mortality requires identifying those cases of occlusive stroke that are eligible for this treatment, which is contraindicated in hemorrhagic stroke. For identifying barriers to timely access to high-quality care, a hospital-based stroke surveillance program has been established on a limited scale in the United States called the Paul Coverdell National Acute Stroke Registry, operated by several state health departments in cooperation with the Centers for Disease Control and Prevention (CDC). Through collaboration among interested national nongovernmental organizations and other federal agencies, agreement was reached on definition of data elements in 12 core areas. Implementation of this registry and parallel work in Canada, Germany, Japan, and Korea has been reported recently and indicates current issues for hospitalbased surveillance of stroke in various settings.10 A further difficulty is to know how death in a stroke victim is classified, whether as caused by stroke or distinct from the stroke itself. Study of a sample of 200 cases of stroke or TIA in the Brain Attack Surveillance in Corpus Christi (TX) Project indicated poor agreement between raters on this question.11 Agreement between two neurologists as to whether a
93
death would be attributed to stroke was only between 40% and 50%. This finding implied that stroke mortality data may be based on methods of uncertain unreliability and should be interpreted with caution.
RATES In the United States, 143,579 deaths were attributed to stroke as the underlying cause in 2005, while WHO in 2004 estimated more than 5 million stroke deaths from this second leading cause of death worldwide.12,13 The United States’ national target for reduced stroke death rates by 2010 is 50/100,000, a 20% improvement from the baseline rate of 62/100,000 in 1999. By 2002, half of the projected gain for the decade had already occurred for the population as a whole.14 Mortality The United States Numbers of deaths due to stroke by age, sex, and race for non-Hispanic Whites and Blacks in the United States for 2005 are shown with other population-level indicators in Table 5-3, from data reported by the National Center for Health Statistics and other sources.12 For both Whites and Blacks, numbers of stroke deaths among females far exceed those for males. This pattern by sex occurs despite higher age-specific stroke death rates for males (shown elsewhere) because greater numbers of females reach the ages where rates are highest. Figure 5-3 illustrates the steep age gradient for stroke for these four sex-race groups.15 Like the corresponding patterns for coronary heart disease (Figure 2-1), stroke death rates are higher for Blacks
Table 5-3
Stroke Mortality, United States, 2005 Stroke Deaths Total 143,579 All males 56,586 % of total 39.4 All females 86,993 % of total 60.6 Non-Hispanic white males 47,194 % of subgroup 35.2 Non-Hispanic white females 74,674 % of subgroup 64.8 Non-Hispanic black males 7,519 % of subgroup 42.9 Non-Hispanic black females 10,022 % of subgroup 57.1 Source: Data from Heart Disease and Stroke Statistics—2009 Update. A Report from the American Heart Association Statistics Committee and Stroke Statistics Committee. D Lloyd-Jones et al., © 2009, Courtesy of the American Heart Association/American Stroke Association.
46894_CH05_089_110.qxd
94
2/25/10
3:25 PM
Page 94
CHAPTER 5 STROKE
700
Deaths/100,000 Population
600
Black Male White Male Black Female White Female
500 400 300 200 100 0 35–44
45–54
55–64
65–74
75–84
Age (Years)
Figure 5-3 Death Rates for Stroke by Age, Race, and Sex, US, 2004. Source: From Morbidity & Mortality: 2007 Chart Book on Cardiovascular, Lung and Blood Diseases. National Institutes of Health, National Heart, Lung and Blood Institute, June 2007.
than Whites at every age. In contrast to the picture for coronary heart disease, the difference in age-specific rates between males and females is narrower for Blacks and nearly absent for Whites. Stroke mortality does not exhibit the apparent advantage for women found for coronary heart disease. Nearly one-third of stroke deaths in the United States occur at younger than age 75, and slightly more than two-thirds occurred at ages 75 and older. For this reason there is special concern about stroke among the older population. Analogous to the atlases of county-level coronary heart disease mortality by sex and race/ethnicity for the United States illustrated in Figure 2-6, data on stroke mortality by race/ethnicity and geographic area are presented in two publications and on an interactive Web site (accessible at http://www.cdc.gov/chv/ maps).16,17 The underlying data are for United States counties for the period 1991–1998 and demonstrate continuing appearance of the “stroke belt,” the concentration of high stroke death rates in southeastern states. Additional areas of concentration include the Mississippi River Valley and, more prominently for women than men, Washington, Oregon, and Idaho as well as northern counties of California. Special importance is attached to the Atlas of Heart Disease and Stroke Among American Indians and Alaska Natives, a heterogeneous population group for whom the exceptionally high burden of heart disease and stroke has been largely unrecognized. This publication includes mapping of risk-factor information for this population, as obtained through the Behavioral Risk Factor Surveillance System of the Centers for Disease Control and Prevention on a state-by-state basis.
Location of death from stroke—whether out of hospital, in the hospital emergency department, or after hospital admission—is of interest especially from the viewpoint of access to emergency transport and medical services immediately after onset of warning signs for a stroke. On the basis of data from 1999 for the United States, deaths were attributed to subarachnoid hemorrhage (2.4/100,000), intracerebral hemorrhage (9.4/100,000), ischemic stroke (42.2/ 100,000), or sequelae of stroke (7.8/100,000).18 Nearly half of all stroke deaths (48.3%) occurred “prior to transport” or by the time of arrival at the emergency department. This was a much less frequent occurrence among younger than older stroke victims and somewhat less frequent among other race/ethnic groups than non-Hispanic Whites. “Pretransport” deaths, including those in persons who died before transportation could be (or would be) provided, were about twice as frequent for ischemic stroke (23.3%) as for subarachnoid or intracerebral hemorrhage (12–14%). Deaths due to sequelae of stroke (69.1%) were exceptionally common. This may reflect difficulty in assigning cause of death for victims of unobserved strokes who were found dead after significant delays. Another factor to be considered is the occurrence of stroke death in nursing homes, rather than in hospital, especially among older stroke victims. Europe and Elsewhere Population differences in stroke mortality are well documented for the baseline period of the WHO MONICA Project from 1985 to 1987 (Figure 5-4).19 Stroke surveillance was reported for 18 populations
46894_CH05_089_110.qxd
2/25/10
3:25 PM
Page 95
RATES
95
Figure 5-4 Annual Mortality from Stroke in 18 Populations in the WHO MONICA Project, by Sex, 1985–1987. For key to MONICA abbreviations, see Appendix 4-A. Source: Reprinted with permission from P Thorvaldsen, K Asplund, K Kuulasmaa, AM Rajakangas, and M Schroll, Stroke, Vol 26, p 366, American Heart Association.
46894_CH05_089_110.qxd
96
2/25/10
3:25 PM
Page 96
CHAPTER 5 STROKE
in 10 countries located in eastern and western Europe and the People’s Republic of China. Both men and women who died at ages 35–64 years were included. The figure indicates stroke mortality for each country and reporting area, separately for men and women in descending order of their observed rates for definite stroke. Stroke mortality ranged about fourfold from lowest to highest for both men and women. Rates were considerably lower for women than for men in each of the 18 populations. Stroke types were not described. Global estimates of stroke deaths are summarized in Table 5-4.13 Exceptionally large numbers are attributed to portions of southeast Asia and the Western Pacific regions, dominated by the large populations of India and China, respectively, and Europe, with the low child–high adult mortality in its central and eastern states, many in the former Soviet Union. Here, too, differentiation of stroke types is lacking.
and 6.3 for White men and women, respectively, and 12.1 and 10.5 for Black men and women, respectively. Men and Blacks experienced the greater rates at every age. Stroke surveillance in one community (Corpus Christi, TX) provided an opportunity to compare stroke incidence under the same study protocol in Mexican Americans and non-Hispanic Whites. Higher rates occurred for every stroke type in Mexican Americans, another racial/ethnic group at increased risk over Whites.12 Corresponding findings from the Northern Manhattan Stroke Study in the mid-1990s indicated population-based stroke incidence 2.4 times as great among Blacks and 2 times as great among Hispanics as among Whites.21 Additional incidence data have been reported from the Mayo Clinic/Olmsted County population and from the Greater Cincinnati/Northern Kentucky studies. The Chart Book presents age-specific stroke incidence data from ARIC and FRS, the latter spanning the years 1980–2003.20 Rates for Framingham men and women, though including an earlier decade, are lower at comparable ages than those for ARIC: For example, at age 55–64, rates per 1000 personyears were 6.1 and 4.6 for men and women in ARIC but only 4.3 and 2.2 for FRS. These different local estimates would yield markedly different projections for the US population. Perhaps different ascertainment methods account for the variation: ARIC used self-reported physician diagnosis of stroke or TIA, whereas FRS utilized in-hospital examination or review of hospital records. In the absence of standardization of methods, such variation is not surprising but nonetheless limits the value of the data for projection to other populations.
Incidence The United States As is the case with coronary heart disease, incidence of stroke is unknown for the United States nationally for any or all types.12 Compilation of information from the Greater Cincinnati/Northern Kentucky Stroke Study, the Framingham Heart Study (FRS), the Atherosclerosis Risk in Communities (ARIC) Study, and the National Heart, Lung and Blood Institute yields an estimate of 500,000 cases of first stroke each year and 200,000 recurrent strokes, a total of 700,000 new and recurrent attacks at all ages. The Incidence and Prevalence 2006 Chart Book (see Chapter 4) reports stroke incidence from the ARIC cohort, with data pooled over the years 1987–2001 for stroke and TIA combined, at ages 45–84 years.20 The rates, per 1000 person-years of follow-up, were 7.7 Table 5-4
Africa 1 172
2 187
Europe and Elsewhere In addition to mortality, the MONICA Project provides valuable data on incidence and survival from
Numbers of Deaths (ⴛ1000) Due to Cerebrovascular Disease by Mortality Strata in Each WHO Region, 2002 South-East Eastern Western The Americas Asia Europe Mediterranean Pacific Total 3 187
4 239
5 26
4 162
5 897
3 414
4 284
6 749
1. high child, high adult mortality 2. high child, very high adult mortality 3. very low child, very low adult mortality 4. low child, low adult mortality 5. high child, high adult mortality 6. low child, high adult mortality Source: Data from The World Health Report 2004, Statistical Annex, Table 2, pp 122–123.
4 55
5 172
3 150
4 1807
5509
46894_CH05_089_110.qxd
2/25/10
3:25 PM
Page 97
RATES
stroke.19 The total stroke event rates presented in Table 5-5 combine initial and recurrent definite strokes of all types. These rates ranged threefold for men, from 121 per 100,000 population in Italy (Friuli) to 359 in Finland (Kuopio), and fivefold for women, from 58 per 100,000 in Germany (Rhein-Neckar Region) to 294 in Russia (Novosibirsk). These stroke events, all classified as definite stroke in keeping with the MONICA protocol, constituted a very high proportion of all strokes reported in each population, typically well over 90%. The proportion of strokes that were recurrent, rather than truly incident, events was below 10% in only one population for both men and women but above 25% in only two populations each for men and for women. Thus generally from 75% to 90% of these definite strokes were initial events. Two other approaches to assessing incidence of stroke in Europe are noteworthy. The European Registers of Stroke (EROS) Collaboration established population-based registers in three areas—Erlangen, Germany; Dijon, France; and London, UK.22 Firstoccurring strokes of all types in 1995–1997 were ascertained by several standardized methods in all three areas. Relative to reference rates in Dijon, the rate in London was 20% higher and that in Erlangen was nearly 40% higher. Types of stroke varied among the communities and case-fatality ranged from 27% in Table 5-5
Locationa CHN-BEI DEN-GLO FIN-KUO FIN-NKA FIN-TUL GER-HAC GER-KMS GER-RDM GER-RHN ITA-FRI LTU-KAU POL-WAR RUS-MOC RUS-MOI RUS-NOI SWE-GOT SWE-NSW YUG-NOS
97
Dijon to 41% in London after adjusting for differences in age, sex, and stroke type. The potential for rigorous comparison of stroke experience across diverse populations was clearly demonstrated. Another approach was to estimate current and projected stroke incidence and prevalence for the European Union countries plus Iceland, Norway, and Switzerland (members of the European Fair Trade Association), on the basis of review of 44 populationbased studies of stroke incidence from 14 European countries.23 The report of this project outlines details of methods used, shortcomings of data available from the multiple independent studies cited, and the resulting estimates by which the World Health Organization can forecast the stroke burden to 2025. Current incidence was estimated to be 1.1 million new cases per year in the included countries. The urgency of collecting standardized stroke data for the region was emphasized, and surveillance through the WHO STEPS Stroke System was proposed as the mechanism. Case-Fatality Reported case-fatality from stroke varies among data sources depending upon age, sex, race/ethnicity, and aspects of case definition, as well as possible differences in initial severity, accessibility and quality of
Stroke Attack Rates (per 100,000/yr), Recurrence, and 28-Day Case-Fatality in 18 Populations of the WHO MONICA Project, by Sex, 1985–1987 Males Females Stroke % % CaseStroke % % CaseRate Recurrent Fatality Rate Recurrent Fatality 240 177 359 293 264 150 167 136 137 121 286 152 251 229 344 128 216 235
27 21 18 17 25 17 23 17 23 13 25 13 21 25 27 9 20 20
28 22 17 27 21 36 32 30 16 34 23 49 32 38 25 18 15 28
169 93 194 124 105 84 102 74 58 63 146 76 136 123 294 67 115 110
27 17 16 15 15 17 21 19 13 11 14 12 24 26 23 6 18 15
37 26 18 31 22 36 34 36 23 39 22 57 38 39 22 25 21 44
a
For key to MONICA abbreviations, see Appendix 4-A.
Source: Reprinted with permission from P Thorvaldsen, K Asplund, K Kuulasmas, AM Rajakangas, and M Schroll, Stroke, Vol 26, © American Heart Association.
46894_CH05_089_110.qxd
98
2/25/10
3:25 PM
Page 98
CHAPTER 5 STROKE
treatment, and other factors. For persons aged 45–64 in the ARIC Study, death occurred within 30 days (a nonstandard interval) in 8–12% of ischemic strokes and 37–38% of hemorrhagic strokes. Similar rates were described in a four-community study of persons aged 65 years or older enrolled in Medicare: 1-month case-fatality of 8.1% for ischemic strokes and 44.6% for hemorrhagic strokes.12 The finding that hemorrhagic strokes confer substantially higher early mortality is consistent with other sources. This observation also underscores the importance of distinguishing among stroke types, given marked differences in their clinical course. Case-fatality in the MONICA stroke centers was assessed for all strokes, regardless of type, for new and recurrent events together (Table 5-5).19 Casefatality ranged, for men, from 15% to 49%, and for women from 18% to 57%. The upper limits were in Warsaw, Poland, for both men and women, and were “outliers,” being uniquely high values. This suggests exceptional circumstances in Warsaw that were not typical of other centers. Excluding Poland, the range was limited to 38% for men and 44% for women, still at a level seen only for hemorrhagic stroke in the US data, but the comparison may be inappropriate on several grounds. Threefold variation in casefatality raised questions about possible differences among reporting areas in case severity, treatment, or other influences.
stroke among respondents (Table 5-6).24 In response to the interview question “has a doctor ever told you that you had a stroke?” 2.6% of 356,112 respondents in all 50 states and other jurisdictions indicated “yes,” leading to the estimate of 5,839,000 prevalent cases of stroke in the noninstitutionalized US population age 18 years or older. Prevalence increased with age (65 years and older, 8.1%), male sex (2.7%), race/ ethnicity (Blacks 4.0%, American Indian/Alaska Native 6.0%), and education (less than high school 4.4%, college graduate 1.8%). Variation by state was about twofold, with Connecticut (1.5%) and Mississippi (4.3%) the lower and upper extremes. The analysis of European population studies of stroke cited previously included assessment of agespecific prevalence and projection to estimate prevalence for each European Union country and European Fair Trade Association participant, including those lacking relevant data altogether.23 At age 65–74 years, for example, prevalence was estimated to be as high as 14.2% for men in Portugal and as low as 1.5% for women in Cyprus. This nearly 10-fold range calls attention to the need for countries to assess their own burdens of stroke and other major chronic conditions in order to devise and implement effective preventive strategies. Lack of data is a serious handicap, but the interim device of estimation from the most extensive available data—recognizing their limitations—is an expedient approach.
Prevalence The prevalence of stroke in the adult US population aged 20 years and older was estimated for the year 2006 for the total population and separately for males and females: 6.5 million, 2.6 million, and 3.9 million, respectively.12 The basis for these estimates is a selfreported history of a nonfatal stroke by participants in the National Health and Nutrition Examination Survey, 1999–2004. Percent frequencies of persons living with a history of stroke were also given, for the total population and for multiple racial/ethnic groups. The highest prevalence, for American Indians/Alaska Natives at 6.0%, was noted as being unreliable. This suggests that adequate data are not being collected for this population, already seen to have exceptionally high stroke mortality. For Black men and women, prevalence was also relatively high in comparison with non-Hispanic Whites—for men, 3.9% versus 2.3% and for women 4.1% versus 3.2%. Prevalence was lowest for Mexican American males at 2.1%. In the United States, the Behavioral Risk Factor Surveillance System also provided an opportunity for the year 2005 to include inquiry about history of
Disability The aftermath of stroke for those who do survive is often one of significant disability. According to the Atlas of Stroke Hospitalizations among Medicare Beneficiaries, among beneficiaries hospitalized for stroke, just half (51.0%) are discharged home, nearly one in ten die in hospital, and nearly 40% are discharged to a skilled nursing or other care facility.25 At three months after hospital discharge, 20% of stroke survivors require institutional care, and 15–20% are permanently disabled. Within 5 years, 13% of men and 22% of women 40 to 69 years old at the time of stroke have a recurrence.12 Participants in the Medical Expenditure Panel Survey (MEPS) of a representative sample of the US population for the years 2000 and 2002, described in Chapter 4, were asked whether they “had ever been diagnosed as having had a stroke or transient ischemic attack.”26 Compared with survey participants without a history of stroke, after adjustment for differences in sociodemographic factors, risk factors, and comorbidities, stroke survivors had significantly poorer ratings on all four of the HRQoL scales—
46894_CH05_089_110.qxd
2/25/10
3:25 PM
Page 99
RATES
Table 5-6
99
Percentage of Respondents Who Reported a History of Stroke by Selected Characteristics—Behavioral Risk Factor Surveillance System, United States, 2005 Estimated No. of Total No. of Prevalence of US residents Respondents* Stroke† 95% Cl‡ living with stroke§
Characteristic Age (yrs) 18–44 45–64 65 Sex¶ Male Female Race/Ethnicity¶ White, non-Hispanic Black, non-Hispanic Asian Hispanic American Indian/Alaska Native Multiracial Education¶ Less than 12 years High school graduate Some college College graduate Total
128,328 137,738 87,351
0.8 2.7 8.1
0.7–0.9 2.5–2.9 7.7–8.5
852,000 1,926,000 3,036,000
136,201 219,911
2.7 2.5
2.5–2.8 2.4–2.7
2,694,000 3,145,000
279,419 27,925 5974 25,539 5535 6519
2.3 4.0 1.6 2.6 6.0 4.6
2.3–2.4 3.6–4.5 1.0–2.7 2.1–3.3 4.5–7.8 3.7–5.6
4,017,000 772,000 60,000 616,000 126,000 136,000
38,202 109,830 93,228 113,944 356,112
4.4 2.6 2.7 1.8 2.6
4.0–4.9 2.5–2.8 2.5–2.9 1.6–1.9 2.5–2.7
1,365,000 1,863,000 1,474,000 1,108,000 5,839,000
*The sums of the sample sizes in each category may not add up to the total number of respondents because of unknown or missing information. There are 2695 respondents with unknown or missing age, 5201 with unknown or missing race/ethnicity, and 908 with unknown or missing level of education. † Weighted percentage of respondents who report a history of stroke. ‡ Confidence interval. § Estimated number of US residents with a stroke history. ¶ Weighted percentages are age-adjusted to the 2000 US standard population. Source: Centers for Disease Control and Prevention. The Prevalence of Stroke—United States, 2005. MMWR 2007; Vol 56, pp 470–474.
mental health, physical health, health utility, and selfrated health (Table 5-7). Black–White disparities in HRQoL among survey participants overall were amplified among those who were stroke survivors. On a global level, disability due to stroke was described in Chapter 1 as contributing increasingly to disability worldwide as measured by disabilityadjusted life years (DALYs), ranking sixth in 1990 and projected to rank fourth by 2020.27 Disparities Variation in the frequency and burden of stroke by age, sex, race/ethnicity, and geography has been alluded to previously in this chapter and in Chapter 2 (Figure 2-5), which illustrates striking excess prevalence of stroke among Native Hawaiians/Pacific Islanders and American Indians/Alaska Natives relative to non-Hispanic Whites.28 Vital statistics and health-related surveys have provided evidence of racial/ ethnic disparities in death from each major stroke subtype, with different groups being most affected;29 in overall stroke death rates, age at death and years
of potential life lost;30 and prevalence of cardiovascular risk factors among those who have survived a stroke.31 Multiple groups experience these differential risks and consequences of stroke, which are to be eliminated as one of two overarching goals of the health blueprint for the decade in the United States, Healthy People 2010.32 Attention has been called especially to the burden of stroke in Blacks, for whom numbers of deaths increased in the 1990s as the decline in rates (see below) slowed, and for whom the excess burden is especially great in nonmetropolitan areas of the southeast.33 Achievement of the Healthy People objective of 20% reduction in the overall stroke death rate and elimination of disparities would entail reduction by 40% among Blacks in the same time period. Review of studies on stroke incidence or mortality in relation to socioeconomic status in Europe, the United Kingdom, and the United States points to increased risks among lower socioeconomic groups.34 This aspect of disparities in atherosclerotic and hypertensive diseases is discussed in Chapter 16.
51.1 (0.1) 51.3 (0.2) 52.5 (0.2) 51.8 (0.3) 51.8 (0.5) 52.2 (0.1) 50.4 (0.1) 51.3 (0.1) 51.1 (0.2) 51.2 (0.3) 50.9 (0.2) 51.3 (0.1) 51.2 (0.2) 51.6 (0.2) 51.2 (0.1) 51.3 (0.1) 50.3 (0.2) 51.5 (0.1) 49.2 (0.3) 51.4 (0.1) 49.4 (0.3) 51.4 (0.1) 49.2 (0.2) 51.9 (0.1)
45.1 (1.2) 45.9 (1.0) 48.5 (0.9) 48.9 (0.7) 46.5 (1.5) 49.1 (0.6) 45.0 (0.7) 47.9 (0.5) 45.4 (1.0) 43.3 (3.4) 44.6 (1.4) 47.5 (0.5) 46.7 (1.0) 48.3 (0.9) 47.0 (0.7) 47.4 (1.0) 47.4 (0.5) 47.4 (0.9) 46.1 (0.9) 47.8 (0.5) 46.2 (0.8) 48.0 (0.5) 45.4 (1.0) 48.0 (0.5)
36.7 (1.0) 35.1 (0.5)
31.9 (0.8) 37.4 (0.5)
31.3 (0.7) 36.9 (0.5)
34.1 (0.5) 38.5 (0.8)
34.4 (1.1) 35.8 (1.0) 34.8 (0.9) 37.4 (0.7)
33.6 (1.4) 35.7 (0.5)
35.5 (0.1) 35.3 (1.1) 37.8 (2.7)
48.9 (0.2) 49.8 (0.1)
37.8 (0.4) 50.2 (0.1)
40.9 (0.3) 50.1 (0.1)
43.3 (0.2) 50.1 (0.1)
50.1 (0.2) 50.0 (0.2) 48.9 (0.2) 49.9 (0.1)
50.3 (0.2) 49.5 (0.1)
49.6 (0.1) 49.3 (0.2) 50.1 (0.3)
50.4 (0.1) 48.8 (0.1)
52.0 (0.1) 47.6 (0.2) 43.8 (0.3) 39.4 (0.3) 35.5 (0.5)
49.6 (0.1)
0.71 (0.01) 0.69 (0.01)
0.65 (0.04) 0.71 (0.01)
0.62 (0.02) 0.72 (0.01)
0.68 (0.01) 0.72 (0.01)
0.67 (0.02) 0.71 (0.01) 0.69 (0.01) 0.71 (0.02)
0.59 (0.04) 0.70 (0.01)
0.70 (0.01) 0.67 (0.02) 0.64 (0.07)
0.72 (0.01) 0.67 (0.01)
0.73 (0.02) 0.67 (0.01) 0.72 (0.01) 0.70 (0.01) 0.60 (0.03)
0.69 (0.01)
0.84 (0.00) 0.88 (0.00)
0.74 (0.01) 0.88 (0.00)
0.77 (0.01) 0.88 (0.00)
0.80 (0.00) 0.89 (0.00)
0.88 (0.00) 0.87 (0.00) 0.86 (0.00) 0.88 (0.00)
0.88 (0.00) 0.87 (0.00)
0.87 (0.00) 0.86 (0.00) 0.89 (0.01)
0.88 (0.00) 0.86 (0.00)
0.90 (0.00) 0.84 (0.00) 0.82 (0.00) 0.78 (0.01) 0.73 (0.01)
0.87 (0.00)
Source: Reprinted with permission from Stroke, Vol 37, pp 2567–2572. © 2006, American Heart Association, Inc.
59.1 (1.9) 62.5 (0.9)
55.5 (1.4) 66.7 (0.9)
55.4 (1.8) 63.9 (0.9)
59.7 (0.9) 65.8 (1.4)
59.9 (2.3) 63.3 (1.4) 59.9 (1.4) 63.7 (1.6)
53.9 (3.0) 62.1 (0.8)
61.9 (0.9) 60.5 (2.2) 58.8 (4.1)
62.5 (1.3) 60.9 (0.9)
60.6 (2.2) 61.0 (1.5) 62.8 (1.8) 62.6 (1.4) 58.2 (3.3)
61.6 (0.8)
77.2 (0.3) 81.4 (0.2)
64.7 (0.6) 81.3 (0.1)
67.3 (0.5) 81.3 (0.2)
72.5 (0.3) 82.4 (0.2)
81.0 (0.3) 81.1 (0.3) 80.1 (0.3) 80.0 (0.3)
80.4 (0.4) 80.5 (0.2)
80.6 (0.2) 79.9 (0.4) 80.1 (0.6)
81.5 (0.2) 79.5 (0.2)
82.9 (0.2) 78.1 (0.3) 76.1 (0.4) 71.1 (0.6) 66.0 (1.0)
80.5 (0.2)
3:25 PM
37.0 (0.7) 34.3 (0.5)
40.5 (1.2) 37.0 (1.0) 35.6 (0.7) 34.1 (0.8) 29.9 (1.0)
35.6 (0.5)
Self-Rating of Health (EQ VAS) Stroke Nonstroke Mean (SE) Mean (SE)
2/25/10
Note: The sample included 1040 stroke and 38,640 nonstroke participants. All were significant between stroke and nonstroke populations based on Z tests (p 0.05).
51.3 (0.1)
47.4 (0.4)
HRQoL in Stroke and Nonstroke Populations by Selected Characteristics, MEPS 2000 and 2002 Mental Health Score Physical Health Score Health Utility Score (MSC-12) (PCS-12) (EQ-5D Index, US) Stroke Nonstroke Stroke Nonstroke Stroke Mean Nonstroke Mean (SE) Mean (SE) Mean (SE) Mean (SE) Mean (SE) Mean (SE)
100
Overall Age 18–49 50–64 65–74 75–84 85 Gender Male Female Race White Black Others Hispanic Yes No Region Northeast Midwest South West Hypertension Yes No Diabetes Yes No Coronary artery disease Yes No Smoking Yes No
Table 5-7
46894_CH05_089_110.qxd Page 100
CHAPTER 5 STROKE
46894_CH05_089_110.qxd
2/25/10
3:25 PM
Page 101
RISKS 101
RISKS Factors predictive of the occurrence of stroke have been investigated in many cohort studies, in which characteristics at entry (baseline) were evaluated for their relation to stroke incidence over several years of follow-up. Because of particular interest in the relation of blood cholesterol concentration and blood pressure to the individual risks of stroke, a group of investigators constituting the Prospective Studies Collaboration undertook to review cohort studies of stroke in which these two particular characteristics were measured.35 Altogether, 45 studies were included, with nearly 450,000 participants followed from 5 to 30 years (mean follow-up, 16 years) and a total of 13,397 persons with stroke in 7.3 million personyears of experience. The stroke events were predominantly deaths. By adjustment for variation in measurements of blood pressure (here, diastolic only) and cholesterol within individuals on repeated occasions of observation, the “usual diastolic blood pressure” and “usual total cholesterol” were estimated.
Usual diastolic blood pressure was examined first in terms of proportional rates, or the ratio of the stroke rate in each successive stratum (category) of blood pressure to the rate in the lowest stratum. The analysis addressed the possibility that for younger persons, these ratios at successively higher levels of pressure might increase more steeply than for older persons, as shown in Figure 5-5. For example, in the highest stratum (100 mm Hg or greater) versus the lowest (below 80 mm Hg), the rates were 10 times as high at younger than age 45, five times as high at ages 45–64, and only two times as high at age 65 or older. Thus, the relative importance of blood pressure was greatest for the younger adults and least for the oldest ones. However, the total or absolute impact on the rate of strokes was much greater for the oldest group because the stroke rate even in the lowest blood pressure category was many times greater for older than for younger adults. Figure 5-6 shows that the doubling of the reference rate for those 65 years and older at entry reflected an absolute increase in stroke rate of 8.4
Figure 5-5 Proportional Stroke Risk by Age and Usual Diastolic Blood Pressure. Source: Reprinted with permission from Prospective Studies Collaboration, The Lancet, Vol 346, p 1651, © 1995, The Lancet, Ltd.
46894_CH05_089_110.qxd
102
2/25/10
3:25 PM
Page 102
CHAPTER 5 STROKE
Figure 5-6 Absolute Stroke Risk, by Age and Usual Diastolic Blood Pressure. Source: Reprinted with permission from Prospective Studies Collaboration, The Lancet, Vol 346, p 1651, © 1995, The Lancet, Ltd.
per 1000 (18.4–10.0), whereas the 10-fold increase among the youngest group added less than 2 per 1000. Crude estimation of the numbers of years of life lost due to stroke death emphasized the relative cost of fatal strokes in younger adults. From one or another vantage point, then, the cost of stroke to both younger and older adults can be argued. No difference was reported in results for women and men. The largest single follow-up study of stroke mortality was not included in the Prospective Studies Collaboration. In the six-year follow-up study, strokes were ascertained through the National Death Index for the decedents among 350,977 middle-aged US men screened for the Multiple Risk Factor Intervention Trial, or MRFIT.36 The relation of age, systolic and diastolic blood pressure, serum cholesterol concentration, cigarette smoking, and race (as reflected in the percentage of Blacks in each stroke category) was examined among three types of stroke: subarachnoid hemorrhage (55 deaths), intracranial
hemorrhage (83 deaths), and nonhemorrhagic (that is, thrombotic or embolic) stroke (92 deaths). These comparisons of factors distinguished between men who died of each type of stroke and those who did not die of stroke. For each type of stroke death, there were significantly higher mean values of systolic and diastolic blood pressure, a higher prevalence of diastolic pressure of 90 mm Hg or greater, and a greater proportion of cigarette smokers than among men without stroke death. In the Prospective Studies Collaboration, results for total cholesterol indicated no relation to stroke rates adjusted for study, age, sex, diastolic blood pressure, history of coronary heart disease, and ethnicity. It was noted that this result might reflect opposite effects of high and low cholesterol concentrations on different types of stroke, which could not be distinguished for this analysis. In MRFIT, the mean values of serum cholesterol concentration among the groups with subarachnoid
46894_CH05_089_110.qxd
2/25/10
3:25 PM
Page 103
RISKS 103
and intracranial hemorrhage (212.2 and 211.4 mg/dl, respectively) were slightly less than those of the group without stroke (214.4 mg/dl), but in the group with nonhemorrhagic stroke, they were significantly greater. This observation indicates type-specific differences in the relation between cholesterol concentrations and stroke. There were not equal and opposite effects of cholesterol concentration on hemorrhagic and nonhemorrhagic stroke, as suggested by the Prospective Studies Collaboration. However, in further analysis (not shown), the lowest stratum of cholesterol concentration (less than 160 mg/dl) was associated with a risk three times that of men with cholesterol concentrations of 160 mg/dl or greater. Thus, there were in fact opposite associations as discussed in the report of the collaborative review, but in this study the effect of the inverse relation with hemorrhagic stroke was dominated by the increased risk of nonhemorrhagic stroke with increased cholesterol concentration. Ability to investigate associations of blood lipids, and possibly other factors, on risk of stroke appears to depend crucially on identification of specific stroke types, as urged by Stallones 40 years ago.3 The MRFIT data also show that cigarette smoking was significantly more frequent among subjects in each stroke group than among subjects without stroke. The percentages of Blacks among the stroke deaths were greater in all groups than in the group without strokes, significantly so for both intracranial hemorrhage and nonhemorrhagic stroke. Other prominent predictors of stroke are prior cardiovascular conditions, such as coronary heart
disease, cardiac failure, and atrial fibrillation (chronic irregularity of contraction of the upper chambers of the heart). These conditions, in addition to hypertension, were found in the Framingham Heart Study to occur especially commonly among older persons with stroke (Table 5-8).37 Even these conditions, however, with their direct pathological connections to risks of embolic stroke, contributed less than hypertension to the attributable risk (the proportion of events explained by the presence of that condition). Hypertension (high blood pressure) remains the dominant characteristic in the prediction of stroke. Many of these factors are discussed in subsequent chapters. A compendium of more than 30 “well-documented and modifiable risk factors” for stroke is presented in an American Heart Association/American Stroke Association Guideline, Primary Prevention of Ischemic Stroke, published in 2006.38 Data presented for each factor are its prevalence, population-attributable risk, relative risk, and evidence—when available—of reduction of risk with treatment. For several factors, separate estimates of these measures are given by sex or age. A greatly simplified list constitutes the Modified Framingham Stroke Risk Profile, one of several scoring schemes for clinical use to identify patients at high risk of stroke. Separately for men and women, the scheme incorporates information on eight factors, each weighted according to the strength of its association with stroke as observed in the Framingham Heart Study. Age and systolic blood pressure are both scored from 0 to 10 on a graded scale. Systolic blood pressure is scored differently for a given range of values according to treatment status.
Table 5-8
Contributions of Other Cardiovascular Conditions to Risk of Stroke, by Age, Framingham Heart Study Age (Yr) Cardiovascular Condition 50–59 60–69 70–79 80–89 Number of stroke events 92 213 192 75 Hypertension Attributable risk (%) 48.8 53.2 48.6 33.4 % of events in persons with condition 72.8 80.3 83.9 84.0 Coronary heart disease Attributable risk (%) % of events in persons with condition
11.1 25.0
12.4 32.9
12.6 38.0
0.0 28.0
Cardiac failure Attributable risk (%) % of events in persons with condition
2.3 9.8
3.1 11.7
5.6 18.2
6.0 18.7
Atrial fibrillation Attributable risk (%)a % of events in persons with condition
1.5 6.5
2.8 8.5
9.9 18.8
23.5 30.7
Significant increase with age (p 0.01).
a
Source: Reprinted with permission from PA Wolf, Circulation, Vol 88, p 2475, © 1993, American Heart Association.
46894_CH05_089_110.qxd
104
2/25/10
3:25 PM
Page 104
CHAPTER 5 STROKE
For example, if the systolic pressure is 130 mm Hg in a patient not being treated with antihypertensive medication, three points are entered into the score; if the patient is at that blood pressure level and on treatment, five points are added. (It is noteworthy that risks are unequal under these two scenarios at the same level of blood pressure.) Each of the remaining elements of the score—history of diabetes, cigarette smoking, cardiovascular disease, atrial fibrillation, and left ventricular hypertrophy on the electrocardiogram—is scored in a similar manner but with different levels of risk. Absence of the condition adds zero, whereas its presence adds from two to six points depending on the factor and sex of the patient. The second part of the assessment translates the score into a percentage risk of experiencing a stroke in the next 10 years. Scores of 1 (lowest), 10, 20, and 30 (highest) indicate risks for men of 3, 10, 37, and 88%. For women, scores of 1, 10, 20, and 27, the maximum attainable for women, correspond to risks of 1, 6, 37, and 84%. These risk factors, based on a single community in the United States, have much in common with the main contributors to risk of stroke death globally, as
estimated by the Global Burden of Diseases and Risk Factors Study, illustrated in Table 1-8.39 Blood pressure far exceeds other factors in importance. Cholesterol (perhaps because of mixed stroke types in the mortality data) is second by a wide margin, followed by overweight and obesity and smoking. Low fruit and vegetable intake and physical inactivity are the remaining major factors.
TRENDS The United States Stroke mortality has decreased in the United States throughout the 20th century. This was demonstrated clearly for the period from 1900 to 1960 in the epidemiologic review and analysis of stroke by Stallones published in 1965 (Figure 5-7).3 This reconstruction of historical vital statistics for the United States was based on the nearest equivalent of the category, “vascular lesions of the central nervous system,” in the first through sixth revisions of the International Classification of Diseases. Age-adjusted rates were pre-
Figure 5-7 Secular Trend of Mortality from Stroke in the United States, 1900–1960. Source: Reprinted with permission from RA Stallones, Journal of Chronic Diseases, Vol 18, p 864, © 1965, Elsevier Science Inc.
46894_CH05_089_110.qxd
2/25/10
3:25 PM
Page 105
TRENDS 105
sented to reflect a constant age structure of the population over this period and to remove the effect of the actual increased proportions of older persons, with the highest stroke rates, over time. (This results in a different picture than Figure 1-2, which reflects crude death rates from stroke as essentially constant from 1900 to 1970 at about 100/100,000 population per year.) The figure illustrates for these six decades a continuous decrease from more than 125 stroke deaths per 100,000 population per year to 75 or fewer, allowing for the noticeable effects of change in classification under the Sixth Revision, as shown. Stroke mortality in the United States continued to decline through the late 1980s, as was shown in the context of such changes in 26 other countries in Figure 2-9. A more current update to 2004 shows the still downward trend in age-adjusted stroke mortality reaching approximately 50/100,000 (Figure 5-8).15 The observation made in previous chapters in connection with cardiovascular deaths and coronary heart disease deaths applies to stroke deaths as well: although age-adjusted mortality has declined dramatically, the numbers of deaths have declined much less. The epidemiologic good news about rates is dampened by the public health message about a continuing and expected increasing burden.
Further observations about stroke trends in the United States indicate a geographic shift over the three decades of the 1960s through the 1980s.40 While stroke mortality was declining nationally, the stroke belt of the southeastern United States became less clearly defined. This region continued to have some of the highest stroke mortality in the nation but was less homogeneous at the end of the period. Meanwhile, areas of equally high stroke mortality emerged in the Mississippi River Valley, a trend that has continued. Europe and Elsewhere The WHO MONICA Project was able to monitor trends in stroke mortality and event rates in 15 centers in 9 countries—in western, central, and eastern Europe and China.41 Figures 5-9a, b, and c present 10-year changes in stroke rates from the mid-1980s to the mid-1990s, among 35- to 64-year-old men (left panels) and women (right panels) based on 10,442 stroke events in 23.4 million person-years of observation. Both coronary heart disease (CHD) and stroke events are shown: mortality (a), event rates (b), and casefatality (c). Stroke rates for all three classes of events exhibited mixed trends among the several populations. Increases in case-fatality and mortality occurred in several centers in eastern Europe, while decreases
125
Deaths in Thousands (Bar)
100 150
75 100 50
50 25
0
79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 00 01 02 03 04 Years * The comparability ratio 1.0502 was applied to the deaths and rates reported in vital statistics for 1979–1998.
Deaths/100,000 Population (Line)
200
0
Figure 5-8 Deaths and Age-Adjusted Death Rates for Stroke, US, 1979–2004. Source: From Morbidity & Mortality: 2007 Chart Book on Cardiovascular, Lung and Blood Diseases. National Institutes of Health, National Heart, Lung and Blood Institute, June 2007.
46894_CH05_089_110.qxd
106
2/25/10
3:25 PM
Page 106
CHAPTER 5 STROKE
*
*
*
* 0
4
*
* *
*
*
*
24
*
*
28
212
8
Coronary Stroke
*
*
*
**
* *
*
Coronary Stroke
*
FIN-TUL FIN-NKA RUS-MOC FIN-KUO SWE-GOT ITA-FRI SWE-NSW POL-WAR RUS-MOI DEN-GLO CHN-BEI LTU-KAU RUS-NOC YUG-NOS RUS-NOI
*
212
A
*
Women
*
Men SWE-NSW FIN-NKA FIN-KUO FIN-TUL SWE-GOT ITA-FRI DEN-GLO YUG-NOS RUS-NOC POL-WAR LTU-KAU RUS-MOC RUS-MOI CHN-BEI RUS-NOI
28
24
0
* *
* **
*
*
*
*
210 28 26 24 22
6
8 10
* * *
*
*
4
Coronary Stroke
*
*
* 2
% change over 10 years
POL-WAR ITA-FRI FIN-TUL LTU-KAU FIN-NKA RUS-NOC SWE-NSW YUG-NOS FIN-KUO SWE-GOT CHN-BEI RUS-MOC RUS-MOI DEN-GLO RUS-NOI
*
* 0
4
*
*
*
*
Coronary Stroke
*
28 26 24 22
2
Women
*
Men
C
0
% change over 10 years
% change over 10 years
SWE-NSW ITA-FRI FIN-NKA POL-WAR YUG-NOS FIN-TUL RUS-NOC SWE-GOT FIN-KUO LTU-KAU CHN-BEI RUS-NOI DEN-GLO RUS-MOC RUS-MOI
*
4
Coronary Stroke
*
2
*
* * **
*
0
*
*
* * *
*
26 24 22
*
RUS-MOC FIN-NKA FIN-TUL FIN-KUO SWE-GOT RUS-MOI DEN-GLO SWE-NSW ITA-FRI CHN-BEI POL-WAR RUS-NOI RUS-NOC LTU-KAU YUG-NOS
*
Coronary Stroke
210 28
B
8
Women
*
Men FIN-NKA FIN-KUO SWE-NSW FIN-TUL DEN-GLO SWE-GOT RUS-MOC ITA-FRI RUS-MOI YUG-NOS POL-WAR RUS-NOC LTU-KAU CHN-BEI RUS-NOI
4
% change over 10 years
% change over 10 years
28 26 24 22
0
2
4
6
8 10
% change over 10 years
*Statistically significant changes (p 0.05)
Figure 5-9 Percentage Change Over 10 Years in Age-Standardized CHD and Stroke Mortality Rates (Panel a), Event Rates (Panel b), and Case Fatality (Panel c) in Men and Women Aged 35 to 64 Years. Source: Reprinted with permission from Truelsen T et al., Stroke, Vol 34, pp 1348, 1349, and 1350, © 2003 American Heart Association, Inc.
46894_CH05_089_110.qxd
2/25/10
3:25 PM
Page 107
REFERENCES 107
occurred elsewhere. Changes in event rates were less consistent. Stroke and CHD trends were generally, though not always, in the same direction in any given population. Trends in blood pressure levels and other risk factors were of particular interest in seeking an explanation of the stroke trends.42 For women, but not for men, trends in blood pressure accounted for a significant component (38%) of the variation in stroke event trends, after including a 3- or 4-year time lag in the analysis. Combining other risk factors—smoking, serum cholesterol, and body mass index—did not add significantly to explanation of the trends. When the components of stroke occurrence were investigated, it appeared that changes in case-fatality were more important than changes in event rates in explaining the overall mortality trends.43 The original expectation was that case-fatality would unambiguously indicate improvement in case management rather than risk-factor improvement or, more generally, primary prevention. It was concluded instead that reduced severity of incident cases could affect case-fatality favorably, so that effects of primary prevention could not be excluded. On a broader scale, investigation of the World Health Organization Data Bank for stroke mortality from 1968 to 1994 provided information on trends especially in the latest 5-year period.44 Mortality for all strokes was analyzed for 51 industrialized and developing countries, each of which submitted data to WHO for at least 80% of all deaths over this 27-year period. Among all 51 countries, a range of about threefold was observed between the highest- and lowest-rate countries. The highest rates, as in the WHO MONICA Project experience, were in eastern Europe and countries of the former Soviet Union. Lowest rates were in the countries with steepest declining trends—the United States, Canada, Switzerland, France, and Australia. Japan, included in the WHO data analysis, has also been the subject of separate investigation, revealing that the very rapid decline in stroke mortality from the mid-1960s changed abruptly with a greatly reduced pace of change in the 1990s.45 The need for stroke surveillance especially in high-mortality countries is underscored by these findings, in order both to monitor changes in components of stroke occurrence and to study determinants of stroke in the many varied settings that comparative studies can exploit.
ening picture for decades to come. The forecast of the Global Burden of Disease Study ranks stroke second among causes of death worldwide in 1990 and 2020; seventh in years of life lost in 1990 and third in 2020; and sixth in DALYs in 1990 and fourth in 2020 (see Table 1-6).27 With special concern about impending increases in the burden of stroke in low- and middle-income countries, The Lancet Neurology presented a series of reports in early 2007 addressing needs for improved surveillance and application of preventive strategies.46,47 It is observed that success in reducing case-fatality from stroke will have the effect of increasing prevalence and disability, increasing the demand for rehabilitation and long-term care. Primary prevention of stroke becomes in this scenario a fundamental and urgent approach in addressing stroke at the population level.
CURRENT ISSUES Among current issues in the epidemiology and prevention of stroke, the foremost questions from the US perspective are: Can the historical decline in mortality due to stroke be sustained and, while continuing, achieve elimination of the longstanding racial/ ethnic disparities among populations? On a global basis, especially in regions where stroke predominates over coronary heart disease among the atherosclerotic and hypertensive diseases, can stroke be prevented effectively on a populationwide scale, so as to avert the forecasts of continuing contributions of stroke to death and disability? Can the experience of investigators in many centers with diverse methods of population surveillance be exploited to establish a global network of community-based stroke surveillance centers, collaborating under a common protocol, so as to monitor and evaluate trends in occurrence of stroke as well as the impact of change in policies and practices aimed at stroke prevention? REFERENCES 1. Albers GP, Cutler RWP. Cerebrovascular diseases. In: Dale DC, Federman DD, eds. Scientific American Medicine. New York: Scientific American, Inc; 1994:1–13.
FORECASTS
2. National Center for Health Statistics. Health, United States, 2008. With Chartbook. Hyattsville, MD: 2009.
The future of the world’s burden of stroke is uncertain, but currently adopted projections suggest a wors-
3. Stallones RA. Epidemiology of cerebrovascular disease: a review. J Chronic Dis. 1965;18:859–872.
46894_CH05_089_110.qxd
108
2/25/10
3:25 PM
Page 108
CHAPTER 5 STROKE
4. Gorelick PB, Alter M. The Prevention of Stroke. New York, NY: The Parthenon Publishing Group; 2002. 5. Gillum RF, Fortmann SP, Prineas RJ, Kottke TE. International diagnostic criteria for acute myocardial infarction and stroke. Am Heart J. 1984;108:150–158. 6. Asplund K, Tuomilehto J, Stegmayr B, et al. Diagnostic criteria and quality control of the registration of stroke events in the MONICA Project. Acta Med Scand. 1988;728(suppl): 26–39. 7. Thorvaldsen P, Kuulasmaa K, Rajakangas AM, Rastenyte D, Sarti C, Wilhelmsen L. Stroke trends in the WHO MONICA Project. Stroke 1997;28(3):500–506. 8. World Health Organization. International Statistical Classification of Diseases and Related Health Problems. 10th rev. Geneva (Switzerland): World Health Organization; 1992:1. 9. Chalela JA, Kidwell CD, Nentwich LM, et al. Magnetic resonance imaging and computed tomography in emergency assessment of patients with suspected acute stroke: a prospective comparison. Lancet. 2007;369:293–298. 10. Labarthe DR, Broderick JP, Atkins D, Zheng ZJ, Yoon SS, eds. Paul Coverdell National Acute Stroke Registry Prototypes. Assessing acute stroke care in the U.S. and beyond. Am J Prev Med. 2006;31(suppl):S189–S259. 11. Brown DL, Al-Senani F, Kisabeth LD, et al. Defining cause of death in stroke patients. The Brain Attack in Corpus Christi Project. Am J Epidemiol. 2007;165:591–596. 12. Lloyd-Jones D, Adams R, Carnethon M, et al. Heart disease and stroke statistics—2009 update. A report from the American Heart Association Statistics Committee and Stroke Statistics Subcommittee. Circulation. 2009; 119:e1–e161.
Washington, DC: US Government Printing Office, December 2006. 15. National Heart, Lung and Blood Institute. Morbidity & Mortality: 2007 Chart Book on Cardiovascular, Lung, and Blood Diseases. Bethesda, MD: US Department of Health and Human Services. Public Health Service, National Institutes of Health; June, 2007. 16. Casper ML, Barnett E, Williams GI Jr, Halverson JA, Braham VE, Greenlund KJ. Atlas of Stroke Mortality: Racial, Ethnic, and Geographic Disparities in the United States. Atlanta, GA: Department of Health and Human Services, Centers for Disease Control and Prevention; 2003. 17. Casper ML, Denny CH, Coolidge JN, et al. Atlas of Heart Disease and Stroke Among American Indians and Alaska Natives. Atlanta, GA: Department of Health and Human Services, Centers for Disease Control and Prevention and Indian Health Service; 2005. 18. Centers for Disease Control and Prevention. State-specific mortality from stroke distribution of place and death—United States, 1999. MMWR. 2002;51:429–433. 19. Thorvaldsen P, Asplund K, Kuulasmaa K, Rajakangas AM, Schroll M. Stroke incidence, case fatality, and mortality in the WHO MONICA project. World Health Organization Monitoring Trends and Determinants in Cardiovascular Disease. Stroke. 1995;26: 361–367. 20. National Heart, Lung and Blood Institute. Incidence & Prevalence: 2006 Chart Book on Cardiovascular and Lung Diseases. Washington, DC: US Department of Health and Human Services, Public Health Service, National Institutes of Health; 2006.
13. World Health Organization. World Health Report 2004. Geneva (Switzerland): World Health Organization; 2004.
21. Sacco RL, Boden-Albala B, Gan R, et al. Stroke incidence among White, Black, and Hispanic residents of an urban community. The Northern Manhattan Stroke Study. Am J Epidemiol. 1998;147:259–268.
14. US Department of Health and Human Services. Healthy People 2010 Midcourse Review.
22. Wolfe CDA, Giroud M, Kolominsky-Rabas P, et al. Variations in stroke incidence and sur-
46894_CH05_089_110.qxd
2/25/10
3:25 PM
Page 109
REFERENCES 109
vival in 3 areas of Europe. European Registries of Stroke (EROS) Collaboration. Stroke. 2000;31:2074–2079. 23. Truelson T, Piechowski-Józ´wiak B, Bonita R, Mathers C, Bogousslavsky J, Boysen G. Stroke incidence and prevalence in Europe: a review of available data. Eur J Neurol. 2006;13: 581–598. 24. Centers for Disease Control and Prevention. The prevalence of stroke—United States, 2005. MMWR. 2007;56:469–474. 25. Casper ML, Nwaise IA, Croft JB, Nilasena DS. Atlas of Stroke Hospitalizations among Medicare Beneficiaries. Atlanta, GA: US Department of Health and Human Services, Centers for Disease Control and Prevention; 2008. 26. Xie J, Wu EQ, Zheng ZJ, et al. Impact of stroke on health-related quality of life in the noninstitutionalized population in the United States. Stroke. 2006;37:2567–2572. 27. Murray CJL, Lopez AD. Alternative visions of the future: projecting mortality and disability, 1990–2020. In: Murray CJL, Lopez AD, eds. The Global Burden of Disease: A Comprehensive Assessment of Mortality and Disability from Diseases, Injuries, and Risk Factors in 1990 and Projected to 2020. Boston, MA: The Harvard School of Public Health; 1996. 28. Mensah GA, Brown DW. An overview of cardiovascular disease burden in the United States. Health Affairs. 2007;26:38–48. 29. Ayala C, Greenlund KJ, Croft JB, et al. Racial/ethnic disparities in mortality by stroke subtype in the United States, 1995–1998. Am J Epidemiol. 2001;154:1057–1063. 30. Centers for Disease Control and Prevention. Disparities in deaths from stroke among persons aged 75 years—United States, 2002. MMWR Morb Mortal Wkly Rep. 2005;54(19): 477–481. 31. McGruder HF, Malarcher AM, Antoine TL, Greenlund KJ, Croft JB. Racial and ethnic disparities in cardiovascular risk factors among stroke survivors. United States 1999 to 2001. Stroke. 2004;35:1557–1561.
32. US Department of Health and Human Services. Healthy People 2010. 2nd ed. With Understanding and Improving Health and Objectives for Improving Health. 2 vols. Washington DC: US Government Printing Office; 2000. 33. Gillum RF. Stroke mortality in blacks. Disturbing trends. Stroke.1999;30(8):1711–1715. 34. Cox AM, McKevitt C, Rudd AG, Wolfe CD. Socioeconomic status and stroke. Lancet Neurol. 2006;5(2):181–188. 35. Prospective Studies Collaboration. Cholesterol, diastolic blood pressure, and stroke: 13,000 strokes in 450,000 people in 45 prospective cohorts. Lancet. 1995;346:1647–1653. 36. Iso H, Jacobs DR Jr, Wentworth D, et al., for the MRFIT Research Group. Serum cholesterol levels and six-year mortality from stroke in 350,977 men screened for the Multiple Risk Factor Intervention Trial. N Engl J Med. 1989; 320:904–910. 37. Wolf PA. Contributions of epidemiology to the prevention of stroke. Lewis A. Conner Lecture. Circulation. 1993;88:2471–2478. 38. Goldstein LB, Adams R, Alberts MJ, et al. Primary prevention of ischemic stroke: a guideline from the American Heart Association/ American Stroke Association Stroke Council: cosponsored by the Atherosclerotic Peripheral Vascular Disease Interdisciplinary Working Group; Cardiovascular Nursing Council; Clinical Cardiology Council; Nutrition, Physical Activity, and Metabolism Council; and the Quality of Care and Outcomes Research Interdisciplinary Working Group: The American Academy of Neurology affirms the value of this guideline. Stroke. 2006;37: 1583–1633. 39. Mazzati E, Vander Hoorn S, Lopez AD, et al. Comparative quantification of mortality and burden of disease attributable to selected risk factors. In: AD Lopez, CD Mathers, M Ezzati, DT Jamison, CJL Murray, eds. Global Burden of Disease and Risk Factors. Washington, DC: The International Bank for Reconstruction and Development/The World Bank; 2006:241–396. 40. Casper M, Wing S, Strogatz D. Variation in the magnitude of Black-White differences in stroke
46894_CH05_089_110.qxd
110
2/25/10
3:25 PM
Page 110
CHAPTER 5 STROKE
mortality by community occupational structure. Epidemiol Community Health. 1991;45: 302–306. 41. Truelsen T, Mahonen M, Tolonen H, Asplund K, Bonita R, Vanuzzo D. Trends in stroke and coronary heart disease in the WHO MONICA Project. Stroke. June 2003;34(6):1346–1352. 42. Tolonen H, Mahonen M, Asplund K, et al. Do trends in population levels of blood pressure and other cardiovascular risk factors explain trends in stroke event rates? Comparisons of 15 populations in 9 countries within the WHO MONICA Stroke Project. World Health Organization Monitoring of Trends and Determinants in Cardiovascular Disease. Stroke. 2002;33(10):2367–2375. 43. Sarti C, Stegmayr B, Tolonen H, Mahonen M, Tuomilehto J, Asplund K. Are changes in mortality from stroke caused by changes in stroke
event rates or case fatality? Results from the WHO MONICA Project. Stroke. 2003;34(8): 1833–1840. 44. Sarti C, Rastenyte D, Cepaitis Z, Tuomilehto J. International trends in mortality from stroke, 1968 to 1994. Stroke. 2000;31(7):1588–1601. 45. Liu L, Ikeda K, Yamori Y. Changes in stroke mortality rates for 1950 to 1997. A great slowdown of decline trend in Japan. Stroke. 2001; 32:1745–1749. 46. Editorial. STEPS in the right direction. Lancet Neurol. 2007;6:93. 47. Strong K, Mathers C, Bonita R. Preventing stroke: saving lives around the world. Lancet Neurol. 2007;6:182–187.
46894_CH06_111_138.qxd
2/25/10
3:26 PM
Page 111
C H A P T E R
6 Related Conditions nificant contributions to the overall burden of cardiovascular diseases—whether in terms of mortality, hospitalizations, or physician office visits, for example—or due to other consequences.
SUMMARY In addition to coronary heart disease and stroke, five other conditions contribute substantially to the overall burden of cardiovascular diseases. Two of these conditions are manifestations of atherosclerosis in regions of the circulation other than the heart and brain: peripheral arterial disease, in which the blood supply of the lower extremities is compromised, and aortic aneurysm, or localized dilatation with potential for rupture of the arterial trunk connecting the heart with the peripheral circulation. The third condition is chronic heart failure, or impairment of the fundamental function of the heart as a pump and prime mover of the circulatory system. Fourth, deep vein thrombosis is a condition affecting the venous side of the circulation with local occlusion of blood flow, especially in the lower extremities and pelvic veins. It has the potential to cause the grave complication of pulmonary embolism, in which a thrombus or fragment dislodges from its point of origin and is carried through the venous circulation to obstruct blood flow in the lung. Fifth, cardiac arrhythmias, or disturbances of heart rhythm, reflect dysfunction of electrophysiologic control of the rate and rhythm of the cardiac cycle. One very serious consequence is disturbance of blood flow through the left atrium of the heart due to atrial fibrillation, promoting formation and dislodging of thrombi that can be carried through the circulation to the brain and result in a thromboembolic/occlusive stroke. Another is an increased rate of ventricular contraction (ventricular tachycardia or fibrillation), with loss of effective pumping action of the heart, potentially leading to cardiac arrest and sudden cardiac death. The importance of these conditions lies in their sig-
INTRODUCTION Four of the five major conditions addressed here are represented in Table 6-1, based on the International Classification of Diseases and Related Health Conditions, Tenth Revision (ICD 10).1 “Peripheral arterial disease” (PAD) appears as a subclass of diseases of the arteries, arterioles, and capillaries (I70-I79). It is likely to be coded nearly always as I70.2, although some cases might be described only as “peripheral vascular disease” or “intermittent claudication” (a condition characterized by pain in the calf muscles during walking) and require coding as I73.9. A single broad category, aortic aneurysm and dissection (I71), includes subclasses according to location of the aneurysm, whether in the abdominal or thoracic portion or both, and whether there is mention of rupture of the aneurysm. The abdominal aortic aneurysms (AAA) alone will be addressed here. “Congestive heart failure,” by contrast, is coded in any of several different categories, depending on the underlying condition. The term refers to left ventricular failure specifically. It includes right ventricular failure if this has resulted from left ventricular failure. This occurs when a sequence of events develops gradually: poor emptying of the left ventricle into the arterial circulation; backup of blood flow from the lungs to the left ventricle; back-pressure on the right ventricle due to passive congestion of the lungs; and failure of the right ventricle. “Deep vein thrombosis”
111
46894_CH06_111_138.qxd
112
2/25/10
3:26 PM
Page 112
CHAPTER 6 RELATED CONDITIONS
Table 6-1
Classification of Peripheral Arterial Disease, Abdominal Aortic Aneurysm, Heart Failure, and Deep Vein Thrombosis and Pulmonary Embolism
a. Peripheral arterial disease I70 Atherosclerosis I70.2 Atherosclerosis of arteries of extremities I73 Other peripheral vascular disease I73.9 Peripheral vascular disease, unspecified b. Aortic aneurysm I71 Aortic aneurysm and dissection I71.3 Abdominal aortic aneurysm, ruptured I71.4 Abdominal aortic aneurysm, without mention of rupture c. Congestive heart failure I50 Heart failure I50.0 Congestive heart failure I50.1 Left ventricular failure I50.9 Heart failure, unspecified I09 Other rheumatic heart disease I09.9 Rheumatic heart disease, unspecified I11 Hypertensive heart disease I11.0 Hypertensive heart disease with (congestive) heart failure I13 Hypertensive renal disease I13.0 Hypertensive heart and renal disease with (congestive) heart failure I13.2 Hypertensive heart and renal disease with both (congestive) heart failure and renal failure d. Deep vein thrombosis and pulmonary embolism I80 Phlebitis and thrombophlebitis I80.1 Phlebitis and thrombophlebitis of femoral vein I80.2 Phlebitis and thrombophlebitis of other deep vessels of lower extremities I80.3 Phlebitis and thrombophlebitis of lower extremities, unspecified I80.8 Phlebitis and thrombophlebitis of other sites I80.9 Phlebitis and thrombophlebitis of unspecified site I26 Pulmonary embolism I26.0 Pulmonary embolism with mention of acute cor pulmonale I26.9 Pulmonary embolism without mention of acute cor pulmonale Source: Data from International Statistical Classification of Diseases and Related Health Problems: Tenth Revision. World Health Organization © 1992. World Health Organization, Geneva, Switzerland.
(DVT, clotting of blood in the larger veins, especially in the lower extremities or pelvis) is especially common among hospitalized patients confined to bed. It is often complicated by transport or embolization of a portion of the thrombus through the venous system and right side of the heart to reach the lung, causing acute respiratory compromise and sometimes death (pulmonary embolism, PE). ICD 10 provides for coding of the initial thrombosis by its venous location and of PE by the presence or absence of mention of acute failure of the right ventricle of the heart (cor pulmonale). Cardiac arrhythmias, the fifth group of cardiac conditions of interest, are classified in six separate categories in ICD 10 (I44-I49). They constitute a heterogeneous group, of which two are addressed here, atrial fibrillation and ventricular arrhythmias. As a measure of the importance of these five conditions, several indicators of their occurrence are pre-
sented in Table 6-2, based on US data for 2003 and 2004.2 The numbers of hospital discharges for which each condition was the first-listed diagnosis, the number of physician office visits, and the number of deaths attributed to each condition are shown. To provide some perspective on these numbers, the corresponding figures are also given for cerebrovascular diseases (mainly stroke) and coronary heart disease. PAD (atherosclerosis of arteries) greatly exceeded aortic aneurysm (abdominal and thoracic together) as a cause of hospitalization but was similar in numbers of deaths and represented only a small fraction of cardiovascular diseases relative to coronary heart disease. Physician office visits were about 50% more frequent for aortic aneurysm than for atherosclerosis of arteries. In contrast, heart failure (of which nearly all cases are specifically coded as congestive heart failure) was about half as frequent a hospital dis-
46894_CH06_111_138.qxd
2/25/10
3:26 PM
Page 113
PERIPHERAL ARTERIAL DISEASE 113
Table 6-2
Condition
Frequencies of Selected Conditions Relative to Cerebrovascular Diseases and Coronary Heart Disease, United States, 2003 and 2004 First-Listed Physician ICD 10 Hospital Discharge Office Visits Code(s) Diagnosis (2004) (2003) Deaths (2004)
Atherosclerosis of arteries Aortic aneurysm Heart failure Deep vein thrombosis Pulmonary embolism Cardiac arrhythmias Cerebrovascular diseases Coronary heart disease
I70 I71 I50 I80.2 I26 I48, Other I43–I49 I60–I69 I20–I25
123,000 61,000 1,099,000 7000 121,000 762,000 906,000 1,981,000
445,000 709,000 2,890,000 — 53,000 4,733,000 3,538,000 9,389,000
11,861 13,753 57,120 2843 8113 37,606 150,074 451,326
Source: From Morbidity & Mortality: 2007 Chart Book on Cardiovascular, Lung, and Blood Diseases. National Institutes of Health, National Heart, Lung and Blood Institute. June 2007.
charge diagnosis as coronary heart disease and led to nearly one-third as many physician office visits. The number of deaths coded as congestive heart failure in one year was a little more than one-tenth of that coded as coronary heart disease, perhaps reflecting diagnostic or coding practices that give preference to coronary heart disease when both conditions are present. Deep vein thrombosis is probably omitted often when pulmonary embolism occurs as a complication. Its frequency as a reason for physician office visits is unknown, and deaths attributed to it are probable miscoding of deaths due to pulmonary embolism. It may be more reliable to combine DVT and PE as a single category in view of these types of misclassification. Pulmonary embolism, on the other hand, appears to have been about as common as peripheral arterial atherosclerosis as a first hospital discharge diagnosis, an infrequent cause of office visits, and less common as a cause of death. There are striking differences between these indicators for deep vein thrombosis and pulmonary embolism since a decade earlier:3 hospitalizations for DVT decreased from 27,000 to 8000, whereas those for PE increased from 59,000 to 99,000. The increase in the combined frequency is consistent with the increase in numbers of deaths attributed to the two conditions. These distinct changes are likely a reflection of changes in diagnostic and treatment practices between 1993 and 2001–2002, such that hospitalizations for the two conditions have changed differentially and are therefore unreliable for understanding of trends. Other notable changes over the ten years were an increase of about 50% in hospital admissions and decrease of 25% in office visits for atherosclerosis of arteries, a decrease of about 50% in office visits for aortic aneurysm, and an increase of about 35% in office visits and deaths due to congestive heart failure.
Coronary heart disease far dominates the distribution of cardiovascular deaths, with stroke following and also far outweighing the other five conditions. Among all cardiovascular hospitalizations, however, heart failure and cardiac arrhythmias are on the same high level of frequency as cerebrovascular diseases (stroke), and coronary heart disease is more than twice as frequent a discharge diagnosis as stroke. Heart failure and cardiac arrhythmias far exceed stroke in numbers of office visits. They represent about 35% and 50% as many visits, respectively, as for coronary heart disease. Each of the five conditions addressed here adds importantly to the burden of cardiovascular diseases in the United States, whether in terms of one or more of these indicators or because of its serious prognostic implications, as discussed hereafter.
PERIPHERAL ARTERIAL DISEASE Typical Course Typically, an older person with recognized PAD, or atherosclerotic impairment of arterial blood flow to the lower extremities, may experience intermittent claudication, or pain in the calf muscles after walking a short distance (e.g., 100 yards or fewer) that is relieved temporarily by stopping. This condition may persist with little change for several years. But the presence of diagnosed PAD is often a marker for advanced coronary or cerebral atherosclerosis, which may lead to death within one to two years. The condition may also progress to the degree that vascular surgery is required to improve blood flow, or amputation may be required because adequate flow cannot be restored.4
46894_CH06_111_138.qxd
114
2/25/10
3:26 PM
Page 114
CHAPTER 6 RELATED CONDITIONS
Background The early epidemiology of PAD was the epidemiology of the symptom, intermittent claudication. Perception and reporting of pain by patients or study participants is highly subjective. Medical histories elicited by physicians or other health personnel are variable in quality. These circumstances make diagnosis of PAD by history unreliable and of very uncertain comparability between observers or between studies. For these reasons, and to support international comparative cardiovascular surveys, standardization was needed. Rose reported in 1962 on studies that led to development of a standard questionnaire to determine presence or absence of intermittent claudication as well as pain typical of myocardial infarction.5 The “Rose questionnaire” (or “London School of Hygiene questionnaire”) was a method which standardized use of the term “intermittent claudication,” if the questionnaire was properly administered. This method was used in many subsequent studies, and intermittent claudication was often incorporated as a component of prevalence of atherosclerosis in population surveys. Development of this questionnaire represented an important early advance in cardiovascular survey methods and is reflected in most of the knowledge of the epidemiology of PAD well into the 1980s. Limitations of this approach to detecting arterial pathology and estimating its prevalence could not be fully appreciated until newer techniques became available and were applied in population studies. Population Studies: Definition and Classification By the mid-1980s, noninvasive techniques to assess peripheral arterial blood flow had advanced to permit evaluating these approaches in comparison with interview/questionnaire methods. One technique was to calculate the ratio of blood pressure measured at each of several points between the thigh and the toes to blood pressure measured in the brachial artery of the arm (the “ankle-brachial index,” or “ABI”). In the absence of impairment of blood flow in the artery to the lower extremity, these measurements would be approximately equal, and the expected ratio would be 1. Obstruction of blood flow would reduce the blood pressure distal to an obstruction in the lower extremity, so the ratio would be less than 1. Several values of this ratio have been proposed as the criterion for detecting PAD. Ultrasound techniques permit measuring blood flow through specific vessels. Other indicators are also available, recently including magnetic resonance angiography, which has been evalu-
ated in the context of clinical use.6 Through these methods, vascular disease can be detected at an earlier stage, before intermittent claudication has developed. Disease of the large, or “major,” vessels to the lower extremity can be distinguished from that of small vessels. As a result, the focus has shifted to large-vessel PAD (LV-PAD) as assessed, for example, by combined ABI and ultrasound measurement of blood flow through one major artery, the posterior tibial artery in the calf. These methods have been compared with the history of intermittent claudication as determined by the Rose questionnaire.7 Physical examination to check the presence and character of the pulse at several points in the extremities has sometimes been included. A wide range of prevalence estimates results, depending on the indices included. Relative to the most extensive set of measurements available, history of intermittent claudication detected only 9.2% of cases identified by other means, and only one-half of those positive by history had demonstrable disturbances of vascular flow. Early studies of causes and outcomes of PAD, based on history of intermittent claudication, were evidently limited by inaccurate identification of true arterial disease.8 Table 6-3 presents results of an assessment of three classes of LV-PAD identified by two criteria advocated as of the mid-1990s: ABI was 0.8 or less and forward blood flow through the posterior tibial artery was 3 cm/sec or less.8 When all evidence of disease was taken as the reference standard, this two-test combination detected 89% of limbs with LV-PAD; 99% of limbs classified as negative by these tests were judged so by the full battery of tests; positive classification by the two tests was associated with 90% confirmation from the full battery; and negative classification was confirmed in 99% of the limbs tested by the full battery. Accuracy, defined as the numbers of true positives added with true negatives, divided by the total number of limbs, was 96%. These favorable findings for all LV-PAD were essentially the same after exclusion of cases with isolated posterior tibial artery disease and restriction to those with asymptomatic disease or those who had previous surgical treatment. Relative to the history of intermittent claudication, then, the combination of ABI and posterior tibial artery flow provided a far superior index of true arterial disease. Interest focuses, then, on studies evaluating PAD by these measures, even though specific criteria have differed somewhat among studies, and ABI alone has become the current standard.
46894_CH06_111_138.qxd
2/25/10
3:26 PM
Page 115
PERIPHERAL ARTERIAL DISEASE 115
Table 6-3
Case Group
Evaluation of Measures of Large-Vessel Peripheral Arterial Disease (LV-PAD) Among 484 Adults Originally Screened Between 1979 and 1981 Ankle-to-Arm Pressure Ratio 0.8 ⴙ Posterior Tibial Peak Forward Flow 3 cm/sec Positive Negative Predictive Predictive Sensitivity Specificity Value Value Accuracy No. No. No. No. No. % of Limbs % of Limbs % of Limbs % of Limbs % of Limbs
All LV-PAD LV-PAD with isolated posterior tibial cases excluded Asymptomatic LV-PAD cases only
89.0 82.4
81/91 42/51
99.0 99.0
867/876 864/876
90.0 82.4
81/90 42/51
99.0 99.0
867/876 867/876
98.0 98.1
948/967 909/927
88.1
59/67
98.9
823/832
86.8
59/68
99.0
823/831
98.1
882/899
Source: Reprinted with permission from HS Feigelson, American Journal of Epidemiology, Vol 140, No 6, p 531, © 1994.
Rates Prevalence of PAD, based separately on history of intermittent claudication or on ABI, was reported for men and women in 11 studies reviewed in 2003.9 Table 6-4, parts (a) and (b), presents results of these studies, all from the United States or Europe, published chiefly in the 1990s. Age groups studied include middle-aged adults in some but not all instances, and data were reported separately for women and men in some but not all studies. Three main observations are that prevalence was estimated to be several times higher by ABI than by history in every study but one; prevalence was greater in men than in women in the majority of studies where reported separately; and studies with age-specific results or including older age groups demonstrated higher prevalence with increasing age. The range of prevalence was from less than 1% to 12.7% by history and from 1.6% to 29.1% by ABI with a criterion of 0.90 in the majority of studies. The distribution of ABI was explored in a more recent report that focused on the concept of “borderline peripheral arterial disease” and the full range of ABI in the population.10 Data were taken from the US National Health and Nutrition Examination Survey, 1999–2002, for 4895 qualifying participants aged 40 years or older. Categories of ABI were defined as: PAD, 0.90; borderline PAD, 0.90–0.99; lownormal ABI, 1.00–1.09; and normal ABI, 1.10–1.29. Higher values were excluded as representing severe arterial rigidity. Distributions of these categories are presented by age (Figure 6-1) and by race/ethnicity and gender (Figure 6-2). The age gradient in prevalence suggested above is clearly shown here, with a fourfold increase from ages 40–59 to 60–74 and dou-
bling from 60–74 to 75 years. Prevalence of borderline PAD also increased with age, although less sharply. Non-Hispanic Blacks exceeded other groups in prevalence of less than normal ABI at each level, whereas non-Hispanic Whites and Mexican Americans were similar in distribution of ABI. Prevalence of less than normal levels of ABI was greater for women than for men at each level. A further general observation about prevalent PAD is its contribution to the spectrum of subclinical atherosclerosis and cardiovascular disease, as discussed in Chapter 3. The report of the Cardiovascular Health Study on the prevalence of subclinical disease indicated that undiagnosed PAD identified 16% of men and 19.9% of women who were free of known cardiovascular disease but were classified as having subclinical disease.11 These cases were identified chiefly by ABI without blood flow measurements. Therefore, they represent an underestimate of the true prevalence of subclinical disease. Rose questionnaire responses identified few cases not detected by ABI. Risks The prognostic importance of PAD was demonstrated nearly 2 decades ago from the mortality experience of persons followed for up to 10 years from diagnosis, as examined by various subclasses in Table 6-5.12 The overall results, not shown in the table, are as follows: Among 569 study participants available for analysis, 67 had been found to have LV-PAD at baseline. There were 32 deaths in this group, constituting 61.8% of the 34 affected men and 33.3% of the 33 affected women. Corresponding frequencies among those without PAD at baseline were 16.9% and
46894_CH06_111_138.qxd
116
2/25/10
3:26 PM
Page 116
CHAPTER 6 RELATED CONDITIONS
Table 6-4 Location, Study Type, Author
Prevalence of Peripheral Arterial Disease (PAD) by (a) Intermittent Claudication and (b) Ankle-Brachial Index (ABI) Prevalence (%) (b) ABI Prevalence (%) (a) Year n Ages Men Women cutoff Men Women 12.7b,c
12.7b,c
0.90d
29.1b
29.1b
55
2.2
1.2
0.90
16.9
20.5
15,792
45–64
1.0
1.0
0.90
3.0
3.0
1996
3171
45–54 55–64 65–74
0.5 0.9 2.4
0.3 2.2 3.7
0.95 0.95 0.95
1.6 7.7 16.4
3.1 6.8 11.2
Cardiovascular Health Study, four US communities, Newman et al.
1993
5084
65
2.0b
2.0b
0.90
13.8
11.4
Pittsburgh, CrossSectional, Vogt et al.
1993
1491
65–93
—
7.4
0.90
SHEPf study, sub-group, Newman et al.
1993
1775
65
6.4b
6.4b
0.90
California, CrossSectional, Criqui et al.
1992
613
38–82
2.2
1.7
Edinburgh, Cohort Study, Fowkes et al.
1991
1592
55–74
4.5b
4.5b
Jerusalem, CrossSectional, Gofin et al.
1987
1592
35–64
1.3
Denmark, CrossSectional, Schroll & Munck
1981
661
60
5.8
PARTNERSa, 320 primary care practices in US, Hirsch et al.
2001
6417
70 or 50–69 (smokers and/or diabetic)
Rotterdam, CrossSectional, Meijer et al.
1998
7715
ARICe study, four US communities Zheng et al.
1997
Limburg, CrossSectional, Stoffers et al.
5.5
—
26.7b
26.7b
11.7
11.7
0.90
18.3b
18.3b
1.8
0.90
4.2
5.4
1.3
0.90
16
13
testsg
a
Peripheral arterial disease Awareness, Risk and Treatment: NEw Resources for Survival Program. Prevalence in the total population; no separate estimates for gender were reported. The PARTNERS program gave data for “chart history of claudication;” Rose claudication numbers were not given, though they do state “charted claudication was much more common than classic (questionnaire-based) Rose claudication.”71 d The PARTNERS program defined PAD as either an ABI 0.09, a previous history of PAD, or prior limb revascularization. e Atherosclerosis Risk In Communities Study. f Systolic Hypertension in the Elderly Program. g Criqui et al. used a different approach to assess the prevalence of PAD; the standared ABI was not used, but rather 4 different noninvasive tests (segmental blood pressure, flow velocity by doppler ultrasound, post-occlusive reactive hyperemia, and pulse reappearance half-time) were used to diagnose PAD. b c
Source: Reprinted with permission from Journal of Epidemiology, Vol 13, Higgins JP and Higgins JA, © 2003, pp 5–6.
46894_CH06_111_138.qxd
2/25/10
3:26 PM
Page 117
PERIPHERAL ARTERIAL DISEASE 117
70
PAD† Borderline PAD‡ Low-Normal ABI§ Normal ABI||
61.9* 60
Prevalence (%)
50 44.9* 40 34.0*
30.8
30
28.8
26.5
20
17.5 16.5
12.3 10
6.0
8.1
1.9 0 40–59
60–74 Age (Years)
751
Percentages do not equal 100% because 3.7% of the population was excluded due to severe arterial rigidity (ABI 1.3). *p trend 0.001. †ABI 0.90; ‡ABI 0.90 to 0.99; §ABI 1.00 to 1.09; ||ABI 1.10 to 1.29.
Figure 6-1 Prevalence of ABI Categories by Age Group. Source: Reprinted from American Journal of Cardiology, Vol 98 (I9), Menke A, Muntner P, Wildman RP, Dreisbach AW, Raggi P, Relation of Borderline Peripheral Arterial Disease to Cardiovascular Disease Risk, page 1227, © 2006, with permission from Elsevier.
11.6%, respectively. Adjustment for differences among groups for age, sex, and other cardiovascular risk factors resulted in relative risk estimates of 3.1 (95% confidence interval, 1.9–4.9) for death from all causes, 5.9 (3.0–11.4) for all cardiovascular deaths, and 6.6 (2.9–14.9) for coronary heart disease death. The detailed subgroup analysis shown in the table indicates greater relative risks for bilateral, symptomatic, and severe disease. A relative risk of 15 was reported for cardiovascular or coronary heart disease death among those with severe, symptomatic disease versus those free of disease. A recent analysis evaluated the contribution of ABI to prediction of 10-year cardiovascular outcomes in men and women beyond those resulting from the Framingham risk score alone.13 It was found that ABI 0.90 approximately doubled the predicted total and cardiovascular mortality and major coronary event rate. Its inclusion in the score would have the effect of modifying treatment recommendations for approximately 1 in 5 men and 1 in 3 women. These findings underscore the importance of PAD in evaluating risk of future cardiovascular outcomes. The cross-sectional relation of PAD to the major risk factors for atherosclerosis and coronary heart disease was compared between two methods of classification, the Rose questionnaire and ABI.14 The results for ABI included findings of less disease among women, taller subjects, and those with greater HDLcholesterol concentrations. More extensive disease
was present in relation to greater non-HDL cholesterol and triglyceride concentrations, combined measures of glucose intolerance or diabetes mellitus, and cigarette smoking. The results were not strikingly different between methods of classifying PAD, except that ABI was the more sensitive indicator for evaluating cross-sectional associations with sex, height, and glucose tolerance and diabetes. From the study of borderline PAD cited above, additional information is provided regarding crosssectional relationships between several risk factors and ABI.10 In Table 6-6, the four categories of ABI defined previously are compared. A trend test based on regression analysis is applied to the distribution of each factor across the four categories of ABI. For example, mean age for “PAD present” was 59.3 years and was 58.1, 57.6, and 56.9 years for successively more favorable ABI, a statistically significant trend. Similarly, the proportion of the “PAD present” group who were men was 39.9%, while it increased across categories to be 54.7% of the “normal” category, also significant. Significant trends indicated association with less favorable classification by ABI for older age, non-Hispanic Black race/ethnicity, less than high school education, physical inactivity, hypertension, diabetes mellitus, current smoking, abdominal obesity, chronic kidney disease, and elevated C-reactive protein. A similar trend for serum total cholesterol 240 mg/dl or greater was suggestive but not significant. Being non-Hispanic White was associated with more
46894_CH06_111_138.qxd
Page 118
CHAPTER 6 RELATED CONDITIONS
70 60 Prevalence (%)
3:26 PM
Non-Hispanic White Non-Hispanic Black Mexican American
50 36.8*
40 30
26.4
20 10
9.7* 7.0 5.4
56.2
55.8 37.2*
118
2/25/10
25.5
14.9* 8.3 8.3
0 PAD†
Borderline PAD‡
Low-Normal ABI§
Normal ABI||
(A) 70
Prevalence (%)
60
Men Women
61.0 47.4*
50 40
33.4*
30 21.8
20 11.7*
10
5.2 5.9
6.0
0 PAD†
Borderline PAD‡
Low-Normal ABI§
Normal ABI||
(B)
Figure 6-2 Age-Adjusted Prevalence of ABI Categories by (A) Race/Ethnicity and (B) Gender. *p 0.05. Source: Reprinted from American Journal of Cardiology, Vol 98 (I9), Menke A, Muntner P, Wildman RP, Dreisbach AW, Raggi P, Relation of Borderline Peripheral Arterial Disease to Cardiovascular Disease Risk, page 1227, © 2006, with permission from Elsevier.
favorable ABI. Being Mexican American also tended to be more favorable. The disadvantage of nonHispanic Blacks was the dominant racial/ethnic trend. In a US study of risk factors in relation to location of PAD in the affected vessel or vessels, it was observed that both smoking and elevated systolic blood pressure were related to stenosis of the aortoiliac or femoropopliteal arteries, but not of the tibioperoneal arteries.15 In the latter region, diabetes was more clearly related, but only among men. Mortality 5 and 10 years after diagnosis was from two to seven times greater for persons with aortoiliac or femoropopliteal lesions than for persons free of disease, and there was no significant increase in mortality for those with tibioperoneal lesions alone. In the study of borderline PAD, a further step in analysis was to calculate a multivariable risk score for probability of coronary heart disease within 10 years
as well as other outcomes.10 The several odds ratios were calculated after adjustment for age, race/ ethnicity, and gender, and trends for the odds ratios across categories of ABI were tested for statistical significance. Significant trends were found for a 20% or greater CHD risk in 10 years as well as CHD and stroke outcomes. After adjustment for confounding factors, the trends for stroke and CHD outcomes were not significant. The authors emphasized the importance of ABI assessment and of risk-factor intervention for those with low-normal ABI or borderline PAD. Recommendations regarding screening for PAD are debated after publication of a report by the United States Preventive Services Task Force in 2005, which recommended against PAD screening. They judged that little or no benefit results from this practice and that it could be harmful.16 A counterargument has since been published that proposes targeted screening to reach high-risk patients defined by age, other risk factors, symptoms, or abnormal peripheral pulses on physical examination. This position is consistent with previously published guidelines of the American College of Cardiology and American Heart Association.17 An analysis by Beckman and others projected mortality reductions after targeted screening under the assumption of prevalence of 29%; mortality rates from 27% to 57%, from previous studies; and benefits of intervention of 25% or 50%: In a targeted population in which 29% of screenees would be found positive, from 2 to 9 lives/100 population would be saved in 7 years through screening and treatment to reduce cardiovascular risk. Further supporting the latter view is a review focusing on the morbidity and mortality implications of PAD.18 Here the argument is advanced that PAD is strongly associated with risk of cardiovascular morbidity and mortality, by having a prognostic importance possibly stronger than prior myocardial infarction (see also the Framingham Heart Study report19); it is less often evaluated than other contributors to risk; when identified it leads less often to appropriate risk reduction intervention than presence of coronary artery disease; clinical trial evidence demonstrates the benefit of lipid and antiplatelet therapy in patients with PAD; and the increasing proportion of older persons in the population will add to the importance of addressing PAD effectively in the future. Current Issues The foregoing discussion underscores the question of whether differences regarding screening recommendations can be resolved and appropriate guide-
Source: Reprinted with permission from MH Criqui, The New England Journal of Medicine, Vol 326, No 6, p 384, © 1992, The Massachusetts Medical Society. All rights reserved.
Note: CVD, cardiovascular disease; CHD, coronary heart disease. Relative risks have been adjusted for age, sex, number of cigarettes smoked per day, systolic blood pressure, HDL cholesterol level, LDL cholesterol level, logarithm of the triglyceride level, fasting plasma glucose level, body mass index, and selection criterion.
CHD
Other LV-PAD (N ⴝ 33) 3.4 (1.9–6.0) 7.0 (3.2–14.9) 6.8 (2.7–17.5)
3:26 PM
Cause of Death All causes CVD
Relative Risk of Death Among Subjects with Various Categories of Large-Vessel Peripheral Arterial Disease (LV-PAD) Relative Risk (95% Confidence Interval) Isolated Posterior Unilateral Bilateral Asymptomatic Symptomatic Moderate Severe Tibial LV-PAD LV-PAD LV-PAD LV-PAD LV-PAD LV-PAD LV-PAD (N ⴝ 34) (N ⴝ 30) (N ⴝ 49) (N ⴝ 18) (N ⴝ 49) (N ⴝ 18) (N ⴝ 31) 3.3 2.9 2.7 4.7 2.8 3.9 2.9 (1.9–5.9) (1.5–5.5) (1.6–4.5) (2.3–9.6) (1.6–4.8) (1.9–8.0) (1.6–5.4) 5.5 5.8 4.7 11.2 4.8 8.4 4.2 (2.5–12.1) (2.5–13.3) (2.3–9.8) (4.5–27.9) (2.3–10.3) (3.4–20.8) (1.7–10.4) 5.5 7.2 5.6 11.4 5.6 8.9 5.5 (2.0–15.2) (2.6–19.7) (2.3–13.5) (3.6–35.8) (2.2–14.2) (3.0–26.8) (1.8–16.7)
2/25/10
Table 6-5
46894_CH06_111_138.qxd Page 119
PERIPHERAL ARTERIAL DISEASE 119
46894_CH06_111_138.qxd
120
2/25/10
3:26 PM
Page 120
CHAPTER 6 RELATED CONDITIONS
Table 6-6
Variable
Age-Adjusted Prevalence of Cardiovascular Disease (CVD) Risk Factors by Ankle-Brachial Index (ABI) Category PAD ABI Present Borderline Low-Normal Normal (ABI 0.90) (ABI 0.90–0.99) (ABI 1.00–1.09) (ABI 1.10–1.29) (n ⴝ 368) (n ⴝ 524) (n ⴝ 1455) (n ⴝ 2548) p Trend
Age (yrs) Men Non-Hispanic White Non-Hispanic Black Mexican American Less than high school education Physically inactive Hypertension Serum total cholesterol 240 mg/dl Diabetes mellitus Current smoking Abdominal obesity Chronic kidney disease Elevated C-reactive protein
59.3 (0.7) 39.9 (6.8%) 77.8 (2.9%) 17.4 (3.3%) 3.2 (1.2%) 37.7 (4.9%)
58.1 (0.4) 31.3 (2.8%) 70.6 (3.0%) 16.4 (2.9%) 3.8 (0.9%) 28.8 (2.7%)
57.6 (0.2) 37.7 (1.1%) 73.9 (2.2%) 12.1 (1.6%) 4.0 (0.7%) 24.5 (1.4%)
56.9 (0.1) 54.7 (1.3%) 80.9 (1.7%) 6.2 (0.8%) 4.6 (0.7%) 19.7 (1.4%)
0.001 0.001 0.001 0.001 0.065 0.001
55.6 (4.9%) 58.4 (5.6%) 25.7 (5.9%)
42.5 (2.8%) 50.6 (3.3%) 25.8 (2.7%)
42.3 (1.7%) 49.2 (1.1%) 22.4 (1.5%)
34.1 (2.0%) 38.3 (1.2%) 19.8 (1.6%)
0.001 0.001 0.119
13.2 (2.1%) 38.3 (3.5%) 62.8 (5.4%) 14.3 (2.6%) 22.1 (5.3%)
10.3 (1.3%) 29.2 (3.6%) 62.6 (2.4%) 8.2 (1.2%) 13.9 (2.4%)
9.2 (0.8%) 24.8 (1.1%) 54.9 (1.5%) 6.1 (0.6%) 11.2 (1.0%)
9.5 (0.6%) 15.8 (1.0%) 53.7 (1.8%) 6.2 (0.7%) 8.2 (0.7%)
0.003 0.001 0.037 0.001 0.001
Data are presented as means or percentages (SE). Source: Reprinted from American Journal of Cardiology, Vol 98 (I9), Menke A, Muntner P, Wildman RP, Dreisbach AW, Raggi P, Relation of Borderline Peripheral Arterial Disease to Cardiovascular Disease Risk, page 1228, © 2006, with permission from Elsevier.
lines then implemented to reduce risks of severe and fatal cardiovascular outcomes associated with PAD and suboptimal ABI. Prevention of PAD itself has been a less prominent topic and warrants consideration in parallel with atherosclerosis in other regions of the circulation. Could wider use of ABI as a screening tool serve this purpose, building on the evidence presented previously on the gradient of ABI? Is it a reasonable inference that “low normal ABI” represents a stage at which progressive disturbance of lower extremity circulation might be arrested or reversed by effective intervention?
AORTIC ANEURYSM Unlike PAD, aortic aneurysm may be directly fatal as a consequence of rupture or complications of attempted surgical repair. Its often rapid course may explain the data in Table 6-2 showing equal contributions from these two conditions to numbers of cardiovascular deaths, although aortic aneurysm leads to only half as many hospitalizations or physician office visits. Until about 1950, aortic aneurysm in the United States was predominantly due to
syphilis and was most often located in the thoracic aorta. A transition in factors causing aortic disease led to dominance of atherosclerosis as the underlying process and more common appearance of abdominal aneurysms, consistent with the anatomic distribution of this disease. Pathology of atherosclerosis in the abdominal aorta was discussed in Chapter 3, and discussion here also focuses specifically on abdominal aortic aneurysm (AAA). Typical Course An aneurysm may first be recognized when it causes pain in the abdomen or lower back in an adult aged 60 years or older and leads to physical examination and X-ray. More often, detection is incidental and precedes any symptoms, a result of routine palpation of the abdomen during physical examination by an examiner sensitized to the possibility of such an unanticipated finding. The typical course involves progressive enlargement of the aneurysmal mass, with mounting risk of rupture because increasing diameter of the aneurysm increases tension on the aortic wall. The normal diameter of the abdominal aorta is 2.5 cm. Aneurysms of less than 6 cm in diameter have a 15–20% risk of rupture within 10 years. If the aneurysm is not surgically repaired when
46894_CH06_111_138.qxd
2/25/10
3:26 PM
Page 121
AORTIC ANEURYSM 121
a diameter of about 6 cm has been attained, the 10year risk of rupture increases sharply, to 45–50%. Surgery carries significant risks even when done as an elective procedure, with 5–10% mortality. But rupture followed by emergency surgery may be fatal in half of the cases.20 In the absence of adequate facilities, the risk of death is 100%. Background An early investigation of the epidemiology of aortic aneurysm was based on mortality statistics for the United States from 1951 to 1981.21 These early data included thoracic as well as abdominal aneurysms. They also included “dissecting aneurysms,” a different disease now classified separately. Age-adjusted mortality due to aortic aneurysm increased from 1951 to a peak in about 1970, followed by a decline to 1981, for White and Black males and females. Over this period, persons born in more recent years were more likely than those born earlier to die of aortic aneurysm on attaining any given age; that is, a cohort effect was evident, with increased risk among more recent cohorts. Abdominal aneurysm rates, analyzed separately from 1968 to 1981, were several times higher than thoracic aneurysm rates and remained fairly constant. This contrasted with the trend in mortality from coronary heart disease, whose sharp decrease in the United States began in the mid- or late 1960s. Typical of the race–sex differences in mortality from abdominal aneurysm over this period were the rates for 1981, the latest year in that study: for White males, 4.97/100,000; for nonWhite males, 1.49/100,000; for White females, 0.91/100,000; and for non-White females, 0.64/ 100,000. Deaths from unspecified types of aneurysms were relatively more frequent than those for abdominal aneurysm in some groups. Population Studies: Definition and Classification Data on AAA depend on physical examination, noninvasive imaging techniques or contrast aortography, and statistical coding of deaths and hospital discharges by use of the International Classification of Diseases. Because methods differ among studies, results must be considered carefully. On physical examination, a characteristically pulsating mass may be felt in the midabdomen, and on X-ray the image of a calcified band may appear that marks a zone of advanced atherosclerosis in the wall of the aorta. More detailed evaluation is possible by ultrasound examination, computed tomography, or arteriography.
Rates The community of Rochester/Olmsted County, Minnesota, has an integrated medical records system for inpatient and outpatient experience of the total population. Incidence of both abdominal and thoracic aortic aneurysm was studied there over the period from 1951 to 1980.22 The diagnosis was accepted on the basis of the medical record if made by physical examination and confirmed by a second physician; if the radiologist made a firm diagnosis by ordinary X-ray (KUB, or kidney-ureter-bladder, X-ray); or if it was based on ultrasonography, aortography, surgery, or autopsy. Total incidence increased over this period, despite a continual decrease in incidence of thoracic aneurysm, because of changing incidence of abdominal aneurysm. For investigating the possibility that the increased incidence of abdominal aneurysm was due to greater frequency of smaller, previously less detectable lesions, the trends were examined specifically for small, medium, and large aneurysms. The greatest increase was for small lesions, but medium and large ones increased in incidence also. This issue was further evaluated by considering the basis for diagnosis. Increased incidence was apparent from each method from the 1950s to the 1960s. During the 1960s and 1970s, ultrasound examination became a major component of the diagnostic procedure and tended to displace physical examination and X-ray. Ultrasound contributed disproportionately more to detection of small lesions (less than 5 cm diameter) in contrast to large ones. Overall, the smaller, asymptomatic, and uncomplicated lesions became more readily detected. These observations indicate that data regarding occurrence of abdominal aneurysm should not be expected to be comparable over time or between settings, when methods of detection differ. On the basis of coded hospital discharges and deaths, investigators in England and Wales similarly observed increased incidence of abdominal aneurysms over the period from 1950 to 1984.23 The increase in mortality was 20-fold for men and 11-fold, beginning a decade in age later, for women. By 1981–1983, the numbers of hospital admissions had increased to nearly three times the numbers in 1968–1971. Over this interval the percentage of admissions that were emergencies decreased only slightly (from 63% to 56%) and case-fatality also diminished only slightly from 45% to 39%. Thus the marked rise in hospital admissions could not be explained by increased admission of cases milder than those admitted in the earlier period. It was noted that hospital data on deaths from AAA in the United Kingdom gave only limited insight because one-half to two-thirds of such deaths occurred out of the hospital.
46894_CH06_111_138.qxd
122
2/25/10
3:26 PM
Page 122
CHAPTER 6 RELATED CONDITIONS
Prevalence studies of AAA at autopsy have also shown increased occurrence of these lesions in recent decades, in Sweden, Australia, and Japan. From 1965 to 1989, in Japan rates of unsuspected abdominal aneurysm discovered at death in national autopsy surveys increased by 80% in men and by 50% in women.24 Only AAA became more frequent over this period, whereas thoracic aneurysm became much less frequent and the occurrence of dissecting aneurysm fluctuated with no overall change. What changes have occurred in the population as a whole, beyond the hospitalized population alone, remains unknown. Prevalence of AAA in the United Kingdom based on several surveys through the mid-1990s varied widely depending on age and the measurement criterion selected.25 In six of seven studies, only men were included, mainly age 65 years or older; several hundred to several thousand persons were screened; and the aortic diameter exceeded 40 mm in three instances, 29 or 30 mm in six, and 25 mm in one. Even with the same measurement criterion and similar age range, 29 mm and ages 65 to 75 years, prevalence varied from 1.3% to 8.4%. As noted above, comparability of studies is limited by differences in criteria, but even with apparently similar design studies present quite divergent results. Analogous data from several European and one American study indicated a range of estimates from 0.7% to 10.7%. Here diagnosis involves more than a single measurement, with dimensions compared between different segments of the aorta. These limitations leave open to question interpretation of any reported population differences or trends in measures of occurrence of AAA, as well as possible risk-factor associations. With that qualification, reported risk factors for this condition include male sex, smoking, hypertension, PAD, and cardiovascular disease. In contrast with coronary heart disease, differences in risk factors for this condition and differences in trends of occurrence raise the question of whether AAA is fundamentally an atherosclerotic condition. A more recent assessment of frequency in the United Kingdom indicates increasing hospitalization rates and mortality from AAA in men, and even greater increases in women, from 1979 to 1999.26 Case-fatality declined from about 25% to 10% for elective repair of nonruptured aneurysms and from about 70% to 50% for ruptured lesions. Risks On the basis of a cross-sectional survey of more than 5000 adults aged 55 years and over, investigators in the Rotterdam Study in the Netherlands identified cases of AAA by ultrasound examination.27 Aneurysm was diagnosed if the diameter of the most distant sec-
tion of the abdominal aorta was 35 mm or greater or if that diameter exceeded by 50% or more that of the most proximal section. Aneurysm was found in 91 of 2217 men and 21 of 3066 women examined. Prevalence increased sharply with age and was several times greater for men than women at every age, from 55–59 years to 80 years and older. Concurrent assessment (separately for men and women) of characteristics of the cases relative to noncases, or controls, indicated associations as shown in Table 6-7. Current smoking was the most striking associated factor in both men and women, with serum cholesterol concentration in men and past stroke in women being additional statistically significant associated factors. Aortic aneurysm has also been found to aggregate in families, as illustrated by a study of 91 first-degree relatives (parents and siblings but not, in this case, offspring) of cases identified in a regional hospital in Pittsburgh, Pennsylvania.28 Compared with the corresponding relatives of persons without AAA, the relative risks among fathers and mothers of cases were approximately 4, but with wide confidence limits that included 1. For siblings, however, the relative risks were large and the lower confidence bounds were greater than 1, being 9.9 (4.3–19.5) for brothers and 22.9 (8.4–50.0) for sisters. This is strong evidence of a familial component, but further investigation would be necessary to establish a meaningful pattern of inheritance. How much of this striking familial resemblance might reflect increased probability of detection due to diagnostic suspicion after the initial case is an unanswered question. A recent study of smoking and aortic aneurysms pursued the issue of whether this condition might be more specifically smoking-related than are other cardiovascular diseases.29 Ten studies were reviewed. Pooled estimates of relative risk were calculated for aortic aneurysm, coronary artery disease, cerebrovascular disease, chronic obstructive pulmonary disease (COPD), and lung cancer. Relative risk of aortic aneurysm among smokers was 2.5 times as great as that for coronary artery disease and 3.5 times as great as for cerebrovascular disease; it was 0.56 times the relative risk of COPD. The authors concluded that the findings were consistent with a nonatherosclerotic cause of aortic aneurysms. It is noteworthy, however, to recall discussion of atherosclerosis in Chapter 3 and evidence of its relation with smoking. Another recent report, from the Health Professionals Follow-up Study, addressed the relation of alcohol to AAA.30 After adjustment for other risk factors (smoking, hypertension, and body mass index), and with use of updated alcohol exposure from periodic questionnaires, alcohol consumption of two
46894_CH06_111_138.qxd
2/25/10
3:26 PM
Page 123
CHRONIC HEART FAILURE 123
Table 6-7
Potential Risk Factors in Men and Women with and Without an Aneurysm of the Abdominal Aorta, Adjusted for Differences in Age, Rotterdam Study, 1989–1993 Men Women Aneurysm of the Aneurysm of the Abdominal Aorta Abdominal Aorta Present Absent P Present Absent P Risk Factor (N ⴝ 91) (N ⴝ 2126) Value (N ⴝ 21) (N ⴝ 3066) Value Body mass index (kg/m2), mean 25.4 25.7 0.29 27.4 26.6 0.30 Systolic blood pressure (mm Hg), mean 142.0 138.6 0.14 142.8 139.5 0.48 Diastolic blood pressure (mm Hg), mean 76.5 74.7 0.14 75.5 73.5 0.41 Current smoking (%) 37.6 23.9 0.01 56.0 19.1 0.01 Serum cholesterol (mmol/l), mean 6.6 6.3 0.04 7.3 6.9 0.11 Serum HDL cholesterola (mmol/l), mean 1.2 1.2 0.53 1.4 1.5 0.32 Hypertension (%) 29.2 26.5 0.59 42.1 32.9 0.37 Stroke (%) 1.8 3.9 0.31 9.0 2.3 0.05 Diabetes mellitus (%) 8.6 10.4 0.61 0.0 9.4 — Intermittent claudication (%) 4.8 1.8 0.04 4.5 1.0 0.12 History of angina pectoris (%) 8.3 6.1 0.39 13.4 6.8 0.24 History of myocardial infarction (%) 15.7 11.0 0.17 8.7 3.3 0.37 a
HDL cholesterol, high-density lipoprotein cholesterol.
Source: Reprinted with permission from HJCM Pleumeekers, American Journal of Epidemiology, Vol 142, No 12, p 1297, © 1995, The Johns Hopkins University School of Hygiene and Public Health.
or more drinks per day was found to be associated with AAA (relative risk 1.65, 95% confidence interval: 1.03–2.64). How best to reconcile this information with other aspects of alcohol use in relation to cardiovascular diseases requires further consideration (see Chapter 15). As encountered in connection with PAD, the question arises of screening for this condition, in view of its risk of fatal complications. Subsequent to an earlier inconclusive assessment by the US Preventive Services Task Force in 1996, four trials of screening among men age 65 or older by ultrasound imaging were reported. The new studies were reviewed in an evidence synthesis that concluded by recommending screening for men aged 65 to 75 years.31 It was noted that although screening would reduce mortality from this condition by 43%, harms of treatment include an operative mortality rate of 2–6%, as well as risks of myocardial infarction, respiratory and renal failure, and others. Insufficient information was available on screening for women to support a recommendation. Questions regarding surgical repair versus surveillance of aneurysms in the 30–54 mm range and harms associated with these were not addressed in this report. Current Issues Intervention when an AAA of critical size is identified reduces the risk of rupture and death by less than half, and intervention itself has substantial other risks.
At or above the threshold of 55 mm, surgical intervention is considered to have a favorable balance of risk against no treatment. Rupture can occur with fatal outcome in lesions of smaller size. Given that prevention is preferable, two appropriate questions are: How can development of AAA be prevented in the first place? And how can lesions now detectable by ultrasound while below the intervention threshold be arrested or reversed in their progression, thereby diminishing the risks of the advanced lesions and avoiding the risks of treatment?
CHRONIC HEART FAILURE “Chronic heart failure” or more simply, “heart failure,” is a condition that reflects impairment of the pumping function of the left ventricle of the heart, as described previously. As a result, blood flow from the left ventricle into the aorta and to the peripheral arterial circulation is reduced. In addition, failure to eject the blood from the left ventricle leads to increased back-pressure in the pulmonary circulation, with reduced blood flow through the lungs and exudation (seepage of fluid) from the blood to the tissue spaces in the lung, exacerbating the accompanying impairment of respiratory function. Mechanisms of heart failure include both “systolic dysfunction,” with a diminished proportion of blood contained in the
46894_CH06_111_138.qxd
124
2/25/10
3:26 PM
Page 124
CHAPTER 6 RELATED CONDITIONS
left ventricle being ejected with each contraction (diminished ejection fraction), and “diastolic dysfunction,” with incomplete relaxation of the left ventricle and therefore reduced volume of blood entering the left ventricle to be ejected with the next contraction. These are features of left heart failure; right ventricular failure can occur as well with adverse consequences for the pulmonary circulation, as a consequence of obesity or sleep-disordered breathing. Heart failure may result from any of several underlying processes, such as myocardial infarction, which causes significant localized damage to the ventricular wall; longstanding high blood pressure, inadequately controlled; cardiomyopathies such as Chagas disease, which results in more generalized loss of heart muscle cell function; or valvular heart disease, such as in chronic rheumatic heart disease, which can cause valvular obstruction to outflow from and leakage or regurgitation of blood flow back into the left ventricle.3 Typical Course Compensatory physiologic changes can maintain adequate left ventricular function, even while the underlying disease progresses. Once heart failure develops, its main clinical manifestations are generally similar regardless of its cause. However, in the case of coronary heart disease, as in acute myocardial infarction, the onset of heart failure may be very sudden. The presence of heart failure in the acute phase of myocardial infarction is a poor prognostic sign, and its effective treatment is important for immediate survival. Recovery may be complete, however, with no recurring signs or symptoms. Chronic heart failure may persist after an acute onset and partial recovery, or progressive ventricular decompensation may occur over weeks, months, or years. Fatigue and shortness of breath on minimal exertion are among the clinical indications of heart failure. Treatment may improve function and prolong life for several years, but progressive decompensation or other complications result in death in a large proportion of cases. In 2005, with an update in 2009, a new practice guideline for diagnosis and management of heart failure in adults was developed by the American College of Cardiology Foundation and American Heart Association (ACCF/AHA Guidelines).32,33 This guideline presented a schematic view of the progressive development of heart failure from being at risk to exhibiting symptomatic disease. Four stages were defined: A, “At high risk for HF [heart failure] but without structural heart disease or symptoms of HF”; B, “Structural heart disease but without signs or symptoms of HF”; C, “Structural heart disease with prior or current symptoms of HF”; D, “Refractory HF re-
quiring specialized interventions.”33 Approaches to therapeutic intervention were described for each stage. Background Epidemiologic investigation of heart failure has been impeded by some of the features just described. Its character as an end-stage development in the course of several distinct diseases has prevented its clear and consistent identification in mortality or hospital statistics. Gradual onset delays its recognition, so case incidence is difficult to define. Given its often prolonged clinical course, death due to an intervening coronary event or stroke may overshadow the presence of heart failure and lead to its omission from the diagnosis entered or coded on the death certificate. Alternatively, a death may be attributed simply to heart failure when a more specific disease could have been indicated. Clinical definitions and classification have also been limited by the nonspecific nature of the symptoms of heart failure, in contrast, for example, to the classical (if not universal) pain of myocardial infarction or intermittent claudication. A comprehensive overview of the Framingham Study experience, based on 32and 40-year follow-ups, addresses the main features of the epidemiology of heart failure and preventive approaches to reduce the mounting public health burden of this condition.34 Population Studies: Definition and Classification Epidemiologic studies of heart failure have required criteria by which to identify cases. A prominent example is the Framingham Study, whose 40-year follow-up was published in 1993 and included criteria for congestive heart failure as shown in Table 6-8.24 They take the form of a list of symptoms, physical findings, and physiological measurements grouped as “major” and “minor” criteria. Classification as a case required that at least two major or one major and two minor criteria be present, with no other medical explanation than heart failure for the presence of the minor criteria. (“Dyspnea” is shortness of breath; “edema” is swelling especially in the feet and ankles, due to fluid retention and reduced efficiency of venous return of blood to the heart; and “rales” are sounds heard by stethoscope on physical examination that indicate presence of fluid near the base of the lungs.) The 2005 ACCF/AHA Guidelines defined heart failure as “a clinical syndrome that is characterized by specific symptoms (dyspnea and fatigue) in the medical history and signs (edema, rales) on the physical examination. There is no single diagnostic test for HF because it is largely a clinical diagnosis that is based on a careful history and physical examination.”32, p 1828 The 2005 ACC/AHA report adopted the
46894_CH06_111_138.qxd
2/25/10
3:26 PM
Page 125
CHRONIC HEART FAILURE 125
Table 6-8
Criteria for Congestive Heart Failure, the Framingam Heart Study
Major Criteria Paroxysmal nocturnal dyspnea Neck vein distention Rates Radiographic cardiomegaly (increasing heart size on chest X-ray film) Acute pulmonary edema Third sound gallop Increased central venous pressure ( 16 cm water at the right atrium) Circulation time 25 seconds Hepatojugular reflux Pulmonary edema, visceral congestion, or cardiomegaly at autopsy Weight loss 4.5 kg in 5 days in response to treatment of CHF Minor Criteria Bilateral ankle edema Nocturnal cough Dyspnea on ordinary exertion Hepatomegaly Pleural effusion Decrease in vital capacity by 33% from maximal value recorded Tachycardia (rate 120 beat/min) Note: The diagnosis of congestive heart failure (CHF) required that two major and two minor criteria be present concurrently. Minor criteria were acceptable only if they could not be attributed to another medical condition. Source: Reprinted with permission from the American College of Cardiology, Journal of the American College of Cardiology, 1993, Vol 22, No 4, p 7A.
simpler term “heart failure” because not all cases present evidence of “congestion” at a given time. This term then included structural or functional cardiac disorders impairing the ability of the ventricle either to fill with or to eject an adequate volume of blood. The 2009 Update included discussion of certain biomarkers now considered useful in diagnostic evaluation of patients with possible heart failure, the natriuretic (sodium-excreting) peptides produced and released by the heart.33 These substances, BNP (B-type natriuretic peptide) and NT-proBNP (Nterminal pro-B-type natriuretic peptide), may be increased in concentration due to other factors than heart failure, however, and are therefore not specific to this condition. Also discussed is increasing use of the echocardiogram in detection and evaluation of heart failure. Several elements of the Framingham Study criteria as well as the ACCF/AHA definition depend on subjective and qualitative judgment. This approach is unlike case definition for PAD or AAA, for example, by measurement of ABI or aortic dimensions. Coding of heart failure as a hospital discharge diagnosis or cause of death also poses difficulties. This is because heart failure may be the result of any of several underlying conditions—hypertension, coronary heart disease, rheumatic heart disease, cardiomyopathies (heart muscle disorders), and others. Heart failure
may not always be indicated when appropriate as the first diagnosis or underlying cause of death. Kannel has estimated that four times the number of coded heart failure deaths include this condition as a contributing cause, indicating a major underestimate of the impact of heart failure in currently available mortality data.34 The extent of misclassification and admixture of cases would be expected to vary with the relative frequencies of the underlying diseases in different populations. Clearly, comparison of studies on heart failure between populations or over periods of decades could be misleading if diagnostic practices or variations in underlying disease frequencies were not taken into account, and to do so is difficult. Rates With this understanding of case identification, several sources of data can be highlighted regarding mortality, hospitalization, case-fatality, incidence, and prevalence. Mortality Interpretation of mortality data in heart failure is compromised by variation in methods for coding or analysis of heart failure, which may be listed as the underlying cause of death, a contributing cause, or an incidental condition unrelated to the cause of death.
46894_CH06_111_138.qxd
126
2/25/10
3:26 PM
Page 126
CHAPTER 6 RELATED CONDITIONS
This limitation should be borne in mind in considering such data as the following: Data compiled by the National Heart, Lung and Blood Institute for deaths due to heart failure in 2004 are shown in Figure 6-3 for four sex-race groups.2 The nearly 57,000 deaths shown in Table 6-2 are seen here as strongly related to age, with highest rates across the age span for Black males and lowest rates for White females. Black males and Black females both reach higher death rates from heart failure at earlier ages than Whites. Age-adjusted death rates for heart failure between ages 35 and 74 years for 15 countries are compared, separately for females and males, in Figure 6-4.2 The rates for the United States are intermediate between the extremes of Poland and Romania, and in all countries except Denmark, rates for males appear to exceed those for females. The extreme high rates for Poland may reflect differences in coding practices or real variation in frequency of the underlying causes of heart failure. A special aspect of mortality in heart failure is the high proportion of deaths that are sudden. The Framingham Heart Study reported that 40–50% of deaths in the presence of heart failure were sudden, defined as occurring within one hour in a previously stable patient.35 On the basis of 30 years of followup in the Framingham population, the presence of previous heart failure increased the risk of sudden death sevenfold when coronary heart disease was absent and nearly ninefold in the additional presence of coronary heart disease.
Another estimate of the impact of heart failure on subsequent mortality is based on long-term follow-up of participants in the first US National Health and Nutrition Examination Survey (NHANES I) of noninstitutionalized persons aged 25–74.36 Here the strategy was to identify cases of heart failure by criteria applied at the initial survey, conducted from 1971 to 1975, and to determine their subsequent mortality over a 15-year follow-up period. All deaths were included, irrespective of their cause according to the death certificate. Two methods were used for case definition at the baseline. The first method was selfreport, based on survey responses to the medical history questionnaire that asked participants whether a physician had ever told them they had heart failure. The second method used a clinical score based on an adaptation of the Framingham criteria, discussed previously, which took into account the relevant observations from the survey. For women aged 55–64 and 65–74 years, respectively, the 15-year postsurvey mortality rates were approximately 40% and 60% for those identified by self-report and 25% and 50% for those identified by clinical score. For men, the corresponding frequencies were 50% and 75% (selfreport) and 65% and 80% (clinical score) for age groups 55–64 and 65–74 years, respectively. Mortality was clearly very high in all age-sex groups, especially for men, by either method of case identification. Especially for the younger age group, mortality was higher for self-reported than for clinically identified heart failure at baseline for women, suggesting that
Deaths/100,000 Population
200
150
Black Male White Male Black Female White Female
100
50
0 35–44
45–54
55–64
65–74
75–84
Age (Years)
Figure 6-3 Death Rates for Heart Failure by Age, Race, and Sex, United States, 2004. Source: From Morbidity & Mortality: 2007 Chart Book on Cardiovascular, Lung, and Blood Diseases. National Institutes of Health, National Heart, Lung and Blood Institute. June 2007.
46894_CH06_111_138.qxd
2/25/10
3:26 PM
Page 127
CHRONIC HEART FAILURE 127
POL JPN GER SPA CZR NTH FRA (2003) DEN (2001) USA NOR (2003) HUN (2003) EW AUL (2003) SCO KOR SWE (2002) FIN ROM (2000)
Male Female
30
20
10
0 10 20 30 Deaths/100,000 Population
40
50
60
Figure 6-4 Age-Adjusted Death Rates* for Heart Failure by Country, and Sex, Ages 35–74, 2004†. Source: From Morbidity & Mortality: 2007 Chart Book on Cardiovascular, Lung, and Blood Diseases. National Institutes of Health, National Heart, Lung and Blood Institute. June 2007. *Age adjusted to European standard. † Data for 2004 unless otherwise noted in parentheses.
the score identified milder cases, but the reverse was true for men. Hospitalization As with mortality data, hospitalizations attributed to heart failure may represent different circumstances in data from different sources. Heart failure may be the primary reason for admission or only an ancillary condition; admission may reflect a case of unstable heart failure; or it may represent another illness, such as pneumonia, threatening effective management of otherwise stable heart failure. The 2009 ACCF/AHA Guidelines identified several factors that could lead to readmission of patients with heart failure that might be considered in interpreting such data.33 Given these qualifications, the available data can be discussed as follows. The rate of hospitalization for heart failure has increased more than threefold over the past three decades (Figure 6-5).2 This pattern is almost entirely due to cases aged 65 years or older, although the direction of the trend is seen, at much lower rates, in the 45–64 year age group as well. Heart failure is now the leading cause of hospitalization in the United States among the elderly. Among objectives for heart disease and stroke prevention in Healthy People 2010 is a
50% reduction in age-specific hospitalization rates for heart failure in each of three age groups, 65–74, 75–84, and 85 years and older.37 The specific targets would in effect reduce the rate in each group to the baseline rate for the next younger age group. Relative to the 1997 reference data, progress was insufficient at the midcourse review in 2007 (based on earlier data) to anticipate complete success by 2010. Hospital discharge data presented earlier by Gillum indicated the effect of choosing only the firstlisted diagnosis or all diagnoses to identify cases.38 On the basis of the first-listed discharge diagnosis alone, from 1986 to 1990 heart failure was responsible for 103,000 to 117,000 hospitalizations per year at ages 45–64 and 461,000 to 560,000 per year at ages 65 and older. The estimated frequency of such discharges was about two and one-half to three times greater when all diagnoses were identified rather than when using the first diagnosis alone. Undercounting of hospitalizations to which heart failure contributes is analogous to the undercounting of deaths based on underlying cause alone. Hospital case-fatality rates for heart failure have fallen by more than 50% during the 1980s and 1990s, for both age groups, 45–64 years and 65 and older (Figure 6-6).2 More common hospitalization of milder cases and more frequently repeated admissions for
46894_CH06_111_138.qxd
128
2/25/10
3:26 PM
Page 128
CHAPTER 6 RELATED CONDITIONS
250
Hospitalizations/10,000 Population
Ages 45–64 Years Ages $65 Years
200
150
100
50
0 1970
1975
1980
1985
1990
1995
2000
2005
Years
Figure 6-5 Hospitalization Rates for Congestive Heart Failure, Age 45–64 and 65 and Over, US, 1971–2004. Source: From Morbidity & Mortality: 2007 Chart Book on Cardiovascular, Lung, and Blood Diseases. National Institutes of Health, National Heart, Lung and Blood Institute. June 2007.
individual cases could explain both increasing rates and decreasing in-hospital case-fatality. To what extent this and other factors such as more effective treatment contribute to these trends is presently unclear. Data on trends in severity of cases at admission, numbers of persons rather than hospitalizations, and numbers of deaths occurring in and out of hospital
would be helpful to understand the reported changes in these measures. Incidence Incidence of heart failure is reported from the Framingham Heart Study (FHS), the Atherosclerosis Risk in Communities (ARIC) Study, and the Cardio-
14 Ages 45–64 Years* Ages $65 Years
Percent Discharged Dead
12 10 8 6 4 2 0 1980
1985
1990
1995
2000
2005
Years * Unreliable estimate for 1981
Figure 6-6 Hospital Case-Fatality Rates for Congestive Heart Failure, Ages 45–64 and 65 and Over, US, 1981–2004. Source: From Morbidity & Mortality: 2007 Chart Book on Cardiovascular, Lung, and Blood Diseases. National Institutes of Health, National Heart, Lung and Blood Institute. June 2007.
46894_CH06_111_138.qxd
2/25/10
3:26 PM
Page 129
CHRONIC HEART FAILURE 129
vascular Health Study (CHS) over the past one to two decades, in the Incidence & Prevalence 2006 Chart Book.39 Incidence approximately doubled with each decade of age from 65–74 to 85–94 years in the FHS, from 9.2 to 43.0 per 1000 person-years among men and from 4.7 to 30.7 per 1000-person years for women. Men had higher rates than women of the same race/ethnicity in all studies and age groups except in ARIC, where at age 65–74 the rate for Black women exceeded that for Black men (15.9 versus 15.2 cases per 1000 person-years). Prevalence Prevalence of heart failure in the United States as of 2006 is indicated in Table 1-4a and included 5.7 million persons, or 2.5% of adults aged 20 years or older, ranging from 1.4% of Mexican American females to 4.2% of Black females and males.40 The recent history of prevalence, from 1971–1974 through 1999–2004, is illustrated in Figure 6-7.2 Separately for Whites, Blacks, males, and females, estimates from successive National Health and Nutrition Examination Surveys indicate a major increase in prevalence at the midpoint of this period, between 1976–1980 and 1988–1994. Prevalence in the first three groups approximately doubled, with a somewhat lesser increase for females than for males. A slight decline in prevalence is apparent for the most recent period. The abrupt change in the midperiod is so striking as to suggest a change in case definition. Risks The predictors of heart failure based on 40 years of follow-up in the Framingham Heart Study are shown
for men in Figure 6-8 and for women in Figure 6-9.41 These figures show, separately by age groups 35–64 and 65–94 years, the relative risk of developing congestive heart failure in relation to cholesterol concentration, smoking, hypertension, diabetes, and presence of left ventricular hypertrophy (enlargement) as determined by electrocardiography (ECG-LVH). In addition, the age-adjusted incidence of heart failure in the presence or absence of each factor is shown. Inclusion of left ventricular hypertrophy as a risk factor is questionable, both because this finding is indicative of the disease process itself and because it appears to diminish the relative importance of the other predictors. Among the other factors, those most strongly related to the risk of heart failure were hypertension and diabetes, for both men and women at both age levels. A more recent presentation of the Framingham experience provides further insight into these results.34 Regarding ECG-LVH, when the hazard ratio was adjusted for age and the other major risk factors, the values were 2.2 for men and 2.9 for women, greatly reduced from the univariate estimates in the earlier report. Further evaluation of these relations to include prevalence of each factor showed ECG-LVH to occur in 4% of men and 3% of women, resulting in populationattributable fractions of 4% and 5%, respectively. High blood pressure (systolic or diastolic pressure 140 90 mm Hg) with hazard ratios of 2.1 and 3.4 and prevalence of 60% and 62% resulted in populationattributable fractions of 39% and 59% for men and women, respectively. For prior myocardial infarction, the population-attributable fractions were 34% for men and 13% for women. On the basis of this analy-
Percent of Population
4
3
1971–1974 1976–1980 1988–1994 1999–2004
2
1
0
White
Black
Male
Female
Figure 6-7 Prevalence (Age-Adjusted) of Congestive Heart Failure by Race and Sex, Ages 25–74, US, 1971–74 to 1999–2004. Source: From Morbidity & Mortality: 2007 Chart Book on Cardiovascular, Lung, and Blood Diseases. National Institutes of Health, National Heart, Lung and Blood Institute. June 2007.
46894_CH06_111_138.qxd
130
2/25/10
3:26 PM
Page 130
CHAPTER 6 RELATED CONDITIONS
Risk Factor Absent Risk Factor Present
Age 35–64 yrs
RR
Age 65–94 yrs
RR
1.2
Cholesterol
0.9
*1.5
Smoking
1.0
*4.0
Hypertension
1.9*
*4.4
Diabetes
2.0*
*14.9
ECG-LVH
4.9*
50 40 30 20 10 0 *P<0.0001 0 10 20 30 40 50 Age-adjusted incidence of CHF Age-adjusted incidence of CHF per 1,000 person years per 1,000 person years
Figure 6-8 Risk Factors for Congestive Heart Failure (CHF) in Men in the Framingham Study, 1948–1988. Relative risks (RR) for the development of heart failure in the presence of the specified risk factor are displayed at the margins; values with asterisks are significant at P 0.001. Cholesterol, serum cholesterol 6.2 mmol/l (240 mg/dl), ECG-LVH, electrocardiographic left ventricular hypertrophy. Source: Reprinted with permission from the American College of Cardiology, Journal of the American College of Cardiology, 1993, Vol 22, No 2, p 9A.
sis, it was concluded that a multivariate risk calculation identified in the highest quintile of the risk distribution was 70% of the expected future cases of heart failure. Trends Figure 6-10 indicates the trend of deaths coded as due to heart failure in the United States from 1979 to
2004.2 Apart from the change in cause-of-death coding instructions noted in 1989, the direction of the trend was clearly upward through the 1980s and increasing only slightly, if at all, in the 1990s. There is no evidence of a decrease in overall mortality from heart failure, despite the reduction by half in hospital case-fatality.
Risk Factor Absent Risk Factor Present
Age 35–64 yrs
Age 65–94 yrs
RR
RR
0.7
Cholesterol
0.8
1.1
Smoking
1.3
*3.0
Hypertension
1.9*
*7.7
Diabetes
3.6*
*12.8
ECG-LVH
5.4*
50
40
30
20
10
0
Age-adjusted incidence of CHF per 1,000 person years
*P<0.0001
0
10
20
30
40
50
Age-adjusted incidence of CHF per 1,000 person years
Figure 6-9 Risk Factors for Congestive Heart Failure (CHF) in Women in the Framingham Study, 1948–1988. Relative risks (RR) for the development of heart failure in the presence of the specified risk factor are displayed at the margins; values with asterisks are significant at P 0.001. Cholesterol, serum cholesterol 6.2 mmol/l (240 mg/dl), ECG-LVH, electrocardiographic left ventricular hypertrophy. Source: Reprinted with permission from the American College of Cardiology, Journal of the American College of Cardiology, 1993, Vol 22, No 2, p 9A.
46894_CH06_111_138.qxd
2/25/10
3:26 PM
Page 131
CHRONIC HEART FAILURE 131
Deaths/100,000 Population*
40 Black Male White Male Black Female White Female
30
ICD-9
ICD-10
20
10
0 1979
1984
1989
1994
1999
2004
Years * The break in the trend lines is intended to signal a discrepancy arising from the adoption of new cause-of-death coding instructions on death certificates in 1989.
Figure 6-10 Age-Adjusted Death Rates for Heart Failure by Race and Sex, US, 1979–2004. Source: From Morbidity & Mortality: 2007 Chart Book on Cardiovascular, Lung, and Blood Diseases. National Institutes of Health, National Heart, Lung and Blood Institute. June 2007.
Note on Right Ventricular Failure Distinct from left-sided heart failure and its consequences is right-sided heart failure resulting from disease of the lungs and designated as chronic cor pulmonale (pulmonary heart disease). Among the terms used in connection with this condition is chronic obstructive pulmonary disease (COPD), which might result especially from chronic bronchitis or emphysema. A major contributor is cigarette smoking, with air pollution, childhood respiratory tract infections, and occupational dust exposures among additional causes.42 This condition, like heart failure as discussed above, is difficult to study epidemiologically because standardization of definition, diagnostic criteria, and classification are lacking. Projections based on demographic considerations alone anticipate doubling the global frequency of death from COPD between 1985 and 2015; taking changes in exposures and long-term effects into account suggests more than three times the deaths from this cause in 2015 relative to 1985— more than 3 million deaths worldwide. Continuation of the tobacco epidemic would add to long-term future risks of right-sided heart failure as a consequence of COPD. Better means are needed to monitor this condition, including its natural history and responsiveness to preventive measures. Prevention An AHA Scientific Statement on prevention of heart failure appeared in 2008. Among its main conclusions were the following:43, p 2544
Identifying and preventing the well-recognized illnesses that lead to HF, including hypertension and coronary heart disease, should be paramount among the approaches to prevent HF. Aggressive implementation of evidence-based management of risk factors for coronary heart disease should be at the core of HF prevention strategies. Questions currently in need of attention include how to identify and treat patients with asymptomatic left ventricular systolic dysfunction (Stage B HF) and how to prevent its development. An earlier collection of reports, Primary Prevention of Heart Failure (2004), provides more extensive discussion of preventive approaches, chiefly from the clinical perspective.44 Current Issues Questions of immediate importance concern taking the burden of heart failure into account in setting public health priorities and understanding the meaning of the increase in hospital discharges for heart failure. How important is heart failure as a component of cardiovascular morbidity and mortality? Indicators of the frequency of heart failure are limited by the nature of the diagnosis and its occurrence as a comorbid condition, for example, with coronary heart disease and stroke. Identifying the true occurrence of heart failure would lead to improved estimates of both
46894_CH06_111_138.qxd
132
2/25/10
3:26 PM
Page 132
CHAPTER 6 RELATED CONDITIONS
mortality and hospitalization rates. Measures of disability and disparities, including health-related quality of life in persons living with heart failure, and differences in such indicators among population groups, would help to make the case for its prevention. Including these improved measures of the impact of heart failure as an outcome of coronary heart disease and hypertension would add importantly to calculations of the cost of these conditions and the rationale for increased investment in their prevention. How many hospitalizations for heart failure should there be? The Healthy People objectives call for reducing age-specific numbers of hospitalizations with heart failure as the first-listed diagnosis by 50% over the current decade.37 However, it is unclear at present what those numbers mean. Problems with hospital discharge data include their enumeration of hospitalizations rather than individuals, as well as susceptibility of discharge codes to influence by their reimbursement value and other changes in coding practices. Does an increase truly represent growing numbers of cases, or more frequent hospitalization per case, or both? How often should a person with heart failure be hospitalized? What proportion of cases may lack access to needed hospital care, such that improved access would be indicated by increasing numbers of admissions?
DEEP VEIN THROMBOSIS AND PULMONARY EMBOLISM As presented in Table 6-1, ICD 10 recognizes separately the occurrence of DVT in various sites (I80 and subclasses) and PE (I26), with or without cardiac complications (cor pulmonale). Frequencies of events related to these two conditions are also indicated separately in Table 6-2, showing numbers of hospital discharges, physician office visits, and deaths. More commonly in the epidemiologic literature these conditions are discussed together as “venous thromboembolism,” or “VTE.” VTE is then partitioned into its two components with the usual statement that two-thirds of persons with symptoms experience DVT alone, and one-third develop PE. Typical Course Typically, DVT is found in hospitalized patients who have been confined to bed, especially those who are immobilized to the greatest degree. The most apparent local complications include swelling and pain, often in the calf and below. The chief concern about DVT is the attendant risk of PE. This is a sometimes fatal complication resulting from venous transport
of a portion of a blood clot from the lower extremity or pelvic veins through the right heart to the lung, where blood flow becomes obstructed, pulmonary function is impaired, and acute right heart failure or cor pulmonale may be precipitated by increased resistance to outflow from the right ventricle. Although this progression is most readily observed in patients while in the hospital, it has been reported that in surgical patients the occurrence of PE was frequently more than two weeks after surgery; the median onset was at the 18th postoperative day. Taking into account events within 30 days of surgery increased the estimated frequency by 30%.45 The trend toward shorter periods of hospitalization after surgery, at least in the United States, raises the possibility that persons remaining relatively immobilized at home after discharge may be at risk of developing these complications without close observation and timely intervention if needed. Not only surgical patients but those in intensive care related to acute coronary events or stroke share a relatively high risk of these complications. Anecdotal reports of cases occurring after air travel of several hours’ duration add to concern about this problem. Background A recent review of the epidemiology of VTE cites a number of studies that have contributed to an expanding literature on this topic, including several studies in the United States.46 The Worcester DVT Study in Worcester, Massachusetts, reviewed hospital discharge records of some 400 cases in the mid-1980s. Investigators in Olmsted County, Minnesota, reviewed records of cases diagnosed from 1966 to 1990 that included autopsy-detected asymptomatic cases. Medicare hospital discharge data were used for estimating incidence of deep vein thrombosis and pulmonary embolism in the population 65 years of age and older, in the late 1980s. Two cohorts discussed earlier (ARIC and CHS) were joined to form the Longitudinal Investigation of Thromboembolism Etiology, addressing incidence, case-fatality, and recurrence rates for these conditions. Record linkage was established between hospital discharge data and vital statistics for the state of California to study racial/ethnic group differences, mainly during the 1990s. The findings of these and other studies were summarized as shown in Table 6-9.46 A common incidence estimate is 1/1000 per year in the general population. Incidence is strongly related to age, with a 10-fold increase in rates from ages in the 30s to the 70s. Little difference by gender has been observed. Certain ethnic groups—Hispanics and Pacific Islanders—appear to have exceptionally low incidence. Diagnoses made at
46894_CH06_111_138.qxd
2/25/10
3:26 PM
Page 133
DEEP VEIN THROMBOSIS AND PULMONARY EMBOLISM 133
Table 6-9
Summary of the Epidemiology of First-Time VTE
Variable Incidence in Total Population (Assuming 95% Caucasian)
≈ 70–113 cases/100,000/year
Finding
Age 25–35 years old 70–79 years old
Exponential increase in VTE with age, particularly after age 40 years ≈ 30 cases/100,000 persons ≈ 300–500 cases/100,000 persons
Gender
No convincing difference between men and women
Race/Ethnicity
2.5–4-fold lower risk of VTE in Asian-Pacific Islanders and Hispanics
Relative Incidence of PE vs DVT
Absent autopsy diagnosis: ≈ 33% PE; 66% DVT With autopsy: ≈ 55% PE, 45% DVT
Seasonal Variation
Possibly more common in winter and less common in summer
Risk Factors
≈ 25% to 50% “idiopathic” depending on exact definition ≈ 15%–25% associated with cancer; ≈ 20% following surgery (3 mo.)
Recurrent VTE
6-month incidence: ≈ 7%; higher rate in patients with cancer Recurrent PE more likely after PE than after DVT
Death After Treated VTE
30 day incidence ≈ 6% after incident DVT 30 day incidence ≈ 12% after PE Death strongly associated with cancer, age, and cardiovascular disease
Source: Reprinted with permission from Circulation, Vol 107, RH White, © 2003 American Heart Association.
autopsy may differ in proportionate assignment of cases to DVT or PE, with a substantially greater relative frequency of PE found at autopsy than by clinical examination. Cases may relate to underlying cancer or surgery, but 25–50% of them have no apparent immediate predisposing illness. Mortality within 30 days of an episode is approximately 6% for DVT and 12% for PE. Population Studies: Definition and Classification Studies of hospital-based patient populations have included investigation of diagnostic approaches to both DVT and PE. Diagnosis of these conditions in many cases remains difficult, and diagnostic practices are likely to differ widely among settings. Clinical history is a key aspect of diagnosis. Detection of thrombosis is more reliable in deep calf veins than in the more proximal veins of the lower extremity (a particular concern in patients undergoing hip surgery, for example).47 Imaging techniques such as infusion of radioiodine (I-125) labeled fibrinogen followed by scanning of the leg are reportedly sensitive to the former type of thrombus but not the latter. PE produces symptoms and signs of varying severity and may be difficult to detect clinically in mild cases.45 A recent review of the current state of diagnostic methods favors ultrasonography for DVT, although in the absence of the required technology
other methods must be used.48 PE is most often detected by use of spiral CT (computed tomography). Occurrence and recognition of these complications in the hospital do not ensure their inclusion in reported hospital discharge statistics or in death certificate data, especially if only the first-listed discharge diagnosis or underlying cause of death is coded and tabulated. This issue clouds interpretation and comparison of reported frequencies of these events between populations or over time. Rates True rates of DVT or PE are difficult to establish in any population. Goldhaber, for example, ascribed to these conditions “hundreds of thousands of hospitalizations annually in the U.S.,”45, p 1582 but this statement contrasts sharply with data in Table 6-2.2 Further, Goldhaber and Sors attributed 50,000 deaths per year in the United States to PE,49 a figure nearly five times that shown in Table 6-2, based on reported deaths for 2001–2002.2 Frequencies of occurrence of fatal postoperative PE in patients with general elective surgery, elective hip surgery, and emergency hip surgery have been reported to range from 0.1% to 0.8%, 0.3% to 1.7%, and 4% to 7%, respectively.47 These figures are given for patient groups not receiving preventive measures to protect against DVT,
46894_CH06_111_138.qxd
134
2/25/10
3:26 PM
Page 134
CHAPTER 6 RELATED CONDITIONS
however, so that in settings where standards of practice include currently recommended prophylaxis, lower frequencies would be expected. Incidence of VTE has been monitored over 30 years in Olmsted County, Minnesota.50 The rate declined by more than 50% quite abruptly after reaching a peak of about 250/100,000 in 1975, with a change in diagnostic practice being the apparent explanation. Introduction of venography and nuclear lung scanning made diagnosis more objective and apparently more specific. The rate has since been relatively stable at approximately 100/100,000 (or 1/1000) per year, consistent with the widely quoted general population frequency. Risks Factors that identify higher-risk individuals undergoing surgery are summarized in Table 6-10.51 VTE appears to be associated with a number of hereditary factors, listed in the first column of the table, for some of which specific gene mutations have been identified. All of the factors in this column relate to blood
coagulation, a topic further discussed in Chapter 15. The second column provides a quite mixed list of conditions identified with VTE taken from a 1990 report from the Prospective Investigation of Pulmonary Embolism (PIOPED). A more current listing of factors associated with mortality in pulmonary embolism adds anticardiolipin antibodies, mutations of cystathione beta-synthase or methylene tetrahydrofolate reductase (MTHFR), and high concentrations of Factors VIII and/or XI. The population significance of these associations is in several cases under investigation in the studies noted above. Current Issues Under what conditions other than hospitalization do DVT and PE occur, and what is the epidemiology of these events? Indications that the postdischarge period is one of high risk and recognition that hospital stays are often fewer days than in the past add importance to the period immediately following discharge for detection and effective treatment of cases. Conditions similar to hospital bed confinement, whether at home,
Table 6-10
Hereditary and Acquired Risk Factors for the Development of VTE Hereditary Risk Factors (Primary) Acquired Risk Factors (Secondary) Factor V Leiden mutation
Immobilization
Prothrombin gene mutation
Surgery within the last 3 months (especially major abdominal, pelvic and orthopedic surgery (hip, knee))
Protein S deficiency
Stroke, paralysis of extremities
Protein C deficiency
History of VTE
Antithrombin (ATIII) deficiency
Malignancy
Heparin cofactor II deficiency
Obesity
Plasminogen deficiency
Cigarette smoking
Factor XII deficiency
Hypertension
Dysfibrinogenemia
Oral contraception, hormone-replacement therapy
Increased factor VIII coagulant activity
Pregnancy and puerperium
Primary hyperhomocyst(e)inemia
Secondary hyperhomocyst(e)inemia Antiphospholipid syndromes (“lupus anticoagulant”) Congestive heart failure Myeloproliferative disorders (e.g., polycythemia vera, essential thrombocythemia) Nephrotic syndrome Inflammatory bowel disease Sickle cell anaemia Marked leukocytosis in acute leukemia
The table contents are based on data provided by the PIOPED study. Source: Reprinted with permission from Respiration, Vol 70, C Kroegel and A Ressig. © 2003 S. Karger AG.
46894_CH06_111_138.qxd
2/25/10
3:26 PM
Page 135
ARRHYTHMIAS 135
in nursing care institutions, or elsewhere would seem to constitute equally likely environments for development of VTE. Brief reference is encountered to risks associated with long-range air travel, and advice is offered by airlines to remain mobile and consume ample quantities of water during such flights. Epidemiologic study of VTE within and outside the hospital setting, and in hospitals without the advanced technological capability now advocated for standard diagnosis of DVT and PE, requires more fully developed and widely employed diagnostic criteria and procedures to advance the needed research. What is the most appropriate protocol for such study, and what are the best prospects for its implementation? Can the disparate estimates of frequency and burden of these conditions be reconciled with improved data availability and greater uniformity of coding and analysis of hospital and mortality data?
ARRHYTHMIAS Atrial Fibrillation Atrial fibrillation is a disturbance of cardiac rhythm that results from dysfunction of the electrophysiologic conduction system in the upper chambers of the heart. The most apparent manifestation of this disturbance is the symptom often described as “palpitations” or “racing of the heart.” Because this type of arrhythmia is often only episodic, or “paroxysmal,” presence of atrial fibrillation may not be detectable at the time of examination. When persistent, however, on physical examination the pulse is classically described as “irregularly irregular,” with a corresponding pattern of irregularity on an electrocardiographic recording of several cardiac cycles. The main consequence of atrial fibrillation is stroke, for which it is a major risk factor. Paroxysmal atrial fibrillation may be especially likely to precipitate a stroke due to the resulting disturbance of blood flow in the atrium but is at the same time likely to escape detection; therefore the true frequency of this condition as a precipitating cause of strokes may well be underestimated. In addition, atrial fibrillation is closely associated with heart failure, coronary heart disease, and stroke, so that its particular contribution to major cardiovascular events is difficult to isolate from others. Heart Disease and Stroke Statistics—2009 Update presents information from multiple sources to complete the picture of atrial fibrillation as a factor in serious cardiovascular disease outcomes.40 There is a very strong relation with age in the risk of atrial fibrillation, according to hospital discharge data. From the Framingham Heart Study, approximately one per-
son in four will develop atrial fibrillation after age 40, with the risk being slightly higher for men than for women. Again referring to hospital discharge data, it appears that African American patients are hospitalized for this condition at younger ages than others. Approximately 15–20% of all ischemic strokes are due to this condition. In long-term monitoring of patient records in Olmsted County, Minnesota, ageadjusted incidence of atrial fibrillation increased by more than 12% from 1980 to 2000. Over this same period, national data indicate that trends in hospitalizations for atrial fibrillation increased sharply.2 The numbers of hospitalizations were several times greater when secondary, versus only primary, diagnoses were counted. Secondary diagnoses increased from about 0.5 to 2.25 million, whereas primary diagnoses increased from about 100,000 to 400,000. Neither of these criteria corresponds to the estimates in Table 6-2. This discrepancy demonstrates the limitations of these data for understanding the population burden of atrial fibrillation. Regardless of these issues, evidence is clear that atrial fibrillation is closely related to risk of stroke, strongly age dependent, and expected to increase greatly in frequency as numbers of older persons increase, not only in the United States but throughout the world. These considerations add to the importance of understanding the epidemiology and potential for prevention of this condition. Current approaches to prevention and control include the possibility that smoking cessation and weight reduction may prevent atrial fibrillation itself, and anticoagulant therapy of patients with this condition has been shown to reduce the risk of stroke. Ventricular Arrhythmias Ventricular arrhythmias are the immediate cause of sudden cardiac death. Figure 4-4 addresses the exceptionally high risk of sudden cardiac death among survivors of myocardial infarction complicated by occurrence of ventricular tachycardia (VT) or fibrillation (VF) in the convalescent phase. Ventricular fibrillation is not often coded as the cause of death—in only about 13,000 instances in 2002—but is considered as the reason for the great majority of some 325,000 sudden deaths per year.40 Discussion of precipitating factors in acute coronary syndromes includes effects of certain conditions or exposures on cardiac rhythm as well as on coagulation mechanisms (Chapter 15). Current Issues Regarding cardiac arrhythmias, atrial and ventricular arrhythmias differ especially in the time relation between their appearance and occurrence of major
46894_CH06_111_138.qxd
136
2/25/10
3:26 PM
Page 136
CHAPTER 6 RELATED CONDITIONS
consequences. Atrial fibrillation is a chronic condition in which, once clinically detected, drug therapy can be instituted with reduction in the risk of complications. Ventricular arrhythmias have serious, often fatal, immediate outcomes in which bystander resuscitation or defibrillation may be the only opportunity to avert death. For both conditions, understanding their epidemiology and potential for prevention is necessary for effective public health action to reduce population risk. A key question is how best to improve the systems of data collection regarding these events to facilitate their identification and prevention on a population-wide basis. REFERENCES 1. World Health Organization. International Statistical Classification of Diseases and Related Health Problems, Tenth Revision. Geneva (Switzerland): World Health Organization; 1992:1. 2. National Heart, Lung and Blood Institute. Morbidity & Mortality: 2007 Chartbook on Cardiovascular, Lung, and Blood Diseases. Bethesda, MD: National Institutes of Health, National Heart, Lung and Blood Institute; June, 2007. 3. Labarthe DR. Epidemiology and Prevention of Cardiovascular Diseases: A Global Challenge. Gaithersberg, MD: Aspen Publishers, Inc; 1998. 4. Criqui MH. Peripheral arterial disease. In: Pearson TA, Criqui MH, Luepker RV, Oberman A, Winston M, eds. Primer in Preventive Cardiology. Dallas, TX: American Heart Association; 1994:83–91. 5. Rose GA. The diagnosis of ischaemic heart pain and intermittent claudication in field surveys. Bull World Health Organ. 1962; 27:645–658. 6. Koelemay MJ, Lijmer JG, Stoker J, Legemate DA, Bossuyt PM. Magnetic resonance angiography for the evaluation of lower extremity arterial disease: a meta-analysis. JAMA. 2001; 285(10):1338–1345. 7. Criqui MH, Fronek A, Klauber MR, et al. The sensitivity, specificity, and predictive value of traditional clinical evaluation of peripheral ar-
terial disease: results from noninvasive testing in a defined population. Circulation. 1985;71: 516–522. 8. Feigelson HS, Criqui MH, Fronek A, et al. Screening for peripheral arterial disease: the sensitivity, specificity, and predictive value of noninvasive tests in a defined population. Am J Epidemiol. 1994;140:526–534. 9. Higgins JP, Higgins JA. Epidemiology of peripheral arterial disease in women. J Epidemiol. 2003;13(1):1–14. 10. Menke A, Muntner P, Wildman RP, Dreisbach AW, Raggi P. Relation of borderline peripheral arterial disease to cardiovascular disease risk. Am J Cardiol. 2006;98:1226–1230. 11. Kuller L, Borhani N, Furberg C, et al. Prevalence of subclinical atherosclerosis and cardiovascular disease and association with risk factors in the Cardiovascular Health Study. Am J Epidemiol. 1994;139:1164–1179. 12. Criqui MH, Langer RD, Fronek A, et al. Mortality over a period of 10 years in patients with peripheral arterial disease. N Eng J Med. 1992;326:381–386. 13. Ankle Brachial Index Collaboration. Ankle brachial index combined with Framingham risk score to predict cardiovascular events and mortality. A meta-analysis. JAMA. 2008;300: 197–208. 14. Fowkes FGR, Housley E, Riemersma RA, et al. Smoking, lipids, glucose intolerance, and blood pressure as risk factors for peripheral atherosclerosis compared with ischemic heart disease in the Edinburgh Artery Study. Am J Epidemiol. 1992;135:331–340. 15. Vogt MT, Wolfson SK, Kuller LH. Segmental arterial disease in the lower extremities: correlates of disease and relationship to mortality. J Clin Epidemiol. 1993;46:1267–1276. 16. United States Preventive Services Task Force. Recommendation Statement: Screening for Peripheral Arterial Disease. Washington, DC: Agency for Healthcare Research and Quality; 2005:1–8.
46894_CH06_111_138.qxd
2/25/10
3:26 PM
Page 137
REFERENCES 137
17. Beckman JA, Jaff MR, Creager MA. The United States preventive services task force recommendation statement on screening for peripheral arterial disease: more harm than benefit? Circulation. 2006;114(8):861–866. 18. Golomb BA, Dang TT, Criqui MH. Peripheral arterial disease: morbidity and mortality implications. Circulation. 2006;114(7):688–699. 19. Kannel WB, Wolf PA. Peripheral and cerebral atherothrombosis and cardiovascular events in different vascular territories: insights from the Framingham Study. Curr Atheroscler Rep. 2006;8(4):317–323. 20. Dalen JE. Diseases of the aorta. In: Braunwald E, Isselbacher KJ, Petersdorf RG, Wilson JD, Martin JB, Fauci AS, eds. Harrison’s Principles of Internal Medicine. 11th ed. New York: McGraw-Hill Book Co; 1987. 21. Lilienfeld DE, Gunderson PD, Sprafka JM, Vargas C. Epidemiology of aortic aneurysms: I. mortality trends in the United States, 1951 to 1981. Arteriosclerosis. 1987;7:637–643. 22. Melton LJ III, Bickerstaff LK, Hollier LH, et al. Changing incidence of abdominal aortic aneurysms: a population-based study. Am J Epidemiol. 1984;120:379–386. 23. Fowkes FGR, Macintyre CCA, Ruckley CV. Increasing incidence of aortic aneurysms in England and Wales. Br Med J. 1989;298: 33–35. 24. Hu Y, Shimizu H, Takatsuka N, et al. Trends in prevalence rates of asymptomatic aortic aneurysms in Japan based on autopsy series. Am J Epidemiol. 1995;5:159–163.
27. Pleumeekers HJCM, Hoes AW, van der Does E, et al. Aneurysms of the abdominal aorta in older adults. Am J Epidemiol. 1995;142: 1291–1299. 28. Webster MW, St. Jean PL, Steed DL, et al. Abdominal aortic aneurysm: results of a family study. J Vasc Surg. 1991;13:366–372. 29. Lederle FA, Nelson DB, Joseph AM. Smokers’ relative risk for aortic aneurysm compared with other smoking-related diseases: a systematic review. J Vasc Surg. 2003;38:329–334. 30. Wong DRF, Willett WC, Rimm EB. Smoking, hypertension, alcohol consumption, and risk of abdominal aortic aneurysm in men. Am J Epidemiol. 2007;165:838–845. 31. Fleming C, Whitlock EP, Beil TL, Lederle FA. Screening for abdominal aortic aneurysm: a best-evidence systematic review for the U.S. Preventive Services Task Force. Ann Intern Med. 2005;142(3):203–211. 32. Hunt SA, Abraham WT, Chin MH, et al. ACC/AHA guideline update for the diagnosis and management of heart failure in the adult— summary article: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to Update the 2001 Guidelines for the Evaluation and Management of Heart Failure). Circulation. 2005;112: 1825–1852. 33. Jessup M, Abraham WT, Casey DE, et al. 2009 Focused update: ACCF/AHA guidelines for the diagnosis and management of heart failure in adults. A report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. Circulation. 2009;119:1977–2016.
25. Wilmink AB, Quick CR. Epidemiology and potential for prevention of abdominal aortic aneurysm. Br J Surg. 1998;85(2):155–162.
34. Kannel WB. Incidence and epidemiology of heart failure. Heart Fail Rev. 2000;5(2): 167–173.
26. Filipovic M, Goldacre MJ, Roberts SE, Yeates D, Duncan ME, Cook-Mozaffari P. Trends in mortality and hospital admission rates for abdominal aortic aneurysm in England and Wales, 1979–1999. Br J Surg. 2005;92(8): 968–975.
35. Kannel WB, Plehn JF, Cupples A. Cardiac failure and sudden death in the Framingham Study. Am Heart J. 1988;115:869–875. 36. Schocken DD, Arrieta MI, Leaverton PE, Ross EA. Prevalence and mortality rate of congestive
46894_CH06_111_138.qxd
138
2/25/10
3:26 PM
Page 138
CHAPTER 6 RELATED CONDITIONS
heart failure in the United States. J Am Coll Cardiol. 1992;20:301–306. 37. US Department of Health and Human Services. Healthy People 2010 Midcourse Review. Washington DC: US Government Printing Office, December 2006. 38. Gillum RF. Epidemiology of heart failure in the United States. Am Heart J. 1993;126: 1042–1047. 39. National Heart, Lung and Blood Institute. Incidence & Prevalence: 2006 Chart Book on Cardiovascular and Lung Diseases. Washington, DC: US Department of Health and Human Services, Public Health Service, National Institutes of Health; 2006.
Research; Quality of Care and Outcomes Interdisciplinary Working Group; and Functional Genomics and Translational Biology Interdisciplinary Working Group. Circulation. 2008;117:2544–2565. 44. Narula J, Yancy CW, Young JB, eds. Primary prevention of heart failure. Med Clin N Am. 2004;88(5):1129–1390. 45. Goldhaber SZ. Pulmonary embolism. In: Braunwald E, ed. Heart Disease: A Textbook of Cardiovascular Medicine. 5th ed. Philadelphia, PA: WB Saunders Co; 1997:1582–1603. 46. White RH. The epidemiology of venous thromboembolism. Circulation. 2003;107(23 suppl 1): I4–I8.
40. Lloyd-Jones D, Adams R, Carnethon M, et al. Heart disease and stroke statistics—2009 update. A report from the American Heart Association Statistics Committee and Stroke Statistics Subcommittee. Circulation. 2009;119:e1–e161.
47. Hull RD, Kakkar VV, Raskob GE. Prevention of venous thrombosis and pulmonary embolism. In: Fuster V, Verstraete M, eds. Thrombosis in Cardiovascular Disorders. Philadelphia, PA: WB Saunders Co; 1992: 451–464.
41. Ho KKL, Pinsky JL, Kannel WB, Levy D. The epidemiology of heart failure: the Framingham Study. J Amer Coll Cardiol. 1993;22 (suppl A):6A–13A.
48. Zierler BK. Ultrasonography and diagnosis of venous thromboembolism. Circulation. 2004; 109(12 suppl 1):I9–I14.
42. Bumgarner JR, Speizer FE. Chronic obstructive pulmonary disease. In: Jamison DT, Mosley WH, Measham AR, Bobadilla JL, eds. Disease Control Priorities in Developing Countries. Oxford, England: Oxford University Press; 1993:595–608. 43. Schocken DD, Benjamin EJ, Fonarow GC, et al. Prevention of heart failure: a scientific statement from the American Heart Association Councils on Epidemiology and Prevention, Clinical Cardiology, Cardiovascular Nursing, and High Blood Pressure
49. Goldhaber SZ, Sors H. Treatment of venous thrombosis and pulmonary embolism. In: Fuster V, Verstraete M, eds. Thrombosis in Cardiovascular Disorders. Philadelphia, PA: WB Saunders Co; 1992:465–483. 50. Heit JA. Venous thromboembolism: disease burden, outcomes and risk factors. J Thromb Haemost. 2005;3(8):1611–1617. 51. Kroegel C, Reissig A. Principle mechanisms underlying venous thromboembolism: epidemiology, risk factors, pathophysiology and pathogenesis. Respiration. 2003;70(1):7–30.
46894_CH07_139_158.qxd
2/25/10
3:26 PM
Page 139
The Main Determinants
139
P A R T
3
46894_CH07_139_158.qxd
2/25/10
3:26 PM
Page 140
46894_CH07_139_158.qxd
2/25/10
3:26 PM
Page 141
C H A P T E R
7 Genes and Environment environmental effects to modify the effects of genes, leading to variation in disease associations among groups within a population, between populations, or over time. Applications of genomic epidemiology to family studies, investigations of gene–environment interaction, and specific cardiovascular conditions— atherosclerosis and coronary heart disease, blood lipids and blood pressure, and stroke—provide examples of work in this area. Several current issues in genomic epidemiology include implications for cardiovascular epidemiology and prevention. One broad issue is that of managing expectations. The often-cited limitations of studies to date, inherent in the complex phenomena being addressed, warrant circumspection. Prospects for the “genomic revolution” to have an imminent impact on medical practice, including genotypic diagnosis and personalized treatment of common diseases, appear more distant than many proponents have claimed. To conceptualize common diseases as highly individualized conflicts with generalized recommendations, guidelines, and policies at the population level, an issue deserving thoughtful consideration. Already, concern is mounting in regard to genetic testing. Individual purchase of testing services marketed on the Internet and elsewhere, in the absence of regulatory controls or sound scientific guidance, risks misinformation and inappropriate action as a consequence. Meanwhile, work of the National Human Genome Research Institute, NIH, and the National Office of Genomics and Disease Prevention, CDC, continues to address such issues in translation of evidence from genomics into clinical and public health practice.
SUMMARY An introduction to genetic, or genomic, epidemiology and its application to cardiovascular diseases provides insight into this rapidly advancing field. At one end of the spectrum of genomic epidemiology are rare disorders that may be caused by a single dominant gene mutation. At the other are common “complex” diseases whose genetic underpinnings may involve only weak effects of single genes and influential interactions among multiple genes and with the environment. The first are studied in familial pedigrees to trace their patterns of inheritance. The latter may involve large case-control or cohort studies to identify genetic effects through a population-wide search, with or without a family dimension. The potential scale of such studies includes a vision of very large internationally linked cohort studies to investigate genes and gene–environment interactions, viewed as jointly determining the occurrence of these diseases. Current concepts and strategies of research in genomic epidemiology are highlighted, including emphasis on the theoretical importance and practical utility of family history. Basic sources in the field include not only frequently updated texts but also series of articles published recently to keep pace with developments. Discussion of methods in genomic epidemiology focuses especially on linkage analysis in family studies and genome-wide association studies in whole populations. Major limitations of these methods include small sample sizes of single studies, a high probability of false-positive results, and frequent inability to replicate findings in subsequent studies. Compounding these difficulties is the potential for
141
46894_CH07_139_158.qxd
142
2/25/10
3:26 PM
Page 142
CHAPTER 7 GENES AND ENVIRONMENT
INTRODUCTION Familial Risks Conventional wisdom holds that diseases “run in families.” Cardiovascular conditions including myocardial infarction, stroke, and their major risk factors have long been regarded as familial, in that individual risk in the current generation is increased if these have been present in a sibling or parent, especially at early ages. Familial risks expressed across generations are often thought of as genetically determined. However, not all familial resemblance has a genetic basis, and “cultural heritability”—broadly embracing environmental conditions such as social and behavioral factors shared among family members—is recognized as an important component of heredity. Consequently, genes are considered together with environment in assessing their role in cardiovascular diseases. This view is illustrated by Khoury and others in Fundamentals of Genetic Epidemiology, which reviews basic concepts and methods of both populationbased and family-based genetic studies (Figure 7-1).1 Here the phenotype—any observable characteristic, such as LDL-cholesterol concentration—appears as a direct effect of the genotype and is subject to modification by environmental conditions. Also in play are the population dynamics that determine parental mating patterns and the combination of genes that re-
sults for a given individual. This scheme implies that gene mutations or variations may or may not result in recognizable changes or differences in phenotypes, depending on these additional influences. Questions about the genetics of cardiovascular and other diseases, at least from the perspectives of epidemiology and public health, therefore concern both genes and environment, or “gene–environment interactions.” Recent Reviews of Methods Study of hereditary or genetic factors and disease at the population level utilizes concepts and methods that identify the specialized discipline of genetic epidemiology. A brief discussion of these concepts and methods and some examples of their application to cardiovascular diseases will serve to introduce this aspect of cardiovascular epidemiology and prevention. For more general discussion of genetic epidemiology, further background is available from many sources. Especially useful are two texts and two sets of articles, one appearing in the International Journal of Epidemiology (IJE) in 2004, the other in The Lancet in 2005.2–15 The texts and first set of articles are highlighted here for their general interest, whereas The Lancet’s series addresses specific research strategies and is discussed in the following section. Because this is a rapidly developing area, however, and current literature is quickly outdated, emphasis here is on general principles.
Figure 7-1 The Scope of Genetic Epidemiology. Source: From Fundamentals of Genetic Epidemiology by Muin J Khoury, Terri H Beatty, Bernice Cohen. © 1993 by Oxford University Press, Inc. Used by permission of Oxford University Press, Inc.
46894_CH07_139_158.qxd
2/25/10
3:26 PM
Page 143
INTRODUCTION 143
Genetic, Genomic, and Epigenetic Epidemiology Although neither of the texts listed here is specific to cardiovascular diseases, each provides valuable insights to the field of genetics that are useful for the epidemiologist or public health practitioner. Genetics and Public Health in the 21st Century (2000) presents chapters on implications of genetics for public health; public health assessment of genetic predispositions to selected diseases; evaluation of genetic testing; development, implementation, and evaluation of genetic services as population interventions; ethical, legal, and social issues; and communication, education, and information dissemination regarding genetics for public health.2 Human Genome Epidemiology (2004) introduces the term proposed by Khoury and colleagues in preference to “genetic epidemiology” on the basis that it embraces epidemiologic study of the whole human genome and not only single genes.3,16 This reflects the engagement of growing numbers of epidemiologists in research on the human genome, first published in Science on February 16, 2001, including a 40˝ 57˝ foldout map.17 The collected chapters address fundamental concepts; methods for assessing disease associations and interactions; assessing genetic tests for disease prevention; and case studies demonstrating use of human genome epidemiology to improve health. The October 2004 issue of the IJE focuses on the theme of genetic epidemiology. It presents the view that in consequence of “the genomics revolution” epidemiology and genetics are experiencing “a growing union” in which each discipline is transforming the other.5, p 925 Khoury and others further develop the idea that epidemiology is entering a new era because of its discovery by geneticists (a perhaps exaggerated assessment in view of the continuing significant contributions of epidemiology in long-established areas).7 Although the major achievements of recent years have doubtless accelerated collaborations between epidemiology and genetics, these have origins as early as the 1960s, as reviewed by Schull and Hanis in 1990.18 Another dimension of the evolving discipline discussed in some detail in the IJE series is the concept of “epigenetic epidemiology . . . the part of epidemiology that studies the effects of heritable epigenetic changes on the occurrence and distribution of diseases.”6, p 929 Epigenetic inheritance adds to the complexity of understanding hereditary effects by recognizing defects that are attributable not to genetic mechanisms, i.e., changes in DNA sequence, but to cellular ones. Several mechanisms are considered plausible as “epigenetic inheritance systems,” and the implications of this mode of inheritance for epidemiologic research—beyond those
already identified in connection with genomics—are explored. For example, it is suggested that variations in intrauterine environment might function as influences capable of inducing heritable changes in cells that could persist throughout life and over multiple generations. Genetic Complexity of Coronary Heart Disease Even without the added consideration of epigenetic inheritance, complexity of genetic influences specific to cardiovascular diseases is suggested in a second schematic representation, from Sing and others (Figure 7-2).19 They portray an array of genes acting in combination to determine each of four intermediate traits—hemostasis, lipid metabolism, carbohydrate metabolism, and blood pressure regulation. These traits are shown as converging to determine the probability of developing coronary artery disease (CAD), conditional on environment and age. However, the intermediate traits themselves are influenced importantly by environmental factors, as will be seen in subsequent chapters. One implication is that gene– environment interactions are operating throughout the process of developing atherosclerotic and hypertensive diseases. A key challenge is to discover and understand these interactions and the potential for interventions to modify them favorably. Genetics, Genomics, and Cardiovascular Diseases A current assessment of progress in this pursuit is presented in a 2007 American Heart Association Scientific Statement, Relevance of Genetics and Genomics for Prevention and Treatment of Cardiovascular Disease.20 The report summarizes current knowledge as follows:20, p 2878 Evidence accumulated over decades convincingly demonstrates that family history in a parent or a sibling is associated with atherosclerotic CVD, manifested as coronary heart disease, stroke, and/or peripheral arterial disease. . . . Most common forms of CVD are believed to be multifactorial and to result from many genes, each with a relatively small effect working alone or in combination with modifier genes and/or environmental factors. The identification and the characterization of these genes and their modifiers would enhance prediction of CVD risk and improve prevention, treatment, and quality of care. This and similar statements regarding expectations of genetics or genomics make it important to appreciate the concepts and strategies of investigation in
46894_CH07_139_158.qxd
144
2/25/10
3:26 PM
Page 144
CHAPTER 7 GENES AND ENVIRONMENT
Figure 7-2 Biological Complexity of Coronary Artery Disease (CAD): Genes, Intermediate Traits, and Probability of CAD. Source: Reprinted with permission from CF Sing et al., Annals of Medicine, Vol 24, p 541, © 1992, The Finnish Medical Society Duodecim.
this field. Examples of findings on family history, gene–environment interaction, and other relationships in the genomics of cardiovascular diseases will then illustrate recent work in this area.
CONCEPTS AND STRATEGIES OF GENETIC EPIDEMIOLOGY The contribution of Burton and others in The Lancet’s 2005 series on genetic epidemiology addresses key concepts, including a glossary and interpretation of basic terms in genetics.9 Phenotypic aggregation of diseases within families is discussed with emphasis on “familial aggregation,” in which the frequency of a disease is greater on average among relatives of cases than among relatives of unaffected individuals. Familial aggregation of continuous traits, such as blood pressure, can also be assessed by appropriate analytic approaches. Finding evidence of aggregation leads to further investigation to estimate the magnitude of genetic contribution to the observed pattern. To detect a major gene, with a strong effect on disease susceptibility, the pattern of transmission can be de-
termined by the method of segregation analysis. However, in “complex diseases” in which susceptibility is presumed to be determined by multiple genes, each with only weak effects, other methods are required. These include genetic linkage studies and genetic association studies.10,11 Linkage Analysis Linkage analysis is used to assess whether presence of a disease or trait is accompanied consistently across individuals by presence of a particular genetic marker, a DNA or protein sequence with a specific chromosomal location.1 If so, as described by Teare and Barrett, there may be a gene or combination of genes near this locus contributing to causation of the disease or trait.10 Analysis is based on observations in related individuals and proceeds in distinct ways depending on whether a major gene disorder or complex disease is being studied. In the latter case, as for example in coronary heart disease, no clear mode of inheritance is present and so-called model-free methods are used. Genotyping of sibling pairs or other pairs of related persons with rigorous ascertainment of disease status provides the data needed to assess whether
46894_CH07_139_158.qxd
2/25/10
3:26 PM
Page 145
CONCEPTS AND STRATEGIES OF GENETIC EPIDEMIOLOGY 145
linkage is present. Such studies of complex diseases are considered not to have been very successful to date, possibly because gene effects are weak and sample sizes of the studies have been small. Genome-Wide Association Studies Cordell and Clayton distinguish genetic association studies from linkage analysis as seeking to establish whether one or more alleles is associated with a disease or trait across a whole population and not just among related persons.11 Association studies are capable of detecting weaker genetic effects but require that many markers be investigated. Issues in interpreting such studies include the consequence of vast numbers of gene loci that can now be examined and the likely great predominance of negative findings. This point is emphasized by Pearson and Manolio, who describe the genome-wide association (GWA) study as:21 revolutionary because it permits interrogation of the entire human genome at levels of resolution previously unattainable, in thousands of unrelated individuals, unconstrained by prior hypotheses regarding genetic associations with disease. However . . . [it] presents an unprecedented potential for false-positive results, leading to a new stringency in acceptable levels of statistical significance and requirements for replication of findings. Genome-wide association studies, described by Palmer and Cardon, have come to take advantage of a particular type of marker, the single-nucleotide polymorphism, or SNP.12 This is a form of genetic variation characterized by difference within a DNA sequence of a single nucleotide or base molecule, which is easily typed by current laboratory methods. A “common SNP” is considered to be one in which each of two alleles is present at a frequency of 1% or greater; it is estimated that 10 million such SNPs may occur across the human genome and together account for 90% of human genetic variation. A SNP found in association with disease may be causally related to disease by altering production of a critical protein, or it may mark the approximate location of such a gene. This latter feature underlies use of SNPs for mapping of genetic disorders within the genome, among other applications. It has been noted that so-called whole-genome studies are not strictly that, because not all SNPs have yet been identified and made available for study. Additionally, the 100,000 or more SNPs available in current commercial panels (perhaps 1% of the total
of common SNPs) are selective in their coverage. Even so, the cost per study subject remains high. One suggested strategy for reducing the cost of large genomic case-control studies (those with 1000 or more cases) is to develop a source of “universal controls” based on a very large bank of samples from which subsets could be matched to controls in many studies. Factors influential in the quality of a genetic association study are discussed by Hattersley and McCarthy and offer guidance to those who wish to appreciate and critically evaluate reports of such studies.13 They list among key qualities study design, implementation and interpretation, based on adequately powered samples, as well as sample recruitment strategy, logic of genotypic variant selection, reliability of genotyping, relevance of data analysis, and validity of interpretation. Means of strengthening studies with regard to these influences are also proposed. Population-Based Family Designs Coming full circle to family studies once again, Hopper and others describe “population-based family designs,” whose fullest form is the “case-controlfamily” design.14 Cases and controls are sampled from a defined population and their relatives are included in data collection, potentially via interview, examination, and genotyping. In a compromise with the practical reality of control recruitment from the general population, controls may be drawn from cases’ spouses or partners and their relatives. Cases may be most readily identified through existing registries—an approach that to date is less often feasible in the cardiovascular arena than in cancer, where population-based registries are relatively common. Various strategies of investigation may focus on the cases and controls primarily, on the disease history of the relatives up to the time of diagnosis of the case (“retrospective cohort studies”), or on the future development of disease in the relatives of cases (and controls) (“prospective cohort studies of sets of relatives”). Although there are difficulties in design, conduct, analysis, and interpretation of such studies, the authors speculate that population-based case-family designs:14, p 1404 could be the future of epidemiology, not just genetic epidemiology. Because of their versatility, retrospective and prospective populationbased family studies may become the principal framework for epidemiology in the future and move genetics from its traditional focus on socalled high-risk families to give it a wider clinical and population health relevance.
46894_CH07_139_158.qxd
146
2/25/10
3:26 PM
Page 146
CHAPTER 7 GENES AND ENVIRONMENT
FAMILY HISTORY From the clinical perspective, and in earlier epidemiologic studies of coronary heart disease, attention to genetic aspects was largely limited to questions of family history. For the atherosclerotic and hypertensive diseases, the early focus was on the degree of resemblance among differently related persons with respect to levels of particular risk factors as an indication of possible inheritance of these traits. Studies of twins gave further insight into the role of the genotype in determining risk factor levels through contrasts in the degree of resemblance between monozygotic (identical) versus dizygotic (fraternal) twins. Further studies compared twins raised together and apart, and by natural or adoptive parents, to assess environmental and genetic contributions to observed phenotypes including risk-factor levels. On the basis of such studies, estimates of the heritability of these traits were derived, including partition into genetic and “cultural” components, the latter reflecting nongenetic or, broadly, environmental contributions.22 The potential value of family history of coronary heart disease for purposes of clinical intervention was emphasized by Williams and colleagues.23 Concentrating on familial hypercholesterolemia (see Chapter 11), this group advocated identification of pedigrees or family trees with high risk of this condition as an intervention strategy. The “MED PED” program (tracking MEDical PEDigrees to Make Early Diagnoses and Prevent Early Death) obtained “health family trees” from nearly 90,000 Utah families. They found some 3000 families with a “strong history” of early coronary heart disease and 50,000 persons related closely enough to be considered as high-risk family members. Their approach also identified, from 100 index cases of heterozygous familial hypercholesterolemia, 500 other affected individuals, twothirds of whom had not been diagnosed or were not effectively treated to reduce LDL-cholesterol levels. The MED PED protocol was outlined in sufficient detail to be implemented by others, as Williams and colleagues urged. Studies of Coronary Heart Disease A recent review summarized 19 of the larger epidemiologic studies (more than 500 subjects per study) on family history of coronary heart disease reported from 1975–2001 (Table 7-1).24 The purpose of the review was to “provide a rationale for why family history, mundane as it might sound in this genomic era, should be considered an important measure of a person’s genomic and total environmental risk of future CHD.” The cited studies were located mainly in the
United States and also included four in Finland, Italy, Israel, and South Africa. Studies ranged in size from hundreds to tens of thousands of participants, based in groups of men, women, or families. African Americans were noted to be included in two of the studies. Cross-sectional, case-control, and cohort designs were represented. The criteria for positive family history differed among studies with respect to disease (expressed as CHD, MI, MI or stroke, or CHD risk score, and fatal or not specified), age (younger than 55 or 60 years or not specified), and relation to study subject (parent, first- or seconddegree relative, or multigenerational score). Despite these important differences, there was general consistency in finding a positive association—often explicitly independent of major risk factors—between family history and CHD in these populations. The authors’ interpretation of these observations is that family history constitutes “one of the best indicators of a person’s genomic and ecologically derived risk.” Further:24, p 148 it is quite possible that even with our ability to measure hundreds and thousands of genes and environments we may find that family history is the best, low-cost way to identify the at-risk groups in the population. This will be especially true if gene-gene and gene-environment interactions play a major role in determining risk of future disease. In this genomic context, the measurement of a few individual risk factors pales in comparison to the compressed information held in a simple family history. A High-Risk Family One example of a study linking family history with genetic assessment appeared in Science, 2007, and concerns a family of Iranian ancestry. The subjects were selected for study from a registry of coronary artery disease (CAD) cases and families because the index case of myocardial infarction occurred at the early age of 48 years.25 Of 58 identified blood relatives, 28 had MI, angina, or sudden death before age 50 for men or 55 for women, and 23 of them had died from CAD. For examining risk-factor status in this family, data could be collected on 27 individuals—13 with early CAD, 5 that were older than 50 or 55 years and that were free of CAD, and 9 younger individuals free of known CAD. Genome-wide linkage was investigated in 19 family members, although it was not reported whether they were from the group with clinical data. A segment of chromosome 12p appeared to be linked and includes the gene, LDL receptor-related protein 6,
46894_CH07_139_158.qxd
2/25/10
3:26 PM
Page 147
FAMILY HISTORY 147
Table 7-1
Citation
Selected Studies That Establish the Importance of Family History as a Risk Factor for Coronary Heart Disease Sample Definition of Study Study Design Family History Conclusions
Sesso et al., 2001
Physicians’ Health Study (PHS) 22,071 men; Women’s Health Study (WHS); 39,876 women
Prospective cohort. Men (40–84 yrs.) Women ( 45 yrs.)
Parental MI
Early parental MI ( 60 yrs.) confers greater CHD risk than parental MI at older ages. Maternal history of MI may be important at any age.
Williams et al., 2001
The Health Family Tree Study 128,733 families; NHLBI Family Heart Study 1442 families
Family-based: Crosssectional
FRS from 3generation pedigrees
14% of families (FRS 0.5) contain 72% of early CHD cases in the population.
Li et al., 2000
The Atherosclerosis Risk in Communities (ARIC) Study and NHLBI Family Heart Study 3958; African Americans; 10,580 Caucasians
Prospective cohort: African American and Caucasians (45–65 yrs.)
FRS from first-degree relatives
The FRS predicts incident CHD in African Americans and Caucasians equally.
Pohjola-Sintonen et al., 1998
Study in Helsinki, Finland; 707 cases, 716 controls
Case–control: MI Survivors (29–59 yrs.); Clinical controls (26–59 yrs.)
CHD in first-degree relatives
History of CHD in firstdegree relatives is a greater risk factor in women than men.
Higgins et al., 1996
NHLBI Family Heart Study; 657 high-risk families, 588 randomly sampled families
Retrospective cohort: Probands (45–69 yrs.)
Family risk score (FRS) based on first-degree relatives
FRS can be used to identify high-risk families in the population at large.
Jousilahti et al., 1996
12-year follow-up study performed in eastern Finland, 19,894 study participants
Prospective cohort: Men and women (30–59 yrs.)
Parental CHD 60 yrs.
History of early CHD in parents is a greater risk factor for women than men.
Rotimi et al., 1994
Study in Chicago, IL, 232 families
Case–control: African American and Caucasian
CHD in first-degree relatives
History of CHD in firstdegree relatives are associated with increased CHD risk in African Americans.
Roncaglioni et al., 1992
Gruppo Italiano per lo Studio della Sopravivenza nell’Infarto (GISSI)2 Trial, Italy; 916 hospital cases, 1106 hospital controls
Case–control: MI survivors ( 75 yrs.); Clinic controls ( 75 yrs.)
MI in first-degree relatives
The number and age of relatives with CHD are related to the strength of association with MI. Family history interacts with known CHD risk factors. (continues)
46894_CH07_139_158.qxd
148
2/25/10
3:26 PM
Page 148
CHAPTER 7 GENES AND ENVIRONMENT
Table 7-1
Citation
Selected Studies That Establish the Importance of Family History as a Risk Factor for Coronary Heart Disease—continues Sample Definition of Study Study Design Family History Conclusions
Colditz et al., 1991
Health Professionals Follow-up Study 45,317 study participants
Prospective cohort
Parental MI 60 yrs.
Early parental MI confers greater CHD risk than MIs at older ages.
Rossouw et al., 1991
Coronary Risk Factor Study (CORIS) Baseline Study, South Africa; 2722 men, 3173 women
Cross-sectional
CHD in first-degree relatives
Family history of MI interacts with known risk factors in determining risk of CHD.
Myers et al., 1990
Framingham Study 30–year follow-up; 5209 study participants
Prospective cohort: Caucasian (28–62 yrs.)
Parental death due to CHD
Family history of CHD is an independent predictor of CHD.
Schildkraut et al., 1989
Framingham Study— 28-year follow-up; 5209 study participants
Prospective cohort: Caucasian (28–62 yrs.)
Parental death due to CHD
Early parental CHD ( 60 yrs.) confers greatest risk, especially in women.
Hopkins et al., 1988
Cardiovascular Genetics Research Clinic Study, Salt Lake City, UT, 1196; study participants
Prospective cohort
Number of first-and second-degree relatives with early CHD ( 55 yrs)
Family history of CHD is an independent predictor of CHD.
Khaw and Barrett-Connor, 1986
Rancho Bernardo, CA, Study, 9-yr followup; 4014 study participants
Prospective cohort: Men and women (40–79 yrs.)
MI in first-degree relatives
Family history interacts with smoking to predict CHD risk.
Colditz et al., 1986
Nurse’s Health Study; 117,156 study participants
Prospective cohort: Nurses (30–55 yrs.)
Parental MI
Early parental CHD ( 60 yrs.) is associated with greater risk. Family history is an independent predictor of CHD.
Friedlander et al., 1985
Jerusalem Lipid Research Clinic Prevalence Study; 1044 study participants
Case-control
MI in first-degree relatives 60 yrs.
Family history of CHD is an independent predictor of CHD.
Barrett-Connor and Khaw, 1984
Rancho Bernardo, CA, Study; 4014 study participants
Prospective cohort: Men and women (40–79 yrs.)
MI in first-degree relatives
Family history of CHD is an independent predictor in men but not women.
Rosenberg et al., 1983
Boston, New York Study of MIs and oral contraceptive use; 255 cases; 802 controls
Case-control: Women MI survivors (25–49 yrs.); Clinic controls (25–49 yrs.)
MI or stroke in firstdegree relatives
Early maternal ( 60 yrs.) but not paternal MI is associated with CHD risk.
46894_CH07_139_158.qxd
2/25/10
3:26 PM
Page 149
FAMILY HISTORY 149
Table 7-1
Citation Sholtz et al., 1975
Selected Studies That Establish the Importance of Family History as a Risk Factor for Coronary Heart Disease—continues Sample Definition of Study Study Design Family History Conclusions Western Collaborative Group Study; 3524 study participants
Prospective cohort: Men (39–59 yrs.)
Parental CHD
Family history of CHD is an independent predictor of CHD risk.
CHD, coronary heart disease; FRS, family risk score; MI, myocardial infarction. Source: Reprinted with permission from American Journal of Preventive Medicine, Vol 24, SLR Kardia, SM Modell, PA Peyser. p 146–147, © American Journal of Preventive Medicine.
or LRP6. The index case was found to have a homozygous mutation of this gene that was in turn associated with high LDL-cholesterol concentration in all five carriers of this genotype. They also had high triglycerides, high systolic and diastolic blood pressure, fasting blood glucose, and history of diabetes when compared with noncarriers, conforming to criteria for the “metabolic syndrome.” The authors conclude that their findings “establish a causal link between LRP6 mutation and early CAD with high LDL . . .” The suspected rarity of this mutation in the general population may make it very difficult to replicate the finding and strengthen the conclusion that
mutation of LRP6 fully explains the pattern of CAD seen in this family. Maternal and Paternal Influences A further example illustrates exploration of family history to evaluate sex-specific contributions of inheritance of CHD between parents and offspring.26 The study was conducted through the Swedish Multigeneration Register of all births since 1932 and their parents, linked with the national hospital discharge register to identify cases of fatal or nonfatal CHD in both generations (Tables 7-2a and 7-2b). Standardized incidence ratios (SIRs) were calculated
Table 7-2a
Standardized Incidence Ratios with 95% CIs and Number of Cases for CHD by Parental CHD, in Men (Panel a) Father with CHDa Mother with CHDa Both Parents with CHDa Cases SIR 95% CI Cases SIR 95% CI Cases SIR 95% CI Age at diagnosis of CHD (years) 30 32 2.72 1.86–3.84 6 2.10 0.76–4.60 3 5.33 1.00–15.76 30–39 323 2.08 1.86–2.31 148 2.99 2.53–3.52 85 5.00 4.00–6.19 40–49 1757 1.68 1.60–1.76 903 1.94 1.82–2.07 609 3.10 2.86–3.35 50–59 2580 1.31 1.26–1.36 1790 1.50 1.43–1.57 993 1.90 1.78–2.02 60–69 734 1.10 1.03–1.19 662 1.19 1.10–1.29 321 1.42 1.27–1.59 Occupational status Farmers 591 1.44 1.33–1.56 374 1.65 1.49–1.83 225 2.22 1.94–2.53 Unskilled/skilled workers 2698 1.48 1.42–1.53 1700 1.56 1.49–1.64 1005 2.13 2.00–2.27 White collar workers 1000 1.37 1.29–1.46 645 1.53 1.41–1.65 349 2.01 1.81–2.24 Professionals 64 1.23 0.95–1.58 47 1.81 1.33–2.41 15 1.5 0.86–2.56 Self-employed 279 1.57 1.39–1.77 126 1.46 1.22–1.74 93 2.48 2.00–3.04 Others 794 1.22 1.14–1.31 617 1.47 1.36–1.59 324 1.91 1.71–2.13 Region Big cities 922 1.73 1.62–1.85 606 1.92 1.77–2.08 319 2.47 2.21–2.76 Southern Sweden 1245 1.59 1.50–1.68 827 1.80 1.68–1.93 475 2.33 2.13–2.55 Northern Sweden 3259 1.29 1.24–1.33 2076 1.39 1.33–1.45 1217 1.93 1.82–2.04 All 5426 1.41 1.37–1.45 3509 1.55 1.50–1.60 2011 2.09 2.00–2.18 Notes: Reference group consists of men with both parents not affected by CHD during the study period. a Models adjusted for all the explanatory variables. CHD, coronary heart disease; CI, confidence interval; SIR, standardized incidence ratio. Source: Reprinted with permission from American Journal of Preventive Medicine, Vol 30, K Sundquist, X Li. © 2006 American Journal of Preventive Medicine.
46894_CH07_139_158.qxd
150
2/25/10
3:26 PM
Page 150
CHAPTER 7 GENES AND ENVIRONMENT
Table 7-2b
Standardized Incidence Ratios with 95% CIs and Number of Cases for CHD in Women (Panel b) Father with CHDa Mother with CHDa Both Parents with CHDa Cases SIR 95% CI Cases SIR 95% CI Cases SIR 95% CI Age at diagnosis of CHD (years) 30 9 0.87 0.40–1.67 7 2.69 1.07–5.57 1 1.85 0.00–10.62 30–39 100 1.82 1.48–2.21 40 2.31 1.65–3.14 27 4.65 3.06–6.78 40–49 437 1.37 1.25–1.51 254 1.85 1.63–2.09 151 2.64 2.23–3.09 50–59 743 1.12 1.04–1.21 542 1.38 1.27–1.50 288 1.65 1.47–1.86 60–69 261 0.93 0.85–1.05 279 1.18 1.05–1.33 142 1.46 1.23–1.73 Occupational status Farmers 131 1.23 1.03–1.46 93 1.65 1.33–2.02 50 2.02 1.50–2.67 Unskilled/skilled workers 671 1.19 1.10–1.28 451 1.38 1.26–1.51 271 1.87 1.65–2.10 White collar workers 470 1.18 1.07–1.29 352 1.42 1.28–1.58 173 1.71 1.46–1.98 Professionals 20 1.59 0.97–2.46 10 1.27 0.60–2.34 5 2.29 0.72–5.39 Self-employed 69 1.26 0.98–1.59 48 1.73 1.28–2.30 23 2.02 1.28–3.04 Others 189 1.00 0.86–1.15 168 1.40 1.19–1.62 87 1.75 1.40–2.15 Region Big cities 256 1.28 1.12–1.44 164 1.29 1.10–1.51 90 1.76 1.42–2.17 Southern Sweden 359 1.48 1.33–1.64 285 2.00 1.77–2.24 185 2.92 2.51–3.37 Northern Sweden 935 1.06 0.99–1.13 673 1.30 1.21–1.40 334 1.52 1.36–1.69 All 1550 1.17 1.11–1.23 1122 1.43 1.34–1.51 609 1.82 1.68–1.97 Notes: Reference group consists of women with both parents not affected by CHD during the study period. a Models adjusted for all the explanatory variables. CHD, coronary heart disease; CI, confidence interval; SIR, standardized incidence ratio. Source: Reprinted with permission from American Journal of Preventive Medicine, Vol 30, K Sundquist, X Li. © 2006 American Journal of Preventive Medicine.
to compare incidence of CHD separately for women and men, one or both of whose parents had CHD, with persons neither of whose parents had CHD as the reference. Taking all cases into consideration as in the last row of each table, SIRs were significantly greater than 1 for both women and men and increased in relation to whether the father only, mother only, or both parents were affected. In general, these relationships were consistent at all ages of the offspring and strongest for the youngest ages except younger than 30 years, for which few events were observed. These analyses were not restricted by age at occurrence of CHD in the parents, although separate analysis confirmed greater SIRs for both men and women when the parental CHD occurred early. The authors of the Swedish study note the implication that intervention to reduce CHD risk may be reinforced by knowledge of parental history, especially when the mother or both parents have been affected.26 The concept that a positive family history of CHD can motivate preventive action by physicians— and by individuals—is a recurring theme in this area. The 2001 Healthstyles survey in the United States, for example, showed that cholesterol screening and aspirin use, but not behavior related to tobacco, nutrition, or physical activity, were related to having one or two versus no first-degree relatives with CVD.27
Absence of information about family history is an obvious limitation that applies to several guidelines in CVD prevention where recommendations differ on this basis. For this reason, all parties including individuals, physicians, and planners of electronic health records have been urged to take the potential value of this information into account.28 A recently developed tool, My Family Health Portrait, is designed to support use of family history for clinical risk assessment, early disease detection, and prevention and is a webbased, self-administered record of CVD, diabetes, and cancer in one’s family (https://familyhistory.hhs .gov/fhh-web/home.action).29
GENE–ENVIRONMENT INTERACTION Variations on the Theme Life, it could be said, is a gene–environment interaction—genetic material being the essence of living systems and the environment being the necessary milieu in which the potential for life is realized. Pearson and Manolio define gene–environment interactions rather more narrowly, as “modification of genedisease associations in the presence of environmental factors.”21, p 1337 Other expressions of the idea are broader. For example, Stephens and Humphries depict
46894_CH07_139_158.qxd
2/25/10
3:26 PM
Page 151
GENE–ENVIRONMENT INTERACTION 151
the relation between genes and environment graphically (not shown).30 Both genetic and environmental risks are graded, and only in the zone of overlap when both are sufficiently high does premature CHD develop. Khoury and colleagues specified the task of genetic (only later, genomic) epidemiology “to identify which genetic factors interact with which environmental factors in the pathogenesis of disease. The traditional debate of nature versus nurture (or genes versus environment) is replaced by an active search for both genes and environments in disease.”1, p 126 Further, “failure to consider environmental components of the disease, in addition to measurement of the susceptibility genotype, may lead to erroneous inferences concerning the role of genes in disease etiology.”1, p 152 Sing and colleagues’ schematic representation of the relation between genes and environment in predicting CAD (Figure 7-2) was updated in 2003 to incorporate several complexities: Each of the four intermediate traits is shown to have direct environmental influences; the effects of the traits are age dependent, with lipid metabolism having impact at early ages and hemostasis at later ages; for each of the traits, a two-way relation is shown between phenotype and intermediate traits or genotype; and potential for divergence is depicted in the subsequent course of disease development when one of two individuals of similar phenotype at mid-life experiences a distinctive environmental change.31 The “overused, simplistic view that the genome produces an independent, isolated and fixed one-way flow of information from genome to phenotype” has been superseded by this new perspective:31, pp 1191–1192 The phenotypic measures of health are constantly being shaped, changed, and transposed as a consequence of the epigenetic networks of cellular and organismal dimensions that evolve over the lifetime of the individual. . . . Few genetic studies of the CVDs recognize the realities of the dynamic relationships between an individual’s genotype, his/her history of exposures to environmental agents, such as smoking, a high-fat diet, or a statin drug, and the contemporary phenotype in predicting phenotypic outcomes for a future point in time and a particular environmental niche. Finally, regarding the relation between environmental and genetic approaches to understanding disease, they include in a list of “imperative” considerations for geneticists the following:31, p 1194 In the pregenome era, environmental factors were considered to be the major predictors of diseases. In the postgenomic era, genetic fac-
tors have supplanted rather than complemented the environmental approach. We must return to placing equal emphasis on the role of environment and the interactions with the newfound genetic factors. Cases in Point Two examples illustrate gene–environment interactions in which the environmental factor is a common behavior––smoking, and alcohol consumption. The apolipoprotein E (apoE) gene has three common alleles, E2, E3, and E4, with differing effects on blood lipids and occurrence of CHD. The lowest risk is associated with the E2 allele, and highest risk is associated with the E4 allele, relative to the homozygous genotype E3E3. Smoking, with other CHD risk factors, has often been addressed in studies of apoE genotype by adjustment to assess the independent effect of the genotype. However, Humphries and others observed that the association between CHD risk and the E4 allele was systematically different between smokers (hazard ratio 2.79) and ex-smokers (hazard ratio 0.74) relative to never-smokers.30 Thus, although smoking is associated with increased risk of CHD in all apoE genotypes, the E4 allele especially increases the risk for smokers. This interaction was obscured in previous analyses but is demonstrated here because the association was studied within each of the smoking categories, and such an approach is not feasible without sufficient sample size in each stratum of the population, as well as awareness of the possible interaction. Alcohol intake was studied within a casecomparison study of myocardial infarction conducted in the setting of the World Health Organization MONICA Project centers in Ireland and France (Étude Cas-Témoin de l’Infarctus du Myocarde, or ECTIM).32 A gene that controls activity of the cholesteryl ester transfer protein (CETP) gene, prominently involved in HDL-cholesterol metabolism, has two alleles, B1 and B2. The B2 allele was found to interact with alcohol consumption in determining blood concentrations of both CETP and HDL-cholesterol. Among men who drank less than 25 g/day of alcohol, there was no effect on the blood levels of HDL-cholesterol and a weaker effect on CETP concentrations than at higher levels of intake. Among men drinking 75 g/day or more, the HDL-cholesterol concentration was greater by 13% among those with both B1 and B2 alleles (heterozygotes) and greater by 30% in the B2B2 homozygotes. The odds ratio for myocardial infarction for the B2B2 genotype versus B1B1 or B1B2 was 0.34 (95% confidence limits 0.14–0.83) for those drinking 75 g/day of alcohol or more; it was 0.56 (0.22–1.47) at 50–74 g/day intake and was not
46894_CH07_139_158.qxd
152
2/25/10
3:26 PM
Page 152
CHAPTER 7 GENES AND ENVIRONMENT
reduced at lower levels of alcohol intake. (The level of 75 g/day corresponds to more than 3 oz ethanol or nearly eight drinks per day.) Although not wholly consistent with other studies of this genotype, it appears to confirm findings among heavy drinkers in a Finnish study and to support the concept of a gene– environment interaction involving HDL-cholesterol metabolism and alcohol consumption. These are instructive demonstrations of gene– environment interactions that support the admonition of Khoury and others, as previously shown. Interpretation of genetic studies and comparing genetic risks of CVD between populations or subgroups may be misleading unless relevant environmental characteristics— which may differ within or between populations—are taken into account.
CARDIOVASCULAR APPLICATIONS OF GENOMIC EPIDEMIOLOGY More general consideration of contributions from genomic epidemiology (to use Khoury’s term) can be illustrated here by reference to earlier work as well as to the recent American Heart Association Scientific Statement, cited previously, and others.20 The text Genetic Factors in Coronary Heart Disease (1994) assembles the work of some 50 investigators and addresses a broad range of topics as investigated up to that time: family studies of CHD, animal and human studies of risk factors, monogenic traits affecting CHD incidence, molecular approaches to clinical research, studies of vessel wall processes, and clinical, preventive, and public health actions.33 This material remains valuable as background to more recent studies, albeit the “genomic revolution” has greatly expanded concepts and strategies in this field, as discussed previously. The recent review from the American Heart Association synthesizes findings to date in three categories of cardiovascular conditions: atherosclerosis and myocardial infarction, elevated cholesterol and other lipid disorders, and blood pressure and hypertension.20 This and other sources provide some insight to current knowledge regarding genomics and atherosclerosis, coronary heart disease, and stroke. Atherosclerosis and Coronary Heart Disease A familial pattern of occurrence of atherosclerosis has been emphasized, especially when disease appears at early adult ages. Twin studies have shown the identical twin of a case to have a higher relative risk of atherosclerotic CVD than the fraternal twin of a case, and offspring of affected parents have a greater risk than
those of unaffected parents, especially when the parental disease occurred at an early age. Several forms of subclinical atherosclerosis have also been found to have familial occurrence, including carotid artery wall thickness and calcification of either coronary arteries or the abdominal aorta. Several chromosomes have been identified as associated with CHD in linkage analyses within families, but without any of the implicated regions being strongly replicated in multiple studies. Mixed findings following up on an Icelandic study were judged to support as a “strong candidate gene for MI or stroke” a gene designated ALOX5AP, which is related to a protein involved with lipid metabolism. A number of other nonreplicated findings have been reported. Similarly with other candidate genes, findings across studies have been inconsistent. Genome-wide association studies have had some positive findings, but efforts to replicate them had not yet been reported. The many genes identified through a large number of studies to date have offered little understanding of genetics of atherosclerosis and MI. Other possible genetic contributors to increased risk for ischemic heart disease were reviewed in 2005 under the categories shown in Table 7-3.34 Some 25 genes were suggested as being related to risk of CHD, and the conclusion was consistent with the theme of other reviews:34, p 676 . . . much more investigation is required before definitive conclusions can be reached and genetic screening implemented. Because of the complex etiology of CAD/CHD, genes can influence its development at many levels. New genetic associations are being determined all the time, and much is yet to be learned. This fascinating topic will remain timely for many years to come. Could genetic factors be protective against atherosclerosis? This question was the focus of a working group report to the National Heart, Lung and Blood Institute in 2000.35 The hypothesis was discussed that genetic mechanisms might be identified as having protective effects, which could be investigated in laboratory, clinical, and population research. Several recommendations for a research program on this topic were presented. One recent example of such research concerns the paraoxonase 1 (PON1) gene, whose function is characterized as protecting against oxidative stress and thus atherosclerosis.36 Some 1400 patients undergoing diagnostic coronary angiography, the great majority having CVD at baseline examination, were followed for 3–4 years and major cardiovascular events were identified. Of three genotypes, about 45% of participants were in each of the
46894_CH07_139_158.qxd
2/25/10
3:26 PM
Page 153
CARDIOVASCULAR APPLICATIONS OF GENOMIC EPIDEMIOLOGY 153
Table 7-3
Categories of Genetic Contributors to Increased Risk for Ischemic Heart Disease
Inherited disorders associated with an increased risk for CAD/MI Lipid abnormalities Familial hypercholesterolemia (FH) Familial defective apolipoprotein B-100 (FDB) Familial combined hyperlipidemia (FCHL) and familial hypertriglyceridemia (FHTG) Atherosclerosis susceptibility (ATHS)/atherogenic lipoprotein phenotype (ALP) Inherited CAD Coronary artery disease, autosomal dominant, 1(ADCAD1) Genes associated with an increased risk of CAD/MI Genes associated with lipid metabolism Apolipoprotein E gene lipoprotein lipase (LPL) gene Apolipoprotein B gene Low-density lipoprotein genes Genes associated with vascular homeostasis Endothelial cell nitric oxide synthase (ecNOS) gene Angiotensin converting enzyme (ACE), angiotensin II type-1 receptor (AT1R) and angiotensinogen (AGT) genes Aldosterone synthase gene Genes associated with hemostasis GPIa gene Glycoprotein IIb/IIIa platelet receptor genes Thrombospondin genes Factor V Leiden allele (factor V Arg506Gln) and prothrombin variant G20210A Plasminogen activator inhibitor-1 gene Metabolic factors MTHFR gene Genes associated with inflammation Interleukin-6 gene Other genes associated with increased or decreased CHD risk Alcohol dehydrogenase gene (ADH) Source: Data from Journal of Molecular and Cellular Cardiology, Vol 39, MA Nordlie, LE Wold, RA Kloner. © 2005 Elsevier Ltd.
two more common groups and about 10% were in the remaining, highest-risk category. Event rates were associated both with PON1 genotype and with paraoxonase activity levels measured in serum samples. This was taken as “direct evidence for a mechanistic link between genetic determinants and activity of PON1 with systemic oxidative stress and prospective cardiovascular risk, indicating a potential mechanism for the atheroprotective function of PON1.”36, p 1265 Blood Lipids and Blood Pressure It has been estimated that half or more of the variation in serum lipids within a population can be attributed to genetic variation.20 Familial hypercholesterolemia, caused by mutations of the LDL-cholesterol receptor, is one of several specific disorders with demonstrated patterns of inheritance. About 700 such mutations had been identified by 2007. It is presumed that common genetic variation is the cause of most
lipid abnormalities, but the specific genes have been elusive. The picture regarding blood pressure is much the same. In view of the many regulatory systems controlling blood pressure levels, it might be expected that a great many genes would be involved in determining these levels. Altogether some 30% of the population variability of blood pressure has been attributed to genetic factors.20 Rare disorders of blood pressure regulation have been associated with specific genotypes. However, genome-wide linkage analysis in studies of blood pressure levels has not revealed consistent findings, and several regulatory mechanisms and their interactions have yet to be studied at all. Still, occasional observations stand out as promising for further investigation. An example is identification by the collaborative Family Blood Pressure Program of a possible susceptibility gene on chromosome 2, the solute carrier family 4 (sodium bicarbonate cotransporter), member 5 gene (SLC4A5).
46894_CH07_139_158.qxd
154
2/25/10
3:26 PM
Page 154
CHAPTER 7 GENES AND ENVIRONMENT
Stroke A review of genetics and ischemic stroke identifies 53 separate studies of twins or families.37 The pattern of stronger concordance of stroke between monozygotic than dizygotic twins was observed. In both casecontrol and cohort studies, a family history of stroke was associated with increased risk, especially among study subjects younger than 70 years of age. Results of 9 of the 17 cohort studies are summarized in Figure 7-3. Odds ratios among these studies varied in strength and significance, with a summary value of 1.30 (95% confidence interval, 1.2–1.5). Overall, there was marked heterogeneity among the studies, with stronger associations found in the smaller studies. Only about one-half of them distinguished ischemic from hemorrhagic stroke, despite the importance of differences in known causal factors between them. This limitation is addressed in a second review in which several discrete disorders are described in which stroke is a feature, whether affecting small or large vessels.38 One or more specific genes is identified with each of 10 such conditions that, however, are quite rare. Regarding common ischemic stroke, one gene, phosphodiesterase 4D (PDE4D), had been identified that influences endothelial smooth muscle proliferation. With the familiar qualifications regarding lack of replication, weak effect, and com-
plexity of interpretation, the authors summed up this finding as follows:38, p 10 The PDE4D gene contributes to only a minority of strokes, and its association with stroke needs to be replicated in other independent populations. Nevertheless its identification is “proof of principal” [sic] that taking the genetic approach to understanding and eventually treating stroke is sound. The optimism expressed here is counterbalanced again by cautionary notes from two case-control studies of coronary heart disease—one a large study focusing on lipid-related genes and myocardial infarction, the other seeking to replicate relationships found in previous studies appearing to link 85 variants in 70 genes with acute coronary syndrome (ACS).39,40 The first study was designed to overcome the common limitation of study size by including 4685 cases and 3460 controls, all genotyped for each of six polymorphisms of four lipid-related genes. Little correspondence was found between the strong effects of these polymorphisms on blood lipids, on the one hand, and absence of such effects on myocardial infarction, on the other. The authors concluded:39, p 1011 . . . the present findings emphasize the need to assess the relevance of genes to disease risk di-
Stroke/Subjects Positive FHx
Negative FHx
OR
95% CI
Lindenstrom et al 1993
142/2652
554/10348
1.00
0.8–1.2
Morrison et al 2000
86/4081
175/9955
1.20
0.9–1.6
Wannamethee et al 1996
54/1002
224/6681
1.64
1.2–2.2
Jousilahti et al 1997
47/754
406/13617
2.16
1.6–3.0
Khaw and Barrett-Connor 1986
18/1041
34/2374
1.21
0.7–2.2
Welin et al 1987
21/192
36/597
1.91
1.1–3.4
Berger et al 1998
10/1960
29/10867
1.92
0.9–3.9
Kobayashi 1997
15/509
4/424
3.19
1.1–9.7
Brass and Shaker 1991
25/31
26/36
1.60
0.5–5.1
TOTAL
418/12222
1488/54899
1.30
1.2–1.5 0.1
Significance
p 0.00001
Heterogeneity
p ⴝ 0.0001
1 Odds Ratio (95% CI)
10
Figure 7-3 Odds for Subsequent Stroke in Subjects with a Positive Family History (FHx) of Stroke in Cohort Studies (Ordered by Variance). Source: Reprinted with permission from Stroke, Vol 35, E Flo(ss)mann, UGR Schulz, PM Rothwell. © 2003 American Heart Association. Inc.
46894_CH07_139_158.qxd
2/25/10
3:26 PM
Page 155
CURRENT ISSUES 155
rectly in studies involving very much larger numbers of cases than has hitherto been customary (especially if gene-gene and gene-environment interactions are to be assessed appropriately reliably), rather than just extrapolating from the observed effects of genotypes on plasma lipoprotein levels, or on other single traits that lie on (or close to) a causal pathway to disease. The second of these studies, including 811 cases and 650 controls, failed to identify more than a chance frequency of positive associations between the previously reported genetic variants and ACS: “Our null results provide no support for the hypothesis that any of the 85 genetic variants tested is a susceptibility factor for ACS. These results emphasize the need for robust replication of putative genetic risk factors before their introduction into clinical care.”40, p 1551
CURRENT ISSUES Genomic epidemiology has engendered high expectations in many areas, including its application to cardiovascular diseases. Yet, as Palmer and Cardon stated in 2005, “Explosive growth in technical capacity and genomic knowledge has been tempered by initial failures to find genes for complex phenotypes using any strategy and our statistical methods and informatics capabilities lag far behind our ability to produce huge amounts of genomic data.”12, pp 1230–1231 Scaling Up Study Size The tyranny of small effects from each of many genes, conditional on environmental influences over the life course, poses difficulties that appear to require a radical change in research strategy. Initiatives to establish very large biobanks of individual data have been undertaken in the United Kingdom, Japan, and Sweden.41 Manolio and others report the recommendation of an Expert Panel for the National Human Genome Research Institute to establish a new cohort, “broadly representative of the US population” and “selected to represent the entire human lifespan at the time of their entry into the cohort” with “periodic re-examinations and annual follow-up for major disease outcomes.”41, p 816 A cohort size of 500,000 was recommended that would potentially yield more than 4500 cases each of diabetes, stroke, and heart failure and more than 11,000 cases of myocardial infarction in 5 years. Such a study is imaginable, whether or not it is feasible, and if actually implemented with adequate documentation of criti-
cal environmental exposures—such as detailed dietary assessment on all individuals at entry and follow-up—it might be embraced by chronic disease epidemiologists generally and not only by those interested primarily in genomic epidemiology. The possibility that biobanks could pool data internationally adds to the potential power of such a research enterprise, and beginning efforts of this kind were already in place by 2005.15 A more immediate prospect in the United States is the expansion of a genomics component within any of several large cohort studies already in place. To develop recommendations in this area, the Working Group on Genome Wide Association in NHLBI Cohorts met in 2005.42 The report noted that “the two worlds of molecular genetics, which is producing reliable and comprehensive genetic markers for highthroughput genome-wide genotyping, and population-based and clinical epidemiology, which is expert in defining and measuring phenotypes and disease outcomes, must be wedded.”42, p 2 Combining data across studies was anticipated to have difficulties due to between-study differences in rigor of phenotype ascertainment. Issues of cohort selection, informatics and data management needs, statistical analysis, data sharing, access, consent, confidentiality, and reporting were addressed, as well as approaches to genotyping, and sample acquisition. The latter point was expanded to note the need to measure not just molecules but also pathways in individuals, requiring that protein, lipid, carbohydrate, and nucleic acid data be collected, perhaps repeatedly. Technical aspects of the work to undertake such studies could pave the way for the large cohort study discussed previously. Public Health Applications Public health applications of genomics have been under development by the National Office of Genomics and Disease Prevention at the CDC for more than a decade.43 Program activities include the Human Genome Epidemiology Network (HuGENet), Evaluation of Genomic Applications in Practice and Prevention (EGAPP), the Family History Public Health Initiative, and Public Health Genomics Capacity Building, among other initiatives. The mission of the Center is to integrate genomics into public health research, policy, and programs. Goals include assessment of the role of family history in determining risk and for disease prevention and evaluation of genetic tests. The strategic role of this program is to close the gap between gene discovery and public health applications in prevention, early detection, and treatment.
46894_CH07_139_158.qxd
156
2/25/10
3:26 PM
Page 156
CHAPTER 7 GENES AND ENVIRONMENT
Genetic Testing Although family history has been seen to be regarded as a presently applicable clinical and public health tool, applications of genomics in genetic testing have prompted concern. Issues in this arena are the focus of EGAPP, noted previously, and led to harsh criticism in a Science editorial in 2006:44, p 1853 At worst, genetic testing errors can kill; at best, they result in poorly spent health care dollars. Moreover, should the public begin to question the accuracy of genetic tests or insurers begin to question their validity, “personalized medicine” will be nothing more than a postscript on the pages of medical history. We need sensible regulation to secure the future of genetic medicine. In the absence of regulation, “direct-to-consumer marketing of genomic disease profiles seems to have escaped the careful vetting that accompanies the introduction of new biomedical technologies. . . . There remains a fundamental concern about the validity of many of the tests. . . . Assessing and comparing genetic testing quality between laboratories is not straightforward.”45, p 1353 The prospect of medical practice tailored to genetic individuality of patients triggers enthusiasm in much of the literature highlighted here, but from a public health perspective has a possible adverse effect as well. Specific genetic determinants of risk could potentially be evaluated once an individual was found, for example, to have an elevated cholesterol concentration. A concern is that genetic heterogeneity of populations may be emphasized to such a degree that general recommendations lose acceptance in favor of individualization of all interventions, perhaps requiring genotyping before action is taken on behalf of any individual.46 In this circumstance, the concept of the population as a whole could be lost in an atomized view in which any specific intervention is justified only on a highly individualized basis. The implications for public health of dismantling population-wide recommendations warrant careful consideration. Managing Expectations Summarizing the prospects for genomics and the work required to realize them, Gwinn and Khoury conclude as follows:47, p 21 Successful completion of the Human Genome Project has raised public expectations that research findings will translate quickly into health benefits; however, the gap between biomedical
research and clinical and public health application seems wider than ever. Public health scientists now have the opportunity to help create a broad concept of research translation that integrates genomic information into policies, programs and services benefiting the whole population. Important “signposts” along the translation highway include conducting population-based research in genomics, developing evidence on the clinical and public health value of genomic information, and integrating genomics into health practice. REFERENCES 1. Khoury MJ, Beaty TH, Cohen BH. Fundamentals of Genetic Epidemiology. Volume 22. In: Kelsey JL, Marmot MG, Stolley PD, Vessey MP, eds. Monographs in Epidemiology and Biostatistics. Oxford (England): Oxford University Press; 1993. 2. Khoury MJ, Burke W, Thomson EJ, eds. Genetics and Public Health in the 21st Century: Using Genetic Information to Improve Health and Prevent Disease. Oxford (England): Oxford University Press; 2000. 3. Khoury MJ, Little J, Burke W, eds. Human Genome Epidemiology: A Scientific Foundation for Using Genetic Information to Improve Health and Prevent Disease. Oxford (England): Oxford University Press; 2004. 4. Davey Smith G. Genetic epidemiology: an “enlightened narrative”? Int J Epid. 2004;33: 923–924. 5. Palmer LJ. The new epidemiology: putting the pieces together in complex disease aetiology. Int J Epid. 2004;33:925–928. 6. Jablonka E. Epigenetic epidemiology. Int J Epid. 2004;33:929–935. 7. Khoury MJ, Millikan R, Little J, Gwinn M. The emergence of epidemiology in the genomics age. Int J Epid. 2004;33:936–944. 8. Vineis P. A self-fulfilling prophecy: are we underestimating the role of the environment in gene-environment interaction research? Int J Epid. 2004;33:945–946.
46894_CH07_139_158.qxd
2/25/10
3:26 PM
Page 157
REFERENCES 157
9. Burton PR, Tobin MD, Hopper JL. Genetic epidemiology 1. Key concepts in genetic epidemiology. Lancet. 2005;366:941–951. 10. Teare MD, Barrett JH. Genetic epidemiology 2. Genetic linkage studies. Lancet. 2005;366: 1036–1044. 11. Cordell HJ, Clayton DG. Genetic epidemiology 3. Genetic association studies. Lancet. 2005; 366:1121–1131. 12. Palmer LJ, Cardon LR. Genetic epidemiology 4. Shaking the tree: mapping complex disease genes with linkage disequilibrium. Lancet. 2005;366:1223–1234. 13. Hattersley AT, McCarthy MI. Genetic epidemiology 5. What makes a good genetic association study? Lancet. 2005;366:1315–1323. 14. Hopper JL, Bishop DT, Easton DF. Genetic epidemiology 6. Population-based family studies in genetic epidemiology. Lancet. 2005;366:1397–1406. 15. Davey Smith G, Ebrahim S, Lewis S, Hansell AL, Palmer LJ, Burton PR. Genetic epidemiology 7. Genetic epidemiology and public health: hope, hype, and future prospects. Lancet. 2005;366:1484–1498. 16. Khoury MJ, Little J, Burke W. Human genome epidemiology: scope and strategies. In: Khoury MJ, Little J, Burke W, eds. Human Genome Epidemiology: A Scientific Foundation for Using Genetic Information to Improve Health and Prevent Disease. Oxford (England): Oxford University Press; 2004. 17. Jasny BR, Kennedy D. The human genome. Science. 2001;291:1153. 18. Schull WJ, Hanis CL. Genetics and public health in the 1990s. Annu Rev Public Health. 1990;11:105–125. 19. Sing CF, Haviland MB, Templeton AR, et al. Biological complexity and strategies for finding DNA variations responsible for inter-individual variation in risk of a common chronic disease, coronary artery disease. Ann Med. 1992;24: 539–547.
20. Arnett DK, Baird AE, Barkley RA, et al. Relevance of genetics and genomics for prevention and treatment of cardiovascular disease: a scientific statement from the American Heart Association Council on Epidemiology and Prevention, the Stroke Council, and the Functional Genomics and Translational Biology Interdisciplinary Working Group. Circulation. 2007;115:2878–2901. 21. Pearson TA, Manolio TA. How to interpret a genome-wide association study. JAMA. 2008; 299:1335–1344. 22. Rao DC, Vogler GP. Assessing genetic and cultural heritabilities. In: Goldbourt U, de Faire U, Berg K, eds. Genetic Factors in Coronary Heart Disease. Dordrecht (the Netherlands): Kluwer Academic Publishers; 1994:71–81. 23. Williams RR, Schumacher C, Hopkins PN, et al. Practical approaches for finding and helping coronary-prone families with special reference to familial hypercholesterolemia. In: Goldbourt U, de Faire U, Berg K, eds. Genetic Factors in Coronary Heart Disease. Dordrecht (the Netherlands): Kluwer Academic Publishers; 1994:425–445. 24. Kardia SLR, Modell SM, Peyser PA. Familycentered approaches to understanding and preventing coronary heart disease. Am J Prev Med. 2003;24:143–151. 25. Mani A, Radhakrishnan J, Wang H, et al. LRP6 mutation in a family with early coronary disease and metabolic risk factors. Science. 2007;315:1278–1282. 26. Sundquist K, Li X. Differences in maternal and paternal transmission of coronary heart disease. Am J Prev Med. 2006;30:480–486. 27. McCusker ME, Yoon PW, Gwinn M, Malarcher AM, Neff L, Khoury MJ. Family history of heart disease and cardiovascular disease riskreducing behaviors. Genet Med. 2004;6: 153–158. 28. Crouch MA, Gramling R. Family history of coronary heart disease: evidence-based applications. Prim Care Clin Office Pract. 2005;32: 995–1010.
46894_CH07_139_158.qxd
158
2/25/10
3:26 PM
Page 158
CHAPTER 7 GENES AND ENVIRONMENT
29. My Family History Portrait. A tool from the Surgeon General. https://familyhistory.hhs .gov/fhh-web/home.action. Accessed February 28, 2009. 30. Stephens JW, Humphries SE. The molecular genetics of cardiovascular disease: clinical implications. J Int Med. 2003;253:120–127. 31. Sing CF, Stengård JH, Kardia SLR. Genes, environment, and cardiovascular disease. Arterioscler Thromb Vasc Biol. 2003;23: 1190–1196. 32. Fumeron F, Betoulle D, Luc G, Behague I, et al. Alcohol intake modulates the effect of a polymorphism of the cholesterol ester transfer protein gene on plasma high density lipoprotein and the risk of myocardial infarction. J Clin Invest. 1995;96:1664–1671. 33. Goldbourt U, de Faire U, Berg K, eds. Genetic Factors in Coronary Heart Disease. Dordrecht (the Netherlands): Kluwer Academic Publishers; 1994. 34. National Heart, Lung and Blood Institute. Summary report. National Heart, Lung and Blood Institute Working Group on Atheroprotective Genes. March 29, 2000. http:// www.nhlbi.nih.gov/meetings/workshops/ athro_rep.htm. Accessed May 27, 2007. 35. Nordlie MA, Wold LE, Kloner RA. Genetic contributors toward increased risk for ischemic heart disease. J Molecul Cellul Cardiol. 2005; 39:667–679. 36. Bhattacharyya T, Nicholls SJ, Topol EJ et al. Relationship of paraoxonase 1 (PON1) gene polymorphisms and functional activity with systemic oxidative stress and cardiovascular risk. JAMA. 2008;299:1265–1276. 37. Flossmann E, Schulz UGR, Rothwell PM. Systematic review of methods and results of studies of the genetic epidemiology of ischemic stroke. Stroke. 2004;35:212–227.
38. Bevan S, Markus H. The genetics of stroke. ACNR. 2004;4:8–10. 39. Keavney B, Palmer A, Parish S, et al. Lipidrelated genes and myocardial infarction in 4685 cases and 3460 controls: discrepancies between genotype, blood lipid concentrations, and coronary disease risk. Int J Epidemiol. 2004;33:1002–1013. 40. Morgan TM, Krumholz HM, Lifton RP, Spertus JA. Nonvalidation of reported genetic risk factors for acute coronary syndrome in a large-scale replication study. JAMA. 2007; 297:1551–1561. 41. Manolio TA, Bailey-Wilson JE, Collins FS. Genes, environment and the value of prospective cohort studies. Nature Rev Genetics. 2006;7:812–820. 42. National Heart, Lung and Blood Institute. Working Group on Genome Wide Association in NHLBI Cohorts. http://www.nhlbi.nih.gov/ meetings/workshops/genomewide.htm. Accessed June 26, 2007. 43. Centers for Disease Control and Prevention, National Office of Public Health Genomics. 10 Years of Public Health Genomics at CDC 1997–2007. Atlanta, GA: 2007. 44. Hudson KL. Genetic testing oversight. Science. 2006;313:1853. 45. Offitt K. Genomic profiles for disease risk. Predictive or premature? JAMA. 2008;299: 1353–1355. 46. Omenn GS. Comment: genetics and public health. Am J Public Health. 1996;86: 1701–1704. 47. Gwinn M, Khoury MJ. Genomics and public health in the United States: signposts on the translation highway. Commun Genet. 2006;9:21–26.
46894_CH08_159_190.qxd
2/25/10
3:27 PM
Page 159
C H A P T E R
8 Dietary Imbalance duce the impact of other chronic diseases as well) have several common features. The central public health challenge then becomes making the optimal diet the usual diet. Many opportunities for intervention have been identified, but it is less clear which combinations of policy and environmental change will be most feasible and most effective given a complex, multilevel concept of the determinants of what people eat. Levels of the individual, community, state or country, and the global arena all require consideration, and action at one or another level may be altogether ineffective without addressing the others. Striking changes have occurred in patterns of food consumption in recent decades, all of which reflect societal policies and actions that may seem remote from dietary decisions of individuals. Some changes have been purposeful on the part of leadership in the health arena, providing a basis for optimism that needed change can be brought about with enlightened and sustained commitment.
SUMMARY Dietary imbalance refers both to the relative excesses or deficiencies of particular nutrients or foods in the dietary pattern of a population or an individual and also to the relation between energy intake and expenditure. The consequences of dietary imbalance include essential contributions to atherosclerotic and hypertensive cardiovascular diseases as well as other major chronic conditions. Evidence concerning evolution of human dietary patterns from conditions of primitive life suggests that modern diets differ drastically from those to which human beings were adapted before the first and second agricultural revolutions of some 10,000 years and 150 years ago. The very recent “rich” or “affluent” diet entails many adverse consequences, considered at the level of specific nutrients and the known mechanisms by which they influence or disturb normal metabolic or physiologic function. Substantial methodologic difficulties confront population studies of diet because of the complexity of the exposures involved and obstacles to reliable measurement even in the short term, much less over lifetimes of exposure. Nonetheless, extensive observations document adverse effects of many dietary variants. Dietary components of particular interest include total and saturated fats and specific fatty acids, cholesterol, fiber, and salt. Other nutrients are addressed here only briefly, and salt, alcohol, and antioxidants are discussed in other chapters. Interest has grown in the concept of dietary patterns, in terms of foods or food groups, rather than exclusive focus on specific nutrients. Descriptions of the optimal dietary pattern for prevention of cardiovascular diseases (which would be expected to re-
INTRODUCTION Diet in Perspective It is self-evident that consumption of food is an essential and universal human activity. For every society, arrangements for a secure and reliable food supply are fundamental, and diet shapes the daily habits of every individual from birth until death. At one extreme, dietary behavior is so much a matter of habit that having access to food and providing for the times, places, and circumstances for consuming it may be largely automatic, on a day-to-day basis. At the opposite extreme, however, overcoming food scarcity transcends all other daily activities.
159
46894_CH08_159_190.qxd
160
2/25/10
3:27 PM
Page 160
CHAPTER 8 DIETARY IMBALANCE
If human dietary behavior is considered objectively, certain aspects come to light—for example, the lifetime quantity of food and drink consumed by an individual. At an average lifetime rate of only two pounds per day, we consume 700 pounds per year, or 3.5 tons per decade, or a cumulative total of about 25 tons in 70 years. “You are what you eat” does have a figurative meaning, but the body disposes of all but a minute proportion of this alimentary glut. At another level, how much cereal grain does a population require each year? An illustrative estimate of the supply for the United States would be approximately 800 kg per capita. However, in the United States, only about 10% of this amount is used for direct human consumption, and most of the remainder is converted to animal products. By contrast, the corresponding estimates for India were a total cereal
supply of about 125 kg per capita, or less than onesixth the US quantity, of which more than 80% was consumed directly.1 Different national food production policies result in very different dietary patterns. The Evolutionary Scale The evolution of human eating patterns puts the habits of contemporary societies in perspective and demonstrates sharp contrasts between the dietary conditions characteristic of most of human existence and those of very recent times.1,2 The dietary patterns of three human groups, hunter-gatherers, peasant agriculturalists, and modern affluent societies, are represented in terms of their major nutrient composition in Figure 8-1.2 The hunter-gatherer pattern persisted throughout most of human existence until the
Hunter-Gatherers
Peasant Agriculturalists
15–20
10–15 5
50–70
Modern Affluent Societies
40+
Fat
20
Sugar
25–30
Starch
12
Protein
60–75
15–20
10–15
Salt (g/d)
1
5–15
10
Fiber (g/d)
40
60–120
20
Figure 8-1 Percentage of Energy Intake Obtained from Different Food Components and Salt and Fiber Intakes of Different Human Groups. Source: Reprinted from Report of a WHO Study Group, WHO TRS 797, p 43, © 1990, by permission of Oxford University Press.
46894_CH08_159_190.qxd
2/26/10
12:25 PM
Page 161
CONCEPTS AND DEFINITIONS OF DIETARY PATTERNS 161
first agricultural revolution some 10,000 years ago. The peasant agriculturalist pattern was characteristic until the second agricultural revolution of the mid1800s in Europe. Already, however, dietary patterns of privileged classes of early Western civilization were much like those of modern affluent societies and appeared to have produced atherosclerosis, as inferred from examination of mummified coronary arteries. The most striking differences in the figure are between the latter two patterns, in which the proportions of energy intake from fats and sugar became predominant over intake from starch. Relative to the low-fat, high-fiber diet of most of human existence, modern affluent societies consume two times the fat, half the starch and fiber, and 10 times the salt. Energy from sugar, which was absent from the hunter-gatherer diet, has become one-fifth of the total energy intake in the modern diet. Changes in National Dietary Patterns The long-term history of dietary change can be traced broadly for each of the major geographic regions of the world, but, except in very recent years, data on specific dietary components have been available in few countries. Exceptions are the United Kingdom and Japan, for which dietary intakes of 200 and 100 years, respectively, can be estimated. Changes in the United Kingdom are illustrated in Table 8-1.2 In the first interval, marked increases occurred in consumption of fat, sugar, and potatoes; in the second interval, increases in fat and sugar continued, whereas wheat flour and crude fiber intakes decreased. Century-long trends in Japan, accentuated from 1950 onward, resulted in major increases in intake of animal fat and protein and milk. Japan is exceptional in having annual nutrition surveys by which to monitor the national diet closely, and the rate of change has been striking. In most of the world, much less information is available beyond food production data or broad qualitative observations.
Table 8-1
Estimated per Capita Consumption in the United Kingdom of Various Foodstuffs, 1770, 1870, and 1970 Grams per Person per Day Foodstuff 1770 1870 1970 Fat 25 75 145 Sugar 10 80 150 Potatoes 120 400 240 Wheat flour 500 375 200 Cereal crude fiber 5 1 0.2 Source: Reprinted from Report of a WHO Study Group, WHO TRS 797, p 43, © 1990, Academic Press Ltd.
The principal types of fat in the US diet over the past century—saturated fat, linoleic acid, and oleic acid—increased to 1985. This was especially true of linoleic and oleic acids, respectively the predominant polyunsaturated and monounsaturated fatty acids in the contemporary US food supply, provided mainly in salad and cooking oils and margarine. The increase in saturated fat intake over this same period was relatively less than the increase in total energy intake, so by 1985 it contributed a slightly smaller share of total energy as a percentage of calories. The food sources of fats in the United States have also changed during the 20th century, with a decrease in animal fat from 83% to 58% and an increase in vegetable fat from 17% to 42% of all fats. Shifts from whole milk to low-fat and skim milk were prominent among the changes.3 Changes in food technology have brought about even-more-accelerated dietary change in recent decades. Even in developing countries, food availability has resulted in large and rapid increases in consumable energy, as shown in Table 8-2.2 From the mid-1960s through 1997–1999 and projected to 2030, substantial increases in energy availability have been seen and are forecast in every region except the “transition countries.” The changes in diet composition that accompany economic development are clear from data obtained from the Food and Agriculture Organization and the World Bank (summarized in Figure 8-2): Complex carbohydrates are displaced by animal fat in a continuously graded pattern with increasing per capita GNP. The relation between diet and atherosclerosis was established through animal experimental studies and laboratory, clinical, and epidemiologic observations in humans, including the especially prominent contributions of Keys, as reviewed in Chapter 3, “Atherosclerosis.” A comprehensive reference cited at several points below is Diet and Health: Implications for Reducing Chronic Disease Risk, a study undertaken by the National Research Council and published in 1989.3 It provides an invaluable background to more recent developments.
CONCEPTS AND DEFINITIONS OF DIETARY PATTERNS What is the diet? Definition of diet as the total oral intake of nutrient and nonnutritive material is comprehensive but uninformative. Diet can be described in various ways—by reference to specific nutrients or particular foods consumed or avoided, overall composition of the habitual or prescribed diet in terms
46894_CH08_159_190.qxd
162
2/25/10
3:27 PM
Page 162
CHAPTER 8 DIETARY IMBALANCE
Table 8-2
Global and Regional per Capita Food Consumption (kcal per Capita per Day)—1960s, 1990s, and Projected to 2030 Region 1964–1966 1997–1999 2030 Industrialized Countries 2947 3380 3500 Transitional Countries 3222 2906 3180 Developing Countriesa 2054 2681 2980 NENA 2290 3006 3170 SSA 2058 2195 2540 LAC 2393 2824 3140 East Asia 1957 2921 3190 South Asia 2017 2403 2900 World 2358 2803 3050 a
NENA, Near East and North Africa; SSA, Sub-Saharan Africa (excluding South Africa); LAC, Latin America and the Caribbean.
Source: Data from Joint WHO/FAO Expert Consultation, WHO TRS 916, © World Health Organization 2003.
of nutrients or foods, by pattern of consumption by food groups, or by type of regional or ethnic cuisine. Nutrients The categories of nutrients and other dietary components reviewed in Diet and Health are listed in
Table 8-3.3 Calories, or total macronutrients, account for the combined energy contribution of carbohydrates, protein, fat, and alcohol. The other lipids included with fats are fatty acids, phospholipids, and cholesterol. Fat-soluble and water-soluble vitamins are distinguished chiefly because of their associations
100 Protein
Percentage of Energy
80 Carbohydrates Plus Other
60
40 Animal Fat 20
Vegetable Fat 0 <1,200
1,200–2,500
2,500–5,500
5,500–11,500
>11,500
Per Capita GNP (US Dollars)
Note: This diagram is based on an analysis of diet components , GNP, and mortality rates. Fifty-two countries satisfied selection criteria for this analysis: information was available on per capita GNP and on energy and fat consumption, and the population numbered more than 1 million.
Figure 8-2 Components of Diet in Relation to per Capita Gross National Product (GNP). Source: Reprinted with permission from Report of a WHO Study Group, WHO TRS 797, p 35, © 1990, Food and Agricultural Organization of the United Nations.
46894_CH08_159_190.qxd
2/25/10
3:27 PM
Page 163
CONCEPTS AND DEFINITIONS OF DIETARY PATTERNS 163
Table 8-3
• • • • • • • • • • • •
Dietary Components Described as Categories of Nutrients and Reviewed in Diet and Health
Calories: Total macronutrient intake Fats and other lipids Protein Carbohydrates Dietary fiber Fat-soluble vitamins Minerals Trace elements Electrolytes Alcohol Coffee, tea, and other nonnutritive dietary components Dietary supplements
Source: Data from Committee on Diet and Health, Food and Nutrition Board, Commission on Life Sciences. National Research Council, Diet and Health: Implications for Reducing Chronic Disease Risk, National Academy Press, © 1989.
with different food sources in the diet. Dietary supplements are noted because of their frequent use in some population groups and their often predominant contribution to intake of the vitamins and minerals they contain (although bioavailability may differ between these preparations and the natural food sources of the corresponding substances). Foods Dietary composition in terms of foods is illustrated by the Food Guide Pyramid of the Dietary Guidelines for
Americans, 2005 (DGA), which represents the most recent stage in evolution of federal dietary recommendations for the United States (Figure 8-3).4 The previous edition of the pyramid incorporated a quantitative dimension, in addition to itemizing the component food groups, and conveyed the relative amounts as well as types of foods in the recommended American diet: carbohydrates at the broad base; fruits and vegetables in the second tier; dairy products, animal protein, beans, and nuts in lesser quantities; and fats, oils, and sweets at the narrow peak of the pyramid.5 The 2005 version maintains the focus on foods and adds a graphic element representing physical activity but sacrifices the quantitative aspect of the preceding version.6 The importance of foods, in contrast to food constituents, is a common emphasis in cardiovascular nutrition.7 Categories of foods characterized as whole-grain, fiber, or low-glycemic (having little immediate blood glucose-raising effect) can be further described in terms of specific food sources that provide them. On the basis of review of 121 studies of dietary factors and risk of stroke or hypertension, Ding and Mozaffarian formulated what available evidence suggested to be the “optimal dietary habits for the prevention of stroke” (see below). This approach is illustrated in Figure 8-4, not only depicting the intersecting food categories but also representing the several mechanisms through which these foods influence risk of stroke.8 This combination of foods closely resembles the DASH diet (from the Dietary
Figure 8-3 MyPyramid: Steps to a Healthier You. Source: Reprinted from www.MyPyramid.gov accessed January 15, 2009.
46894_CH08_159_190.qxd
164
2/25/10
3:27 PM
Page 164
CHAPTER 8 DIETARY IMBALANCE
↓Blood Pressure
FIBER ↓Inflammation
OL
LY CE MI C
Whole-Grain Breads, Cereals, & Pastas, Oats, Brown Rice, Rye, Barley, Bran, Granola
WH
↓Weight Gain
↓LDL Cholesterol, Triglycerides
E-G
LO
IN
W
RA
-G
↓Thrombosis
↓Homocysteine
↓Glucose Levels ↑Insulin Sensitivity
Figure 8-4 Confluence of Dietary Fiber, Whole Grains, and Low Glycemic Index Foods. Source: Reprinted with permission from Seminars in Neurology, Vol 26, No 1, EL Ding, D Mozaffarian, © 2006 by Thieme Medical Publishers.
Approaches to Stop Hypertension trial—see Chapter 12, “High Blood Pressure”). A dietary pattern in which plant-based foods predominate similarly emphasizes foods rather than specific nutrients—fruits and vegetables, nuts, and whole grains.9 Protective effects of these foods are ascribed to “multiple beneficial nutrients” they contain, only a few of which have been specifically identified and evaluated in clinical trials to determine their health effects. Isolated from their natural food sources, specific nutrients may not confer the expected beneficial effects. Purified or synthetic supplements may lack benefit for the same reason. Dietary Composition The “rich” or “affluent” diet, considered as the sine qua non of atherosclerosis, has been characterized by Stamler:10, p 36 “Rich” diet is a habitual fare high in animal products and processed animal products, high in total fat, hydrogenated fat, and separated (visible) fat, high in cholesterol and saturated fat, high in refined and processed sugars, high in salt, high in alcohol for many in the population, high in caloric density, in “empty” calories, and in ratio of calories to essential nutrients, low in potassium, fiber, and often other essential nutrients, and high in total calories for a low level of energy expenditure in the era of the automobile, television, and mechanized work.
A different characterization describes the “Mediterranean diet,” long identified with low rates of coronary heart disease and widely publicized by Keys:11, p 1321S What is the Mediterranean diet? One definition might be that it is what the Mediterranean native eats. But as we know and think of it now, it is a relatively new invention. Tomatoes, potatoes and beans, for example, came from America long after Christopher Columbus discovered the New World. . . . The heart of what we now consider the Mediterranean diet is mainly vegetarian: pasta in many forms, leaves sprinkled with olive oil, all kinds of vegetables in season, and often cheese, all finished off with fruit, and frequently washed down with wine. “Popular Diets” Widespread interest in diet for health or cosmetic reasons has stimulated a proliferation of “popular diets” that vary greatly in food and nutrient composition. An extensive assessment of nutrient composition of several such diets included comparison with Recommended Dietary Allowances (RDAs), Dietary Reference Values (DRVs), or Dietary Reference Intakes (DRIs) as applicable to specific components.12 Selected nutrients are shown for illustration in Table 8-4. A very wide range of content is evident in such components as total and saturated fats, cholesterol, sodium, and potassium.
46894_CH08_159_190.qxd
2/25/10
3:27 PM
Page 165
MEASUREMENT 165
Table 8-4
Nutrition Analysis of Various Diets: Carbohydrate Addict’s, Sugar Busters!, Weight Watchers, and Ornish Diets Carbohydrate Sugar Weight Nutrient Addict’s Diet Busters! Watchers Diet Ornish Diet RDAs, DRVs, DRIs* Total calories 1476 1521 1462 1273 2000–2200 Total fat, g (% total kcal) 89 (54) 44 (26) 42 (25) 13 (9) 65 (30) Saturated fat, g 35 11 9 2 20 Monounsaturated fat, g 31 20 18 3 20 Polyunsaturated fat, g 15 9 9 5 20 Cholesterol (mg) 853 128 116 4 300 Sodium (mg) 3192 4012 2243 3358 2400 Potassium (mg) 2479 3020 3773 4026 3500 RDAs, Recommended Dietary Allowances; DRVs, Dietary Reference Values: DRIs, Dietary Reference Intakes. Note: Items in bold indicate values different from RDAs, DRVs, and DRIs. *RDAs and DRIs used are those of a female, 31–50 years old. Calculated values (DRV) are based on a 2000 kcal diet based on 30% total calories from fat, 10% total calories from saturated, monounsaturated, and polyunsaturated fat, and 15% total calories from protein. Source: Data from Obesity Research, Vol 9, Suppl 1, MR Freedman, J King, E Kennedy, © 2001 NAASO.
This review concluded with an admonition against a common perception of “diets,” that is often fostered in their promotion:12, p 34S “The American public needs to be told (and believe) that diets are not followed for 8 days, 8 weeks, or 8 months, but rather form the basis of everyday food choices throughout their life.” Dietary Prescriptions The prescriptive concept of diet is illustrated in the recommendations of the Adult Treatment Panel III (ATP III) of the US National Cholesterol Education Program (NCEP).13 Under the broad rubric of “Therapeutic Lifestyle Changes” (“TLC”), the “TLC diet” is described in Table 8-5. In addition to the macronutrient recommendations shown, specific Table 8-5
Macronutrient Recommendations for the TLC Diet Component Recommendation Polyunsaturated fat Up to 10% of total calories Monounsaturated fat Up to 20% of total calories Total fat 25–35% of total calories* Carbohydrate† 50–60% of total calories* Dietary fiber 20–30 grams per day Protein Approximately 15% of total calories *ATP III allows an increase of total fat to 35 percent of total calories and a reduction in carbohydrate to 50 percent for persons with the metabolic syndrome. Any increase in fat intake should be in the form of either polyunsaturated or monounsaturated fat. † Carbohydrate should derive predominantly from foods rich in complex carbohydrates including grains—especially whole grains—fruits, and vegetables. Source: Reprinted with permission from the Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III), National Cholesterol Education Program, NIH Publication No. 02-5215, National Institutes of Health, 2002.
guidance is given regarding saturated fats ( 7 percent of total calories) and dietary cholesterol ( 200 mg/d). If needed to achieve the cholesterol-lowering goal, addition of 2 g/d of plant stanols/sterols and 10–15 g/d of soluble fiber is recommended. The TLC diet is similar to Step Two of a former graduated approach in which lesser restrictions of saturated fat and cholesterol were the first step. ATP III includes detailed information regarding food choices that conform with the recommended nutrient composition. Dietary Imbalance Dietary imbalance is a term implicit in the characterization of the rich diet quoted from Stamler, above.10 The rich diet is, in a sense, internally unbalanced because of its characteristic high ratio of total calories to essential nutrients: Energy consumed greatly exceeds the quantity needed to provide the desired nutrients. This imbalance is compounded, externally, by the common excess of total calories consumed relative to the low level of energy expended through lack of physical activity. The term “dietary imbalance” thus directs attention both to the composition of the diet as a whole, beyond any single specific nutrient, and to the relation between total energy intake and the other factors that determine energy balance, that is, physical activity, body size and composition, and metabolic efficiency.3
MEASUREMENT Diet is complex. As an exposure, in epidemiologic perspective, its measurement concerns not only countless
46894_CH08_159_190.qxd
166
2/25/10
3:27 PM
Page 166
CHAPTER 8 DIETARY IMBALANCE
specific nutrients and food sources, but several dimensions indicated in Table 8-6: level of assessment, from individuals to whole populations; time of assessment, relative to exposure periods of interest; and availability of indirect indicators of nutritional status, including both anthropometric and biochemical measures (personal communication, Shiriki Kumanyika, 1997). Within any population, there may be wide differences among individuals in their typical dietary habits. Within individuals, diet may vary greatly from day to day or from one period of life to another. Clearly, a single inquiry on a given day may be unreliable as a measure of an individual’s current dietary behavior and still less useful for assessing an adult’s lifetime dietary history. Further, assessing not only the diet but also the resulting nutritional status may be important. The intended level of assessment influences the choice of methods. To characterize the dietary intakes of individuals, detailed and repeated assessments may be required. Dietary habits of a population may be represented adequately by obtaining much more limited information from a large number of individuals. Dietary data for individuals are typically collected through personal approaches such as dietary interviews or diaries. At the population level, existing national data may provide such indirect dietary information as estimates of food availability or purchasing patterns. The time reference of interest may be problematic. Some methods are applicable only to current diet; methods for longer-term dietary assessment are more limited in detail and even less subject to validation. As surrogate measures of diet, anthropometric and biochemical indicators may be measured. They may be more reproducible than dietary information but are influenced by factors other than diet, such as energy expenditure. Biochemical indicators may be sampled in blood, urine, or body tissues such as fat depots, hair, or toenails. Especially
Table 8-6
blood or urinary components may be highly variable within individuals and require repeated sampling for reliable estimation. Methods for personal dietary assessment are outlined and critically evaluated in both Diet and Health and the Dietary Assessment Resource Manual.3,14 Diet and Health outlines several inherent problems in dietary assessment of populations: sampling errors, nonresponse bias, reporting errors, errors related to day-to-day variability, interviewer bias, and errors due to use of food composition tables. Such tables may not indicate the actual composition of the foods reported by respondents, given that they offer only “averages of representative samples,” may not account adequately for variation in preparation of particular food items, and may not reflect current varieties of foods available to a given population or group. The Dietary Assessment Resource Manual addresses each of several methods—dietary records, the 24-hour recall, food frequency, brief dietary assessment methods, diet history, and observed intake with biochemical analysis—with comment on the strengths, weaknesses, and validity of each approach. Beyond these methods, the “gold standard” for dietary assessment at the level of nutrient intakes requires that an observer is present as individual meals are prepared and consumed; a “dummy” serving is prepared that duplicates the actual meal and quantities consumed; and the “dummy” meal is subjected to biochemical analysis. Relative to alternative methods, this approach is least likely to err in determining the composition of foods as they are actually prepared, served, and consumed. Presence of the observer may, of course, introduce error by influencing what is consumed. In any case, the cost of this approach is prohibitive for most epidemiologic studies. The Seven Countries Study (see Chapter 4, “Coronary Heart Disease”) is one of few known examples of population studies in which this method
Dimensions of Dietary and Nutritional Assessment Level Anthropometric Indicators Individuals Relative weight Groups within a population Body mass index Whole populations Body fat distribution Body composition Time Current intake Biochemical indicators Intake at a specified prior time or period Blood concentrations of lipids and other substances Change in intake over time Adipose tissue concentrations of fatty acids Urinary excretion of sodium, potassium, and other substances Source: Data from personal communication, Shiriki Kumanyika, 1997.
46894_CH08_159_190.qxd
2/25/10
3:27 PM
Page 167
MEASUREMENT 167
was used. Description of the dietary methods used in this landmark study is instructive:15, pp 1–162 Except with the U.S. railroad men, the dietary studies involved seven-day records with weighing of all foods eaten by men representing statistical subsamples of the cohorts. Repetitions of those seven-day surveys covered seasons of the year. Nutrients consumed were measured by chemical analysis as well as being calculated from tables of food composition. Special tables of food composition, based on locally analyzed foodstuffs, were used for Finland, Italy, and the Netherlands. For Greece, Japan, and Yugoslavia food composition tables for international use were modified for local food peculiarities. Materials for chemical analysis were collected in the Seven Countries Study in two ways—first, by collecting duplicate meals simultaneously with the sevenday records and, second, by making up at the end of the seven-day period a duplicate of the locally prepared foods recorded in the diary. In each case the collected foods were deep-frozen as collected and subsequently homogenized, lyophilized, and analyzed for determining specific nutrient composition (see as follows). Collection, analysis, and interpretation of dietary data were addressed by the Expert Panel on Guidelines for Use of Dietary Intake Data as advice to the Food and Drug Administration.16 These guidelines are of value both to appreciate issues in interpreting published reports and to plan studies in which dietary assessment is included. Dietary assessment and interpretation of dietary data in minority populations and in children and adolescents present particular issues reviewed in a special supplement to the American Journal of Clinical Nutrition.17–19 Thorough reviews of practical issues in dietary assessment in two major studies—a large multicenter clinical trial, the Multiple Risk Factor Intervention Trial (MRFIT), and a multinational collaborative survey of nutrition and blood pressure, the INTERMAP study—address many aspects of the design, implementation, and evaluation of dietary assessment in major epidemiologic studies.20,21 Common to both studies was detailed dietary assessment protocol development, observer training, and quality control throughout the process of data collection, management, and analysis. Challenges to reliable characterization of diet were compounded in INTERMAP, carried out in China, Japan, the United Kingdom and the United States, requiring multilingual study materials, current country-specific nutrient data bases covering 76
specific nutrients for all reported foods, and real-time international communications for ensuring quality control. A further set of issues arose from the specific research focus: the relation between blood pressure and intake of multiple macronutrients and protein, in addition to sodium and potassium. This required multiple 24-hour urinary collections from each participant. Steps taken to achieve the highest practical quality of the study in all these respects are well documented. Still, as the investigators cautioned:21, p 620 . . . it is essential explicitly to underscore a key fact about all dietary/nutrient data acquired by selfreport of free-living people: There is no “gold standard” for evaluation of validity, that is, objective truth. . . . While multiple measurements permit statistical quantitation of reliability and correction for its limitations in regression analyses of nutrient-BP relations, there are no ways to measure and adjust for limitations in validity of the data. Insofar as these supervene, they introduce “noise” . . . tending generally to result in coefficients smaller than true ones. This must be kept in mind in all work with such data. The methods of the Seven Countries Study, going beyond self-report, provide a close approximation to the elusive “gold standard” for such data but, as noted, have little practical applicability in most largescale studies principally because of cost. A different direction in efforts to understand relations between diet and health has been described by Hu and others as “dietary pattern analysis.”22 It is argued that nutrient-specific analysis faces several kinds of limitations because of the complex interplay of multiple nutrients in every food. Further, for practical purposes, people eat foods rather than nutrients, and patterns of food consumption are therefore more relevant to dietary guidelines. Recognizing a complementary relation between these two approaches, Hu concluded:22, p 8 The dietary pattern approach would not be optimal if the effect is caused by a specific nutrient (e.g. folic acid and neural tube defect), because the effect of the nutrient would be diluted. Therefore, the dietary pattern approach may be more useful when traditional nutrient analyses have identified few dietary associations for the disease (e.g. breast cancer). On the other hand, when many dietary associations have been demonstrated for the disease (e.g. CHD), dietary pattern analysis may also prove to be useful because it goes beyond nutrients and foods, and examines the effect of overall diet.
46894_CH08_159_190.qxd
168
2/25/10
3:27 PM
Page 168
CHAPTER 8 DIETARY IMBALANCE
Methodological issues in dietary pattern analysis include approaches to identifying meaningful patterns, examples being the “Western diet” or “prudent diet,” and the extent to which these can be formulated a priori and tested for associations in population data, in contrast to merely exploratory findings being derived through post hoc analysis. Schulze and Hoffmann applied an approach that begins from prior evidence to determine patterns to be sought in a new investigation; the approach continues with exploratory analysis to determine presence and relationships of the postulated patterns, thus bridging the two alternatives.23 Finally, proceedings of a workshop convened in 2007 include several reports addressing measurement of the “food and physical activity environment.”24 Definition of the concept of “food environment” for purposes of the workshop points to aspects of food availability, price, and quality as well as access to restaurants and grocery stores, within a given community setting. Tentativeness of this definition was evident:24, p S82 “It is important to note that no standard definitions of the food and physical activity environments have been developed, and their boundaries are not clearly established. This lack of standard definitions is likely due to the embryonic nature of the field and the absence of solid conceptual underpinnings.” Workshop recommen-
dations to develop measurement tools and protocols regarding the food environment are especially germane, from the perspective of measurement of diet and its population-level determinants:25, p S185 • A common core of measures should be developed and disseminated. . . . • An electronic repository of field-tested, reliable, and validated measurement tools should be developed . . . • Federal, state, and local sources of policy, environmental, and geographic data on the food and physical activity environments should be collected after a consistent protocol is developed, adopted, and made freely available.
DETERMINANTS Especially when viewed from a population perspective, there are many determinants of dietary behavior that influence the ultimate nutritional status of individuals. Many of these are indicated in Figure 8-5: food science, manufacture, distribution, purchasing, preparation, consumption, and consumption outcomes.26 These are depicted as elements in the food chain, which is subject to several major influences listed below. It becomes readily apparent that the next meal is in only a
The boxes depict the major elements in the food chain from the food science base to the outcomes of food consumption. Major influences on the food chain, each with an impact on several of the food chain steps, are listed. Food Science Base
Food Manufacture
Food Distribution
Food Purchasing
Food Preparation
Food Food Consumption Consumption Outcomes
Nutrition Biochemistry Preservation Genetics
Agriculture Synthesis Processing Additives Modifiers Hybridizers Mass preparation
Wholesale Retail Productspecific Routespecific
Cost Culture Advertising Knowledge Health
In-house Restaurants Institutions
Socialization Education Nutritive value Health Culture Taste Cost Mood
Pleasure Health Deficiencies Surpluses
Major Influences on the Food Chain: Agribusiness Conglomerates Grocery Chains Advertising Profitability Special Interest Groups Media
Government Agencies Health Professionals General Education Nutrition Education Food Science Education Culinary Education
Figure 8-5 Eating Pattern Determinants. Source: Reprinted from Report of the Expert Panel on Population Strategies for Blood Cholesterol Reduction, NIH Publication No 90-3046, National Institutes of Health, 1990.
46894_CH08_159_190.qxd
2/25/10
3:27 PM
Page 169
DETERMINANTS 169
limited sense a matter of choice. The familiar expression, “Making healthy choices the easy choices” suggests a level of personal discretion—or to some, personal responsibility—for choices that have largely been dictated elsewhere. The diet of the individual is strongly, and in some circumstances completely, determined by external factors such that one may have little awareness or control over choices of food. An analogous representation of the complex determinants of “food choice” appeared in an AHA Scientific Statement concerning implementation of AHA’s pediatric and adult nutrition guidelines.27 In response to concerns that guidance is lacking on how to make the recommended dietary choices throughout the lifespan, the authors of the report created a schematic multilevel view from the individual to the family, the microenvironment, and the macroenvironment. Each level has its own determinants, as shown in Figure 8-6. The Scientific Statement identifies potential strategies to be applied at each level. The microenvironment corresponds to the food environment discussed above and is characterized, for example, by availability of “competitive foods and beverages” available for purchase in schools. These are foods and beverages that are sold at school but
• Local Community • School Settings • Worksites • Restaurants & Fast Food Outlets
are separate from USDA school meal programs.28 Competition between these items and more nutritionally favorable school meals constitutes an important feature of the food environment for youth. A study of the City and County of Baltimore, Maryland, identified racial and economic differentials in the food environment, such that predominantly Black and low-income neighborhoods had types of stores, and selections within similar stores, that provided less access to healthy foods.29 A healthy food availability index was devised for this assessment, based on the Nutrition Environment Measures SurveyStores (NEMS-S), that took account of eight groups of foods relevant to cardiovascular health. At the level of the macroenvironment, broad economic policies, legal frameworks, and regulatory provisions operate at all levels of government and society, from local to global. National and global agriculture and energy policies have a decisive and deleterious impact on food. This point was argued forcefully by journalist/horticulturist Michael Pollan in an open letter of October 2008 addressed to the “Farmer in Chief,” the yet-to-be-elected President of the United States.30 To redirect policies that are fundamental determinants of the availability, affordability, and
• Economic Policies • Laws MacroEnvironment Level
• Government Policy
MicroEnvironment Level
• Technology
• Industry Relations • Media • Transportation
Family Environment Level
• Biology/Genetics • Flavor Experiences • Learning History • Demographic Factors
Individual Level
• Role Modeling • Feeding Styles • Availability • Culture • etc
Figure 8-6 Influencing Food Choice. A Multi-Level Framework for Identifying Facilitators or Barriers to Attaining AHA Dietary Recommendations. Concentric circles of influence on eating behaviors. The individual level refers to biological, genetic, demographic, and learning history influences within any person. The individual level is nested within the family environment, which includes such as role modeling, feeding styles, provision, and availability of foods, and other aspects of the home food environment. The third level, the microenvironment level, refers to the local environment or community in which the family and home are immediately nested. This includes local schools, playgrounds, walking areas, and shopping markets that enable or impede healthful eating behaviors. Level 4 is the macroenvironmental level. This level refers to broader economic policies, laws, and industry policies that operate at the regional, state, national, and international levels. The influence of level 4 factors can be persuasive and project down to individuals choices. The model recognizes the importance of both the nesting of levels within one anotherer and reciprocal influences among levels. Source: Reprinted with permission from Circulation, Vol 119, SS Gidding, AH Lichtenstein, MS Faith, et al. © 2009 American Heart Association, Inc., p 1162.
46894_CH08_159_190.qxd
170
2/25/10
3:27 PM
Page 170
CHAPTER 8 DIETARY IMBALANCE
quality of food, Pollan proposed “resolarizing the American farm,” “reregionalizing the food system,” and “rebuilding America’s food culture.” Benefits would accrue not only in food but also in energy, transportation, national security, and the US economy, so intertwined with other major sectors is food policy. Change in food availability around the world has been characterized broadly as increases in foods high in fats and sweeteners. The change is attributed to dual processes of “nutrition transition” and globalization.31 Globalization is seen as impacting food supplies through “integration of the global marketplace,” with three component processes: “(I) production and trade of agricultural goods; (II) foreign direct investment in food processing and retailing; and (III) global food advertising and promotion.”31, p 2 As a consequence, it is predicted from analysis of several case studies that, in developing countries, globalization may exacerbate uneven dietary development, expanding food product diversity in high-income markets and product obesogenicity in lower-income areas, as has already been observed in Western countries.
DISTRIBUTION To monitor levels and trends of dietary intakes for a population requires periodic surveys of appropriate population samples and consistency of methods over time. Periodic nutrition surveys in multistage probability samples of the US population began in 1971–1974, when a nutrition component was added to the health examination surveys that were initiated in 1960.32 Thus the National Health and Nutrition Examination Surveys (NHANES) completed from 1971–1974 to 1999–2000 (the first period of the nowcontinuous cycles of NHANES) could be examined in 2004 for 30-year trends in intake of energy, macronutrients, and other dietary components. Other data sources include the Continuing Survey of Food Intakes by Individuals (CSFII) as well as the Nationwide Food Consumption Survey (NFCS) that preceded it. (The CSFII has now been incorporated into NHANES.) Together, these sources provide insight into changes in national eating habits as well, which help to interpret concurrent changes in nutrient intakes. Consistency of methods is problematic, however. Changes in dietary intake methodology have been introduced, including improvement in probing questions, in visual aids for portion size estimation, and in automation of data collection. Nutrient databases have been updated so are not constant over time for the same foods, and changes in food composition have occurred as well. Overweight persons typically un-
derreport intakes, and prevalence of overweight and obesity has increased in the population. Jointly, these two effects may have led to a systematic increase in underreporting. These qualifications should be borne in mind. In addition, the data reviewed were presented for ages from 1–2 to 60–74 years, and by sex and age from 20–74 years, but not by race/ethnicity. Mean daily energy intake for men aged 20–74 was greater in 1999–2000 than in 1971–1974 by 170 kilocalories (kcal). For women, the corresponding increase was more than 300 kcal. Among adults, intake of total and saturated fats declined from 36% to 33% and from 13% to 11%, respectively. Table 8-7 extends these observations further to include the 2001–2004 data from NHANES.33 For no adult agesex group was the intake goal for either total or saturated fat being met. For males 12–15 and 16–19 years old, energy intake decreased, whereas for females, there was an increase in every age group. For children aged 3–5 and 6–11 years, kcal/d decreased from 1970–1974 to 1988–1994 and then increased toward the initial levels. However, for 1–2-year-olds, following a decline to 1988–1994, energy intake increased by more than 220 kcal/d by 1999–2000.32 Sodium intake appears to have increased substantially from 1971–1974 to 1999–2000 in every age group and for both men and women.32 For men age 20–74 years of age, the increase was from 2780 to 4127 mg/d, whereas for women the corresponding increase was from 1774 to 3002 mg. These levels of sodium intake are greatly in excess of the recommended upper limit of 2300 mg/d, or the limit of 1500 mg/d that applies to people with hypertension, African Americans, and all adults aged 40 or older.4 As a measure of overall diet quality, a Healthy Eating Index (HEI) was developed within the CSFII and continues to be monitored through NHANES.32 The HEI includes indicator categories of foods as well as the percentage of calories from total and saturated fat and intake of cholesterol and sodium. Each component contributes to the overall score, which can total up to 100. The mean score for 1999–2000 was 63.6, which is toward the lower end of the range for a rating, “a diet that needs improvement.” Scores declined in recent years for sodium, meat, and milk and were poorest for fruit—regarded by some to be the component of the American diet most in need of improvement. Fruit and vegetable consumption has lagged behind recommendations and has been monitored through the Behavioral Risk Factor Surveillance System (BRFSS), a telephone survey of a population sample in each US state.34 Five servings a day or more were reported by 19.5% to 27.1% of men and by
46894_CH08_159_190.qxd
2/25/10
3:27 PM
Page 171
DISTRIBUTION 171
Table 8-7
Mean Energy and Macronutrient Intake Among Persons 20–74 Years of Age, by Sex and Age: United States, 1971–1974 Through 2001–2004
[Data are based on dietary recall interviews of a sample of the civilian noninstitutionalized population] Sex and Age 1971–1974 1976–1980 1988–1994 Energy intake in kcals 2450 2439 2664 Male, age-ajusted1 Male, crude 2461 2459 2692 20–39 years 2784 2753 2964 40–59 years 2303 2315 2564 60–74 years 1918 1906 2104 1542 1522 1796 Female, age-adjusted1 Female crude 1540 1525 1804 20–39 years 1652 1643 1956 40–59 years 1510 1473 1732 60–74 years 1325 1322 1520 Percent kcals from carbohydrate 42.4 42.6 48.3 Male, age-ajusted1 Male, crude 42.4 42.7 48.3 20–39 years 42.2 43.1 48.1 40–59 years 41.6 41.5 47.8 60–74 years 44.8 44.1 49.7 45.4 46.0 50.7 Female, age-adjusted1 Female crude 45.5 46.1 50.7 20–39 years 45.8 46.0 50.6 40–59 years 44.4 45.0 50.0 60–74 years 46.8 48.6 52.6 Percent kcals from total fat 36.9 36.7 33.9 Male, age-ajusted1 Male, crude 36.9 36.7 33.9 20–39 years 37.0 36.2 34.0 40–59 years 36.9 37.2 34.2 60–74 years 36.4 36.8 32.9 36.1 36.0 33.4 Female, age-adjusted1 Female crude 36.0 35.9 33.3 20–39 years 36.3 36.0 33.6 40–59 years 36.3 36.4 34.0 60–74 years 34.9 34.7 31.6 Percent kcals from saturated fat 13.5 13.2 11.3 Male, age-ajusted1 Male, crude 13.5 13.2 11.4 20–39 years 13.6 13.1 11.5 40–59 years 13.5 13.4 11.3 60–74 years 13.3 13.1 10.9 13.0 12.5 11.2 Female, age-adjusted1 Female crude 12.9 12.5 11.2 20–39 years 13.0 12.6 11.4 40–59 years 13.1 12.6 11.3 60–74 years 12.4 11.8 10.4
2001–2004 2693 2697 2949 2649 2117 1886 1884 2302 1836 1622 48.2 48.2 49.5 47.1 47.3 50.6 50.6 51.4 49.6 51.1 33.4 33.4 32.1 34.1 34.9 33.8 33.8 33.0 34.6 34.0 10.8 10.8 10.7 10.9 11.0 10.9 10.9 10.9 11.1 10.6
1
Age-adjusted to the 2000 standard population using three age groups, 20–39 years, 40–59 years, and 60–74 years. Age-adjusted estimates in this table may differ from other age-adjusted estimates based on the same data and presented elsewhere if different age groups are used in the adjustment procedure. See Appendix II. Age adjustment. Notes: Numbers have been revised and differ from previous editions of Health, United States. Estimates of energy intake include kilocalories (kcals) from all foods and beverages, including alcoholic beverages, consumed during the preceding 24 hours. Individuals who reported no energy intake were excluded. In 2001–2004, only data collected in the Mobile Examination Center were used to calculate dietary intake. Standard errors are available in the spreadsheet version of this table. Available from: http://www.cdc.gov/nchs/hus.htm. Source: CDC/NCHS, National Health and Nutrition Examination Survey. Data from Health, United States, 2008. National Center for Health Statistics, p 317.
46894_CH08_159_190.qxd
172
2/25/10
3:27 PM
Page 172
CHAPTER 8 DIETARY IMBALANCE
27.3% to 35.9% of women across race/ethnicity groups. Generally, then, about 65% to 80% of US adults do not report meeting this dietary goal. Among students in grades 9–12 participating in the Youth Risk Behavior Survey (YRBS), from 20.1% to 26.6% of males and from 17.6% to 23.4% of females reported consuming five or more servings of fruits and vegetables per day, whereas roughly 75% to 85% do not.35 The difference between adolescent and adult females is striking. If it reflects a cohort difference between these two age groups, surveyed at approximately the same time, it is ominous. Several other indicators of eating habits in the United States come from the previously cited data sources or from special studies:32 • Eating away from home—50% more frequent among women and young children in the 1990s than the late 1970s, contributes more than 25% of the intake of energy and fat • Fast food restaurants—40% of males 12–59 years old on any given day; use reported by 37% of adults and 42% of children in the mid1990s • Portion sizes—increased from the late 1970s to mid-1990s for salty snacks, soft drinks, French fries, cheeseburgers, and Mexican food • Energy-dense, nutrient-poor foods—contributed 27% of calories in 1988–1994 • Beverage consumption—consumption of sweetened beverages increased from 37% to 48% of children and adolescents aged 6–17 years; more carbonated beverages consumed at these ages than fruit juices, milk, fruit ades, or drinks; increase in access at fast food restaurants, vending machines, and school cafeterias between late 1970s and mid 1990s • Snacks—by 1994, provided one-fifth of total calories in the diets of Americans, one-sixth of nutrients • Dietary supplements—used by 57% of women aged 20–74 in 1999–2000, although “regular use” was only 28% among women of all ages
CARDIOVASCULAR-RELATED EFFECTS OF DIET The complexity of diet and its possible cardiovascular health effects are reflected in an extensive literature in which one or more specific nutrients, particular foods, or dietary patterns have been stud-
ied for possible association with a potential metabolic or physiologic response or cardiovascular condition. Both observational and experimental studies contribute to the resulting literature. Several examples of dietary components and studies of their cardiovascular-related effects will serve to illustrate topics of particular interest and methods used to investigate them. First, it should be noted that a major trial of dietary intervention alone for prevention of coronary heart disease was proposed in the 1960s and led to the Diet-Heart Feasibility Study at that time. This doubleblind short-term trial of a “heart-healthy” diet in several test locations in the United States was highly successful in demonstrating feasibility. However, it was judged too costly to undertake the full-scale trial, and the desired evidence has never been obtained. Instead, a great many smaller and less definitive individual studies have been carried out over the intervening years. Fats and Total Cholesterol Concentration Evidence that links components of diet listed in Table 8-3, shown previously with risks of chronic diseases, is reviewed in detail in Diet and Health.3 The relation between dietary fat intake and blood total cholesterol concentration, for example, was investigated extensively by Keys, by Hegsted, and by others. A central question was whether, in groups of human subjects, the average change in serum total cholesterol concentration could be predicted from the average change in fat content of the diet. Keys and Hegsted both developed prediction equations based on dietary experiments in volunteers. The equations were similar in taking into account the intake of saturated fat, polyunsaturated fat, and dietary cholesterol. In the Keys formulation, D Chol 1.35 (2S P) 1.5 Z where D Chol (change in serum total cholesterol concentration) is determined by the relation between S (change in saturated fat) and P (change in polyunsaturated fat), each as a percentage of calories, and a term (Z) is added for the difference in the square root of initial and final cholesterol intake in mg/1000 kcal. Approximately 90% of the effect of dietary change on serum total cholesterol concentration was predicted by this equation. The equation expresses quantitatively the direct relation between increase in dietary saturated fat and cholesterol and increase in cholesterol concentration, and the inverse relation or decrease in cholesterol concentration with increased intake of polyunsaturated fats.
46894_CH08_159_190.qxd
2/25/10
3:27 PM
Page 173
CARDIOVASCULAR-RELATED EFFECTS OF DIET 173
Fats and Coronary Heart Disease Among the classic epidemiologic observations suggesting a relation between dietary pattern and cardiovascular disease mortality is the report of Malmros in 1950 on the experience of Nordic countries, contrasted with that of the United States, during the Second World War (Figure 8-7).36 The beginning of sharp decreases in consumption of dietary fats was followed in one to two years by a substantial decline
in coronary mortality in Sweden and Finland. Coronary mortality climbed beginning in 1942 or 1943, within a year of increased total fat consumption in Sweden, although fat consumption continued to decrease in Finland. In Norway, coronary mortality continued to decline while fat intake dropped and began to rise within a year of a marked upturn in fat consumption. In contrast, the US fat consumption continued to rise throughout this period, concurrently
200
Sweden 150
USA
100
Finland 50 Italy Norway
1935
36
37
38
39
40
41
42
43
44
45
46
47
Figure 8-7 Death Rates from Arteriosclerosis (Including Diseases of the Coronary Arteries) per 100,000 Population, Sweden, United States, Finland, and Norway, 1935–1947. Source: Reprinted with permission from H Malmros, Acta Med Scand, Vol 246, p 142, © 1950, Journal of Internal Medicine.
46894_CH08_159_190.qxd
174
2/25/10
3:27 PM
Page 174
CHAPTER 8 DIETARY IMBALANCE
with coronary mortality. Referring to the wartime period as the “lean” years, Malmros drew the implication from these observations that the “luxury” consumption of dairy products should be avoided. The Seven Countries Study has been described previously, including its detailed biochemical investigation of repeated weighted diet samples among
men in 12 of the 16 cohorts (dietary assessment was by questionnaire and food composition tables in the other four cohorts).15 For 12 of the cohorts, the fat content of the diet is shown in Figure 8-8. Saturated fat content was highest in East and West Finland, Zutphen (the Netherlands), and the United States, from 17–22%, and extremely low in Kyushu (Japan)
Kyushu 3
3
3
9%
Velika Krsna 9
24%
3
12
(19 to 30%)
Montegiorgio 9
13
3
25%
Crevalcore 10
14
9
16
27%
3
Dalmatia 32%
7
Slavonia 16
14
3
33%
Corfu 7
22
33%
4
Crete 8
29
3
40%
West Finland 19
13
3
35%
Zutphen 19
16
5
40%
4–6
40%
US Railroad 17–18
17–18
East Finland 22
14
Saturated FA
Monoene
3
39%
Polyene
Figure 8-8 Average Percentage of Calories from Fats, Men, 40-59 Years. Source: Reproduced with permission from American Heart Association Monograph, No 29, p 1–168, © 1970, American Heart Association.
46894_CH08_159_190.qxd
2/25/10
3:27 PM
Page 175
CARDIOVASCULAR-RELATED EFFECTS OF DIET 175
used in conjunction with 1977 mortality data for men aged 55–64 years. Many foods were significantly correlated with differences in coronary heart disease mortality. Of particular interest was an index of cholesterol and saturated fat intake, or CSI, calculated as:
at 3%. The Mediterranean populations tended to cluster around values from 7–10% of calories from saturated fat. The relation of dietary fats and cholesterol to coronary death and all incident coronary events was determined at the 10-year follow-up of the Seven Countries cohorts.37 As shown in Figure 8-9, a strong correlation (r 0.73) was found between dietary saturated fat intake, as a percentage of total calories, and incident coronary heart disease rates. East Finland, with the highest saturated fat intake (more than 20% of calories), exhibited especially high 10-year incidence, approximately 3000 per 10,000 men. Another international study compared variation in diet and coronary heart disease among 40 countries with more than 1 million population. Food disappearance data from the Food and Agriculture Organization (FAO) and coronary mortality data (ICD 8, codes 410-414) from the World Health Organization (WHO) were analyzed.38 Average food disappearance data for the years 1975–1977 were
CSI (1.01 g saturated fat) (0.05 mg cholesterol) In these data, the CSI correlated very strongly (r 0.98) with the Hegsted equation (similar to the Keys equation). Among the 40 countries, marked variation was found in both fat intake and coronary mortality— a sixfold range of CSI and a range of coronary mortality from close to 0 to more than 1000 per 100,000 population. The overall correlation coefficient was r 0.78, a result consistent with the Seven Countries Study, based on altogether different methods. Analysis of the food balance sheets of the FAO was undertaken for 27 countries. This permitted correlation of overall national dietary patterns in 1964–1966 with the countries’ respective coronary
Y = 10-Year Coronary Incidence per 10,000 Men
3000 E Y = 77+78X r = 0.73 2000
W
C
1000
N
M G J T
D
Z
R
B S
V K
0 0
5
10
15
20
X = % Diet Calories from Saturated Fat
Note: B, Belgrade; C, Crevalcore; D, Dalmatia; E, East Finland; G, Corfu; J, Japan; K, Crete; M, Montegiorgio; N, Zutphen; R, Rome Railroad; S, Slavonia; T, Tanushimaru; V, Velika Krsna; W, West Finland; Z, Zrenjanin.
Figure 8-9 Ten-Year Incidence of Coronary Heart Disease in Relation to Percentage of Dietary Calories from Saturated Fat, Seven Countries Study. Source: Reprinted with permission from the publisher of Seven Countries by A Keys, Cambridge Mass: Harvard University Press, © 1980 by the President and Fellows of Harvard College.
46894_CH08_159_190.qxd
176
2/25/10
3:27 PM
Page 176
CHAPTER 8 DIETARY IMBALANCE
heart disease mortality in 1975.39 The analysis focused on men aged 35–74 years. The countries were predominantly European but also included Japan, New Zealand, the United States, Canada, and Venezuela. For the 17 dietary components shown in Table 8-8, significant positive correlations with CHD mortality were found for nine (r 0.491 to 0.736), including several specific fatty acids. Significant negative or inverse correlations were found for five nutrients (r 0.384 to 0.590), including total carbohydrate (fiber), total -tocopherol equivalent (vitamin E precursors), and vitamin C. Animal protein and plant protein were about equally, and oppositely, correlated with coronary heart disease mortality.
Table 8-8
Simple Correlation Coefficients Between Average National Availability of Nutrients in 1964–1966 and Coronary Heart Disease (CHD) Mortality in 1975, for Men Aged 35–74 from 27 Countries Nutrient CHD Mortality (Men) 0.630** Total fata Saturated fatty acidsa 0.718** Palmitic acida 0.699** Stearic acida 0.690** Monounsaturated fatty acidsa 0.509** Oleic acida 0.491** Polyunsaturated fatty acidsa 0.384* Cholesterolb 0.716** Keys scorec 0.736** Total proteina 0.228 Animal proteina 0.703** Vegetable proteina 0.649** Total carbohydratea 0.569** Total -tocopherol equivalentb 0.590** Total vitamin Ad 0.196 Vitamin Cb 0.552** Energy 0.221 Note: Availability of nutrients based on food balance sheets from the Food and Agriculture Organization of the United Nations. CHD mortality rates, from the World Health Organization, are age averaged by 5-year age groups. The 27 countries are Australia, Austria, Belgium, Bulgaria, Canada, Czechoslovakia, Denmark, the Federal Republic of Germany, Finland, France, the German Democratic Republic, Hungary, Ireland, Israel, Italy, Japan, the Netherlands, New Zealand, Norway, Poland, Romania, Sweden, Switzerland, the United Kingdom, the United States, Venezuela, and Yugoslavia. a Percentage of total energy intake. b mg/4184 kJ(mg/1000 kcal). c 1.35 (2SFA PFA) 1.5 √C where SFA is percent energy from saturated fatty acids, PFA is percent energy from polyunsaturated fatty acids, and C is dietary cholesterol in mg/4184 kJ (mg/1000 kcal). d g RE/4, 184 kJ (µg RE/1000 kcal). *p 0.05. **p 0.01. Source: Reprinted with permission from J Stamler, American Journal of Clinical Nutrition, Vol 59 (Supp 1), p 152S, © 1994, American Society for Clinical Nutrition.
Dietary Cholesterol The Chicago Western Electric Study, an early cohort study of cardiovascular disease in the United States, had exceptionally high-quality dietary data, not only at baseline examinations but also on a second occasion 1 year later. This dual assessment of diet by detailed interview methods permitted taking within-person variation into account and adjusting estimates of individuals’ nutrient intakes accordingly. As a result, precision of estimates of diet–coronary heart disease relationships was greatly improved. Among 1824 middle-aged men over a 25-year period, significantly increased risks of coronary mortality were observed in relation to dietary cholesterol. The relative hazard of the fifth versus the first quintile group of dietary cholesterol intake was 1.46 (95% confidence interval, 1.10–1.94) after adjustment for other cardiovascular risk factors, including total cholesterol concentration.40 In a review of several studies of dietary cholesterol intake, Stamler and Shekelle noted an overall increase in risk of coronary heart disease of 30% (relative risk, 1.3; 95% confidence interval, 1.1–1.5), pooling results from the Chicago Western Electric Study, Honolulu Heart Program, Ireland-Boston Study, and the Zutphen component of the Seven Countries Study.41 Trans Fatty Acids Trans fatty acids are predominantly synthetic compounds.42 They are formed during partial hydrogenation of polyunsaturated vegetable oils in manufacture of such commercial products as margarine. Physical and metabolic properties of these compounds are determined by their having one or more double bonds between carbon atoms that are in the so-called trans configuration. Beginning in the 1980s, it was found in animal experiments that trans fatty acids raised blood cholesterol concentrations. Human studies in the 1990s showed increased LDLcholesterol and decreased HDL-cholesterol in response to trans fatty acid consumption. Reports of association of trans fats with coronary heart disease appeared in the early 1990s also from a small casecontrol study predominantly in men and in the large Nurses Health Study in women.43,44 In the Seven Countries Study, repeat dietary data collection in the course of follow-up permitted analysis of fat components not studied at baseline, specifically trans fatty acids and the omega-3 (or n-3) polyunsaturated fatty acids.45 In 1987, new food samples were obtained from markets in the area of each cohort designed to replicate the composition of the original samples but now to be tested for other con-
46894_CH08_159_190.qxd
2/25/10
3:27 PM
Page 177
CARDIOVASCULAR-RELATED EFFECTS OF DIET 177
stituents. The relation of the newly measured components to the 25-year mortality experience of the 16 cohorts was then evaluated. The 18-carbon trans (monounsaturated) fatty acid, elaidic acid, the 12- to 18-carbon saturated fatty acids, and total fat intake were all significantly correlated with coronary heart disease mortality, as were total saturated fatty acids and dietary cholesterol. An international case-comparison study (EURAMIC) used another method to assess trans fatty acid intake—needle aspiration of tissue fat from the buttock—for both newly hospitalized cases with first myocardial infarction and comparison subjects sampled from the same geographic area.46 Ten centers in nine countries collaborated to recruit a total of 671 cases and 717 comparison subjects, all males. Overall comparisons by center indicated wide variation in the proportion of tissue fatty acids of the trans type, but there was no difference between case and comparison subjects when pooled across all centers (multivariate odds ratio, 0.97; 95% confidence limits, 0.56–1.67). For some time, doubt remained whether there was any adverse effect attributable to these substances.3 Review of available data under the Food and Nutrition Board of the Institute of Medicine (IOM) led to a report in 2002 noting concern that trans fatty acids may be “more deleterious with respect to coronary heart diseases than saturated fatty acids.”42, p 14 A 2006 review by Mozaffarian and others reached a stronger conclusion: “On a percalorie basis, trans fats appear to increase the risk of CHD more than any other macronutrient, conferring a substantially increased risk at low levels of consumption (1 to 3 percent of total energy intake).”47, p 1604 Effects on serum lipids, inflammation, endothelial-cell function, and other effects were reviewed and projections were made of the potential improvement in blood lipid profile and reduction of coronary heart disease risk by eliminating trans fats from the diet. Because some trans fats occur naturally in low concentrations of dairy products and meats, they cannot be eliminated totally from the diet without risk of inadequate intake of essential nutrients. The IOM report recommended that trans fatty acid consumption be reduced as far as possible without compromising a nutritionally adequate diet.42 By 2006, the US Food and Drug Administration had acted to require labeling of foods and supplements regarding their trans fat content. The New York City Department of Health and Mental Hygiene took action by requiring restaurants and food suppliers to provide foods free of industrially produced trans fats.
Omega-3 Fatty Acids or Fish Consumption The question of whether fish consumption or other sources of omega-3 fatty acids, including fish oil supplements, is protective against coronary heart disease has been addressed in many studies, especially in the 1980s and subsequently.3 The hypothesis is traced to reports in 1971 and 1980 that Greenland Eskimos experienced a low frequency of coronary heart disease despite consumption of large quantities of marine meat and fat. The Eskimo diet contained less of both saturated fat and cholesterol than the Danish diet with which it was compared. Several specific fatty acids also appeared in different quantities in this diet than in that of the Danes. Particular attention focused on the relatively increased content of omega-3 polyunsaturated fatty acids (usually denoted w-3, or n-3, PUFAs), so called because the first of the multiple double bonds in each of the fatty acids of this class occurs at the third carbon atom in the chain. Specific compounds in this class are alphalinolenic acid (LNA), eicosapentaenoic acid (EPA), and docosahexaenoic acid (DHA). Their food sources differ, LNA being found in plant sources (tofu, soybean, and others) and EPA and DHA being found in marine sources. The immediate effects of consuming fish and fish oil include reducing triglycerides, partly by displacing saturated fatty acids in the diet, and producing some decrease in LDLcholesterol as well.3 The eicosanoid compounds include thromboxanes and prostacyclins, which are related to the functions of platelets and endothelium, the innermost layer of the blood vessel wall. w-3 PUFAs have anticoagulant effects in experimental studies. It is not clear from population studies in the Netherlands (Zutphen), Japan, and Norway that increased fish consumption, even though associated with measured differences in blood concentrations of w-3 PUFAs, consistently affects hemostatic function. Evidence regarding effects of omega-3 fatty acids on risk of coronary heart disease was reviewed in the mid-1990s.48,49 At that time, the epidemiologic evidence was found to be consistent with the idea that consuming fish once or twice a week may decrease risk of coronary heart disease in comparison to rarely or never eating fish, at least in populations with high mean serum cholesterol. However, many gaps remained in the body of evidence. Very little evidence was available with respect to nonfatal coronary events or diseases other than coronary heart disease. Few data were available for women or for persons with non-European ancestry. Effects of other nutrients associated with fish had not been adequately studied. Further studies were needed to establish more clearly
46894_CH08_159_190.qxd
178
2/25/10
3:27 PM
Page 178
CHAPTER 8 DIETARY IMBALANCE
the dose-response relationship with total mortality and with nonfatal as well as fatal coronary events throughout a wide range of fish consumption and the consistency of this association in a variety of populations. Trials were needed to establish firmly whether increasing the consumption of fish reduces risk of coronary heart disease. Two 2006 reviews addressed a substantial body of evidence in observational epidemiologic studies and clinical trials. Hooper and others assessed findings of 48 randomized control trials, largely of supplemental omega-3 intakes, and 41 cohort studies.50 The pooled estimate of relative risk for cardiovascular events was 0.95 (95% confidence interval, 0.82–1.12), consistent with the absence of an association. They concluded that “Long chain and shorter chain omega 3 fats do not have a clear effect on total mortality, combined cardiovascular events, or cancer.”50, p 1 Mozaffarian and Rimm investigated studies of fish or fish oil consumption and cardiovascular risk and also studies of toxic effects of fish and shellfish contaminants.51 They found, contrary to the previous analysis, that 1–2 servings per week of fish with relatively high content of EPA and DHA reduces risk of coronary death by 36% (95% confidence interval, 20–50%) and total mortality by 17% (95% confidence interval, 0–32%). It was considered that benefits of fish intake outweigh the risks, except that for women of childbearing age some fish species would be excluded because of potential toxicity of contaminants. Policy implications as well as estimated costs to consume recommended quantities of desirable fish species were included in the discussion. Differences in findings between these nearly simultaneous reviews may be due to the focus of the latter review specifically on cardiac death (fatal myocardial infarction and sudden cardiac death), and on consumption of fish rather than omega-3 fatty acid supplements, unlike the review by Hooper and others. Dietary Fiber Dietary fiber is complex carbohydrate material— nonstarch polysaccharides and lignins—that resists digestion in the alimentary tract. Food composition data for fiber have been available only recently. Cereals, fruits, and vegetables are the principal food sources of fiber, of which several types are distinguished. Many studies have now addressed the effects of dietary fiber on risks of coronary heart disease, blood lipid concentrations, and other aspects of atherosclerosis. This work is reviewed extensively in Diet and Health, in which no conclusion was reached on health effects of fiber, but more rigorous epidemio-
logic studies were proposed.3 Desirable studies would be larger, in populations with wider ranges of intake, and would use improved dietary assessment methods and nutrient databases. These conditions would permit more detailed analysis regarding the several components of fiber in the diet. Two subsequent studies have more nearly met these criteria than earlier ones. The first was the Health Professionals Follow-up Study, in which a cohort of 43,757 men aged 40–75 years completed detailed dietary questionnaires in 1986 and were followed for six years to ascertain new coronary events.52 The second was a cohort analysis of 21,930 Finnish men aged 50–69 years who were smokers at high risk of lung cancer at entry to a preventive trial of -tocopherol and -carotene supplements (the ATBC Study).53 In both studies, high-level consumers of fiber (the highest quintile group in each study) differed from low-level consumers in many respects. For example, in the ATBC study, the high-fiber group consumed less fat and cholesterol, more of vitamins C and E, two times the quantities of fruit and berries, and five times the quantity of rye products. They also exercised more. Adjustment for such factors led to results in both studies indicating significant inverse associations between maximum fiber intake and risk of coronary heart disease. In the Health Professionals Follow-up Study, it was estimated that the average reduction in risk of coronary death per 10 g increase in fiber intake per day would be approximately 19%. The corresponding result in the ATBC was a reduction of 17%. A 10 g increase in fiber intake would be from about 15 to 25 g/d for women and 18 to 28 g/d for men in the United States, in close agreement with prevailing dietary recommendations. Coffee and Tea Coffee consumption has been found to be associated with risk of coronary heart disease in some studies but not in others. A recent review, a meta-analysis, and a commentary together identify many of the relevant studies and summarize current knowledge.3,54,55 Coffee has been in use for about a thousand years, and as of the mid-1980s was consumed by 52% of the US population. It is the major source of caffeine for those who drink it, delivering between 200 and 300 mg/d. Coffee is chemically complex, however, and may contain 100 or more active substances, depending on the manner of its preparation. This consideration has led to identification of two specific components, kahweol and cafestol, which in human experiments have been found to increase
46894_CH08_159_190.qxd
2/25/10
3:27 PM
Page 179
CARDIOVASCULAR-RELATED EFFECTS OF DIET 179
blood cholesterol, particularly LDL-cholesterol. When paper filtration is used in the brewing process, these compounds are removed. Only coffee prepared by boiling and drunk without filtration would be expected to have this effect. In the absence of riskreducing effects of other constituents of coffee, this effect would be expected to increase coronary heart disease risk. Review and meta-analysis of 8 case-control and 14 cohort studies concluded that only by drinking five cups per day or more would there be a relative risk greater than 1.5, and there was doubt about the existence or size of a true effect. Further studies over a longer period, with multiple assessments of coffee intake and representing different coffee types, would be helpful. More recently, both coffee and tea consumption were investigated further in the Determinants of Myocardial Infarction Onset Study.56,57 Interview histories of coffee and tea consumption in the preceding year were recorded for hospitalized patients after acute myocardial infarction. Post-MI survival was the outcome, which was unrelated to prior coffee drinking but was greater for moderate or heavy tea drinkers than for nondrinkers. Other Foods and Nutrients On the basis of NHANES data collected in 2003–2006, Danaei and others investigated 12 major modifiable dietary, lifestyle, and metabolic risks for their contribution to mortality in the United States.58 Dietary factors included the following intakes: high trans fatty acids, low polyunsaturated fatty acids (PUFA), low dietary omega-3 fatty acid, high dietary salt, and low fruits and vegetables. (Alcohol use was also included—see Chapter 15, “Other Personal Factors.”) By including these multiple dietary factors, as well as high blood glucose and LDL-cholesterol, high blood pressure and body mass index (BMI), physical inactivity, and tobacco smoking, this analysis provided a more detailed and informative assessment than previous reports of comparative contributions of modifiable factors to mortality. In brief, for both men and women, the numbers of deaths in 2005 estimated to be attributable to dietary factors were: high dietary salt (sodium)— 102,000; low omega 3 (seafood)—84,000; high trans fatty acids—82,000; alcohol use—64,000; low fruits and vegetables—58,000; and low polyunsaturated fatty acids (in replacement of saturated fatty acids)—15,000. These numbers of deaths can be compared with those attributed in this study to tobacco and high blood pressure, 467,000 and 395,000, respectively.
Dietary Patterns The nutrient characteristics of several dietary patterns described above suggest that cardiovascular disease outcomes might differ among them. The relation between adherence to the Mediterranean diet and death from coronary heart disease, cancer, and total mortality was studied in the Greek component of the European Prospective Investigation into Cancer and Nutrition (EPIC).59 A scale was devised to score adherence to elements of this dietary pattern with a range from 0 to 9. Over a 44-month period of follow-up of more than 22,000 participants, a high score was found to be associated with lesser coronary mortality, with an adjusted hazard ratio of 0.67 (95% confidence interval, 0.47–0.94). Total mortality was also favorable in higher-scoring participants. It was noted that these favorable associations could not be detected for individual foods, only for the overall pattern. Because popular diets are often marketed and utilized for weight loss, some have been compared for their effects on weight loss as well as cardiovascularrelated outcomes. A meta-analysis of five small trials averaging 90 overweight or obese participants each for 12 months’ follow-up compared low-carbohydrate, non-energy-restricted diets with low-fat, energyrestricted diets.60 One-half to two-thirds of participants continued on the randomly assigned diets through 1 year. Weight loss at 1 year was similar between groups, but those on the low-carbohydrate diets experienced increased LDL-cholesterol levels. A single trial compared four diets—Atkins (low carbohydrate), Zone (macronutrient balance), Weight Watchers (low calorie), and Ornish (low fat) (see Table 8-5).61 Weight loss at 1 year was 2.1 kg for the Atkins diet and 3.0–3.3 for the others. Self-rated adherence on a scale from 1 to 10 began at 6 and fell to 3 or less in a similar pattern for all of the diets. Regardless of diet assignment, but correlated with degree of adherence, weight loss was associated with reduced total to HDL-cholesterol ratio and other risk factors. The Dietary Approaches to Stop Hypertension (DASH) Trial was an outpatient feeding trial in four centers, involving 436 participants.62 The trial was primarily to evaluate the effect of a diet increased in fruit, vegetables, and low-fat dairy products and reduced in saturated fat, total fat, and cholesterol, while maintaining baseline weight. An extension of the trial compared three levels of sodium content, and the primary outcome was change from baseline blood pressure (see Chapter 12, “High Blood Pressure”). The effect of the DASH diet on blood lipids included reduction in total, LDL-, and
46894_CH08_159_190.qxd
180
2/25/10
3:27 PM
Page 180
CHAPTER 8 DIETARY IMBALANCE
HDL-cholesterol, with no change in triglycerides (triacylglycerol). Change in HDL-cholesterol was small relative to total cholesterol, with improvement in the total/HDL-cholesterol ratio. It was concluded that the DASH diet reduces coronary heart disease risk. To evaluate the contribution of diet to risk of acute myocardial infarction on a broad global basis, the INTERHEART Study identified 5761 cases of acute myocardial infarction and 10,646 control subjects in 52 countries.63 Noted briefly in Chapter 4, “Coronary Heart Disease,” the study assessed dietary habits of each participant by use of a food frequency questionnaire based on 19 food groups. Frequency of consumption of foods in each group was determined, but not quantities. Two approaches to analysis were, first, to derive dietary patterns from the data by factor analysis and, second, to develop a diet risk score based on individual foods. The patterns were termed Western, Prudent, and Oriental. Degrees of conformity to each pattern were assessed. For the Western pattern, a strong positive gradient in odds ratios was found, whereas the opposite was found for the Prudent pattern. For the Oriental pattern, no relation was found between conformity to the pattern and the odds ratios. The diet risk score—which included both presumptively atherogenic and nonatherogenic foods—was strongly and positively related to odds ratios overall and separately for women and men. The authors interpreted the findings to show that “An unhealthy dietary intake, assessed by a simple dietary risk score, increases the risk of AMI globally and accounts for ˜ 30% of the population-attributable risk.”63, p 1929
PREVENTION AND CONTROL Several considerations about diet and recommendations to change it provide background for thinking about prevention and control of dietary imbalance. First, if its determinants are as depicted in Figures 8-5 and 8-6, complex relations exist among multiple influences on diet at both individual and population levels. Second, a recommendation regarding what to eat may take the form of a dietary prescription for an individual at increased cardiovascular risk or may be expressed as a policy intended for adoption by people generally. Third, for individuals or populations in optimum cardiovascular health, promoting these recommendations is prevention, in the sense of forestalling development of unhealthy dietary behavior. For individuals at increased risk, or for populations in which increased cardiovascular risk is widely preva-
lent, the same recommendations are aimed at achieving control, or restoration of dietary balance. Two kinds of questions arise: What is the optimal diet for prevention of cardiovascular events and conditions? And what interventions may be useful to support the optimal diet at individual, community or population-wide, and global levels? Given an answer to the first question, discussion of interventions at each of these levels can follow. The Optimal Dietary Pattern The idea of the “prudent” diet for prevention of cardiovascular diseases has a history of several decades. For example, the AHA has played a prominent role in developing dietary recommendations. Its first such report, published by the Central Committee for Medical and Community Program in 1961, recommended reducing blood cholesterol by decreasing intake of calories and saturated fat and adopting a habit of regular, moderate exercise.64 Advice for dietary change was especially directed at those with a family history of cardiovascular disease; with other risk factors such as elevated blood pressure, being overweight, or “sedentary lives of relentless frustration”;64, p 134 and with a prior heart attack or stroke. The recommended goal of a dietary pattern with 25–35% of total calories from fat and substitution of polyunsaturated for saturated fats was to be pursued under medical supervision. Since 1961, several further statements on diet have been provided by the AHA. The most recent update, Diet and Lifestyle Recommendations Revision 2006, was prepared by the Nutrition Committee.65 This report incorporates more comprehensive preventive principles that had evolved over the intervening years, including the broader concept of lifestyle changes. The report also distinguishes public health and clinical recommendations:65, p 83 Maintaining a healthy diet and lifestyle offers the greatest potential of all known approaches for reducing the risk for CVD in the general public. . . . Although great advances have been made in prevention and treatment of CVD through drug therapies and procedures, diet and lifestyle therapies remain the foundation of clinical intervention for prevention. Unfortunately, the latter commonly are neglected, to the detriment of patients. Rigorous application of the principles of diet and lifestyle intervention . . . will contribute significantly to risk reduction and will augment the benefit that may be obtained by other approaches. The clinical
46894_CH08_159_190.qxd
2/25/10
3:27 PM
Page 181
PREVENTION AND CONTROL 181
approach is an extension of the public health approach . . . The elements of the 2006 recommendations, not only for adults but for children over 2 years of age, were to:65, p 83 • Balance calorie intake and physical activity to achieve or maintain a healthy body weight. • Consume a diet rich in vegetables and fruit. • Choose whole-grain, high-fiber foods. • Consume fish, especially oily fish, at least twice a week. • Limit your intake of saturated fat to 7% of energy, trans fat to 1% of energy, and cholesterol to 300 mg per day. • Minimize your intake of beverages and foods with added sugars. • Choose and prepare foods with little or no salt. • If you consume alcohol, do so in moderation. • When you eat food that is prepared outside of the home, follow the AHA Diet and Lifestyle Recommendations. The recommendations were expressed as advice, or directives, to individuals. Only for dietary fats were quantitative limits suggested. Fish oil supplements were considered advisable for persons with documented coronary heart disease, and plant stanols/ sterols were considered an option, if taken daily, to lower LDL-cholesterol. Use of antioxidant supplements, soy protein, folate and other B vitamins, or phytochemicals was not recommended. Is this the optimal dietary pattern for prevention of cardiovascular diseases? It can be compared with dietary approaches in two reports, each of which claimed to present the optimal diet for prevention of coronary heart disease or stroke, respectively.66,8 Regarding coronary heart disease, a review of metabolic studies, epidemiologic studies, and dietary intervention trials concluded that:66, p 2569 . . . diets using nonhydrogenated unsaturated fats as the predominant form of dietary fat, whole grains as the main form of carbohydrates, an abundance of fruits and vegetables, and adequate omega-3 fatty acids . . . together with regular physical activity, avoidance of smoking, and maintenance of a healthy body weight, may prevent the majority of cardiovascular disease in Western populations. The authors refrained from presenting numerical targets such as those for components of fat intake
on grounds that evidence is insufficient to support specific levels of intake and that the public has difficulty acting on them in any case. Instead, it was argued that “A variety of options exist for designing attractive and heart-healthy diets, with varying amounts of fat and carbohydrates, as long as the diet embraces healthy types of fats and carbohydrates and provides an appropriate balance in energy intake and expenditure.”66, p 2575 The optimal diet for stroke prevention, based on a similar review, differs in several respects from the optimal anticoronary diet. It is based on a ranking of proposed dietary components according to the weight of evidence (“strong,” “moderate,” or “limited or insufficient evidence”) favoring each of them for stroke prevention.8 The habits supported by “strong evidence” were:8, p 19 • lower sodium intake (less than 50 mmol [1150 mg] per day) • higher potassium intake (~4 g per day dietary intake) • modest consumption of fatty fish (1 to 2 servings per week) • plenty of fruits and vegetables (~10 servings per day) • higher intake of whole grains and cereal fiber • a diet pattern rich in vegetables, fruits, and low-fat dairy products Although some elements are common to both of these “optimal” dietary approaches, prevention of stroke differs in part by referring to lower sodium and higher potassium intakes because of their important complementary effects in lowering blood pressure. The “optimal” diet for stroke prevention makes no reference to specific types of fats other than fatty fish. It does include quantitative guidance and introduces the somewhat ambiguous concept of “servings.” (Dietary Guidelines for Americans 2005 defines serving size as “A standard amount of a food, such as a cup or an ounce, used in providing dietary guidance or in making comparisons among similar foods” and portion size as “The amount of a food consumed in one eating occasion.”4, p 69) Given the public health interest in preventing not only coronary heart disease, or stroke, or cardiovascular disease, but also other diet-related chronic diseases, the IOM report, Diet and Health: Implications for Reducing Chronic Disease Risk, warrants further consideration.3 The report of the Committee on Diet and Health, though published two decades ago, can be compared with the foregoing recommendations
46894_CH08_159_190.qxd
182
2/25/10
3:27 PM
Page 182
CHAPTER 8 DIETARY IMBALANCE
as follows: total fat, 30% or less of calories; saturated fat, 10% or less of calories; cholesterol, less than 300 mg daily; vegetables and fruits, five or more servings daily; starches and other complex carbohydrates, six or more servings daily; protein, moderate levels; salt, 6 g/d or less; calcium, “adequate”; and fluoride, “optimal.” A balance between food intake and physical activity was recommended in order to maintain appropriate body weight. The Committee did not recommend alcohol consumption and advised only limited intake (less than 1 ounce pure alcohol per day) for those who do use it. Use of dietary supplements in excess of the RDA was discouraged. (“Serving” was defined by the Committee: “An average serving is equal to a half cup for most fresh or cooked vegetables, fruits, dry or cooked cereals and legumes; one medium piece of fresh fruit; one slice of bread; or one roll or muffin.”3, p 672) Individual Measures For supporting the individual dietary behaviors presented in these varied, but in most respects consistent, recommendations, a number of strategies have been proposed. In the 2006 update from the AHA, “practical tips” were offered. These took the form of instruction, or translation of the recommendations, to facilitate making the intended choices by the individual.65 Among “high-priority recommendations to facilitate adoption” of their recommendations, AHA listed several actions to be taken by practitioners: Advocate a dietary pattern consistent with AHA recommendations; encourage regular physical activity; calculate and discuss with patients their BMI; discourage smoking, or promote cessation among smokers; and encourage moderation of intake by those who drink alcohol. Three years after this update, AHA responded to the limited implementation of dietary guidelines.65 In addition to reiterating the “practical tips” from the earlier report, strategies were proposed in order to reach effectively all individuals over 2 years of age. These strategies were translations of the dietary recommendations into practical terms. Factors working against the recommended dietary pattern were also identified, at both individual and family levels. Points to be communicated in teaching and counseling about the desired dietary behaviors were listed and illustrated by example. For establishing and maintaining the preferred dietary pattern from birth, the AHA Schedule for Integrated Cardiovascular Health Promotion in Children proposes taking a dietary history at every age, from 0–2, 2–6, 6–10, and 10 years, with guidance to physicians in offering dietary advice at every regular visit throughout this period of life.67
The US Preventive Services Task Force (USPSTF— see Chapter 19, “Evidence and Decision Making”), however, concluded in 2003 that “the evidence is insufficient to recommend for or against routine behavioral counseling to promote a healthy diet in unselected patients in primary care settings.”68, p 125 For adult patients with known risk factors for cardiovascular and diet-related chronic diseases, “intensive behavioral dietary counseling” is recommended. The proposed approach for such counseling is the “5-A behavioral counseling framework” of Whitlock:68, p 126 Assess dietary practices and related risk factors. Advise to change dietary practices. Agree on individual diet change goals. Assist to change dietary practices or address motivational barriers. Arrange regular follow-up and support or refer to more intensive behavioral nutritional counseling (e.g., medical nutritional therapy) if needed. A meta-analysis of 23 trials was reported in which healthy adults were assigned to receive or not receive advice to improve their diets by reducing fat or salt intake and other measures consistent with a heart healthy diet.69 A wide range of specific intervention approaches was represented in the studies reviewed, including intervention from 1 to 50 hours over periods from 3 months to 4 years. Overall, it was judged that dietary changes occurred that were sufficient to result in modest improvements in blood pressure and total and LDL-cholesterol levels. The lead author had reached a similar conclusion from a prior meta-analysis:70 Dietary advice from health care or health promotion personnel appears to be effective in achieving modest dietary change and accompanying cardiovascular risk reduction. Dietary advice in primary care, together with public health and other populationwide policies, may present the most cost-effective strategy for prevention. The schema of Figure 8-6 anticipates this viewpoint, that population-wide policies would reinforce the effects of individual-level intervention. What is considered to be useful intervention at the community or population-wide level? Community or Population-Wide Measures Measures to promote and facilitate heart healthy nutrition at the community or population-wide level (which may reach whole populations, e.g., at a state or national level) include interventions to improve the food environment in the sense discussed above.24
46894_CH08_159_190.qxd
2/25/10
3:27 PM
Page 183
PREVENTION AND CONTROL 183
The AHA in 2006 listed a wide range of recommended actions by restaurants, the food industry, schools, and local government.65 In the 2009 report on implementing their diet and lifestyle recommendations, AHA addressed community-based interventions in some detail, including: schools; other influences on children’s eating behavior; community food access; and workplace interventions.27 Summary recommendations at the community level were linked to the schema in Figure 8-6. They focused on the need for research to evaluate implementation strategies at this level, elaborating on the intent of each of five recommendations:27, p 1169 Create a healthy food environment means serving items of high food quality in schools and at work places. Collaboration with the various components of industry responsible for the food supply will be critical to achieving this goal. Subsidize AHA-recommended food choices means creating financial and other incentives for consumers to purchase and food producers to generate nutritious foods. Market nutrition means using media to counterbalance unhealthy food messages. Empower consumers means providing more comprehensive labeling of food and portion size. Train professionals in nutrition means improving the skill level of healthcare practitioners commonly consulted for nutrition advice and enlarging the pool of individuals qualified to provide nutrition advice. Schools are a prominent focus of attention for several obvious reasons: Nutrition early in life influences both immediate and lifelong health habits; food service and access at schools is an important component of nutrition for all children and a principal source of healthy food for many of them; the school setting is a competitive arena between federally mandated food programs and commercial food and beverage sales, sometimes under profit-sharing contracts with the schools; and this is a convenient target for interventions that can potentially benefit children and families. The expanding use of preschool facilities and child-care settings outside of schools broadens the potential scope of effective interventions. Reports for the AHA identify a wide range of programs for comprehensive school health or targeted to cardiovascular health including nutrition, physical activity, and prevention of tobacco use.67 A substantial body of experience has accumulated in single- or multicomponent interventions to improve nutrition in childhood and adolescence, usually focusing on the school environment. To note just one example, the Coordinated Approach to
Child Health (CATCH) Program was a randomized control trial in 96 US schools of interventions to change the food environment and increase physical activity, targeting grades 3 to 5 (see also Chapter 21, “The Case for Prevention”).71 Fat intake from school lunches was reduced and physical activity was increased. CATCH and nine examples of other school-based intervention programs were cited from a total of 41 studies qualifying for systematic review by the Task Force on Community Preventive Services in 2004.72 The Task Force characterized the studies as varying widely in combinations of components, target age group, duration of intervention, and length of followup. Results were described as measured in terms of behavioral outcomes assessed by self-report. This latter feature was considered a major weakness, given that outcomes were “probably subject to reporting bias (e.g., social desirability—the possibility that answers may be influenced by what the respondent thinks is socially acceptable). Although reported changes were in the desired direction, they were small and are questionable because of the potential bias of selfreports.”72, p 1 The Task Force reported as follows: “School-based programs promoting nutrition and physical activity: Insufficient Evidence.” Notwithstanding this reservation, efforts to improve the school food environment continue to evolve. For example, in 2006 the Alliance for a Healthier Generation—a partnership of the AHA and the William J. Clinton Foundation—announced a joint agreement between the Alliance and five leading food manufacturers in the United States.73 Illustrating the link between schools and the broader food environment, the agreement called on industry to reformulate and innovate in their product development to “promote the consumption of fruits, vegetables, whole grains, nutrient-rich foods, fat-free and low fat dairy foods.” The adopted guidelines also “place limits on calories, fat, saturated fat, trans fat, sugar and sodium.” The announcement concluded, “With these key companies on board, the guidelines will have a real impact across America.”73, p 1 In an extensive review of factors in the larger food and physical activity environment, French and others addressed food supply trends, habits of eating away from home, food advertising, promotion and education, and food pricing.74 Twenty-six potential intervention strategies were proposed, under categories of community organization/action, financial and economic incentives, food assistance programs, food packaging and labeling, media and advertising, schools and worksites, and transportation and urban/rural development.
46894_CH08_159_190.qxd
184
2/25/10
3:27 PM
Page 184
CHAPTER 8 DIETARY IMBALANCE
From a still broader perspective, Reddy and Katan concluded that:75, p 167 Adequate evidence is available, from studies conducted within and across populations, to link several nutrients, minerals, food groups and dietary patterns with an increased or decreased risk of CVD. . . . Sufficient knowledge exists to recommend nutritional interventions, at both population and individual levels, to reduce cardiovascular disease. That knowledge should now be translated into policies which promote healthy diets and discourage unhealthy diets. This requires coordinated action at the level of governments, international organizations, civil society and responsible sections of the food industry. These latter views extend well into the fourth level of Figure 8-6, the macroenvironment. In their fullest development, they relate to international or global policies that have been articulated for some decades and whose implementation is now urged by WHO and others. Global Strategies At both population and individual levels, it is clear that dietary patterns change and can be modified within a given country. The World Health Organization Study Group Report Diet, Nutrition, and the Prevention of Chronic Diseases of 1990 reviewed experience in Finland, the Netherlands, Norway, the United Kingdom, the United States, Australia, and New Zealand.2 These experiences reveal differences in governmental policies and actions reflecting circumstances of each country. In the cases of Finland, Norway, the United Kingdom, and the United States, some components of diet had changed markedly, such as adoption of low-fat dairy products in place of traditional ones. The extent of change in food availability in the United States in only the 20 years from 1965 to 1985 was striking in several respects, reviewed in The Surgeon General’s Report on Nutrition and Health in 1988.76 Substantial decreases in production occurred for eggs, fluid whole milk, butter, lard, and refined sugars while marked increases occurred for poultry, low-fat dairy products, cheese, vegetable shortening and oils, fresh fruits and vegetables, and corn sweeteners. To some extent, it is plausible that these changes reflect favorable impacts of policy recommendations and public education, aided by such governmental intervention as adoption of more stringent food labeling regulations and other measures.
The 2003 sequel to the WHO report of 1990 cited above was the Report of a Joint FAO/WHO Expert Consultation on nutrition and chronic diseases. It addressed dietary recommendations from a global and historical perspective, tracing developments in nutrition and food policies from the 1940s.2 The report acknowledged coexistence of both nutritional deficiencies and the “affluent” diet in many countries, both developed and developing, but its emphasis was on the latter dietary problem. Published dietary recommendations for both industrialized and developing countries were summarized. Developing countries were represented only by India and Latin America and these by only two recent reports. For India, it was remarkable that for the high-risk, affluent segment of the population it was still considered feasible to target total fat intake at 20% of calories or less, and mention of cholesterol intake was apparently considered unnecessary. For Latin America, the estimated intake of total fats was also low, and the saturated fat intake was especially so. Generally similar components of the diet were addressed in the developing as in the industrialized countries, but goals reflected intervention much earlier in the course of population trends toward increased average fat intake. Prevention of dietary imbalance in the first place remained a conceivable goal. The World Health Assembly (WHA) in 2004 endorsed the WHO Global Strategy on Diet, Physical Activity and Health, the outcome of a WHA resolution in 2002 and extensive intervening consultations.77 The Strategy summarized evidence in support of its recommendations that, regarding diet, were to achieve essentially the same dietary goals advocated by others as discussed above. Principles for action were presented to guide development of national and regional strategies and action plans. Strategies should be science based, comprehensive with respect to noncommunicable diseases taken together, multisectoral and multidisciplinary, consistent with principles of the Ottawa Charter for Health Promotion, and cognizant of “the complex interactions between personal choices, social norms and economic and environmental factors.”77, p 5 The principles also called for a life-course perspective; comprehensive public health efforts addressing all aspects of nutrition and physical activity; priority for activities reaching the poorest populations and communities; supporting evaluation, monitoring, and surveillance; and sensitivity to differences in patterns of diet and physical activity related to gender, culture, age, and local and regional traditions. Actions were outlined for Member States, WHO, international partners, civil society and nongovernmental organizations,
46894_CH08_159_190.qxd
2/25/10
3:27 PM
Page 185
CURRENT ISSUES 185
and the private sector. Prospects for improving health were viewed positively:77, p 14 Changes in diet and physical activity will be gradual, and national strategies will need a clear plan for long-term and sustained diseaseprevention measures. However, changes in risk factors and in incidence of Noncommunicable diseases can occur quite quickly when effective interventions are made. National plans should therefore also have achievable short-term and intermediate goals. From the perspective of developing countries, and prevention of chronic diseases broadly, Willett and others presented six familiar dietary recommendations, adding adequate folic acid intake.78 Interventions proposed were: education, in schools, in worksites, and by healthcare providers; improvement in the food supply, through changes in processing and manufacturing, fortification, making healthy foods more available and less costly, and promoting healthy foods while limiting aggressive marketing of food to children; and economic policies. Cost-effectiveness analysis of policy options indicated estimated costs in US$/DALY averted, based on varying combinations of intervention costs and reduction in risk of coronary artery disease achieved. These were presented for media campaigns to reduce saturated fat content, substituting 2% of energy from trans fat with polyunsaturated fat, and reducing salt content by means of legislation plus public education. Cost effectiveness estimates differed among WHO regions. It was noted that interventions achievable through policy change that do not require public education are far less costly than others. Strategies proposed to bring about these and other proposed changes in developing countries included changes in food supply by collaboration with the agriculture sector and food industry; school food and nutrition programs; labeling of energy content for all packaged and fast food products; and use of tax policies to make healthy foods the more affordable choices. Again, optimistic conclusions followed from this analysis:78, p 848 Many of the ongoing diet and lifestyle interventions in low- and middle-income countries are relatively recent, and few have documented reductions in the rates of major chronic diseases. However, the successes of Finland, Singapore, and many other high-income countries in reducing rates of CAD, stroke, and smoking-related cancers strongly suggest that similar benefits will emerge in the developing countries.
The analysis by Hawkes of globalization and its relation to processes of nutrition transition, noted previously, concluded with several observations regarding policy implications of these processes at the global level:31, p 22 First, policies must take into account the influence of the policies and processes of global market integration on long-term dietary change, and the context in which they operate. Such a process requires looking beyond the health sector as narrowly defined, and entering into debates and policy arenas dealt with by other sectors and disciplines. Second, policies must address, in some way, the behaviour of TFCs [transnational food companies], preferably by creating incentives to improve “healthy” market functioning. Third, policies need to focus on the promotion of healthy diets over the long-term among groups of low SES. The concern of this paper has been groups with access to diets sufficient in energy. But diet quality is important for those at risk from undernutrition; policies that focus on diet quality are therefore important for addressing problems across the whole nutritional spectrum. Policies that are commonly proposed concern labeling of foods and regulating food advertising and promotion. But, Hawkes argues, these interventions close to the consumer may have limited impact:31, p 23 To alter the series of incentives in the global marketplace from farm to fork, there is need for policy to effect change closer to the point of production . . . FDI [foreign direct investment, the process by which a TFC purchases food production and marketing capacity in another country] represents a single, upstream, entry point to many of the dynamics influencing the production, sale, and promotion of foods in the global market place, and thus could be an effective lever for change. The benefits of such approaches are that they influence markets, not just the products sold in markets. Relatively small changes at a macro-scale can have relatively large population-wide impacts. Perhaps most important, they are the approaches that are most likely to benefit groups of low socioeconomic status over the long-term.
CURRENT ISSUES To make the optimal diet the usual diet—combining the essential features of the dietary patterns reviewed
46894_CH08_159_190.qxd
186
2/25/10
3:27 PM
Page 186
CHAPTER 8 DIETARY IMBALANCE
here—would improve health and longevity for people everywhere and greatly reduce the mortality and burden attributable to cardiovascular diseases. This will require change at every level of analysis, from the broadest macroeconomic forces that determine what investments in agriculture and food manufacture are most profitable to the narrowest choices made by individuals at moments of food consumption, given that choices are in fact available. Reducing salt consumption is justifiably a distinct priority and strategy because it can be isolated from the broader dietary policies in a meaningful way; however, this is not a solitary problem nor is it the whole nutritional answer to generation after generation of great numbers of people with high blood pressure worldwide. But high blood pressure will persist at high levels of prevalence until salt consumption is reduced, so it cannot be left out of a comprehensive nutritional approach to prevention of cardiovascular diseases. Research is needed most to establish what policy interventions can be implemented to greatest effect and what in fact are the costs and benefits of doing so. If costs are seen as transfers of expenditure from less to more effective investments in health, and if benefits are fully taken into account, the cost–benefit ratios for policy options in nutrition may be found to favor substantially greater expenditure for prevention in all regions of the world. REFERENCES 1. Stamler J. Population studies. In: Levy RI, Rifkind B, Dennis B, Ernst N, eds. Nutrition, Lipids, and Coronary Heart Disease. New York: Raven Press; 1979:25–88. 2. World Health Organization Study Group. Diet, Nutrition, and the Prevention of Chronic Diseases. Technical Report Series 797. Geneva (Switzerland): World Health Organization; 1990. 3. Committee on Diet and Health, Food and Nutrition Board, Commission on Life Sciences, National Research Council. Diet and Health: Implications for Reducing Chronic Disease Risk. Washington, DC: National Academy Press; 1989. 4. US Department of Health and Human Services and US Department of Agriculture. Dietary Guidelines for Americans, 2005. 6th ed. Washington, DC: US Government Printing Office; January 2005.
5. Nutrition and Your Health: Dietary Guidelines for Americans. 4th ed. Washington, DC: US Department of Health and Human Services; 1995. 6. MyPyramid. Steps to a Healthier You. http:// www.MyPyramid.gov. Accessed October 1, 2008. 7. Sacks FM, Katan M. Randomized clinical trials on the effects of dietary fat and carbohydrate on plasma lipoproteins and cardiovascular disease. Am J Med. 2002;113(9B):13S–24S. 8. Ding EL, Mozaffarian D. Optimal dietary habits for the prevention of stroke. Sem Neurol. 2006;26:11–23. 9. Hu FB. Plant-based foods and prevention of cardiovascular disease: an overview. Am J Clin Nutr. 2003;78(suppl):544S–551S. 10. Stamler J. Established major coronary risk factors. In: Marmot M, Elliott P, eds. Coronary Heart Disease Epidemiology: From Aetiology to Public Health. Oxford (England): Oxford University Publishers; 1992:35–66. 11. Keys A. Mediterranean diet and public health: personal reflections. Am J Clin Nutr. 1995; 61(suppl):1321S–1323S. 12. Freedman MR, King J, Kennedy E. Popular diets: a scientific review. Obes Res. 2001; 9(suppl):1S–40S. 13. National Cholesterol Education Program Expert Panel on Detection, Evaluation and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). NIH Publication No. 02-5215. Bethesda, MD: National Cholesterol Education Program, National Heart, Lung and Blood Institute, National Institutes of Health; September 2002. 14. Thompson FE, Byers T. Dietary assessment resource manual. Am J Nutr. 1994;124(suppl): 2245S–2317S.
46894_CH08_159_190.qxd
2/25/10
3:27 PM
Page 187
REFERENCES 187
15. Keys A. Coronary Heart Disease in Seven Countries. Monograph 29. Dallas, TX: American Heart Association; 1970. 16. Anderson SA. Guidelines for use of dietary intake data. J Amer Diet Assoc. 1988;88:1258. 17. Beaton GH, Burema J, Ritenbaugh C. Errors in the interpretation of dietary assessments. Am J Clin Nutr. 1997;65(suppl):1100S–1107S. 18. Coates RJ, Monteilh CP. Assessments of foodfrequency questionnaires in minority populations. Am J Clin Nutr. 1997;65(suppl): 1108S–1115S. 19. Rockett HRH, Colditz GA. Assessing diets of children and adolescents. Am J Clin Nutr. 1997;65(suppl):1116S–1122S. 20. Cutler JA, Stamler J. Introduction and summary of the dietary and nutritional methods and findings in the Multiple Risk Factor Intervention Trial. Am J Clin Nutr. 1997; 65(suppl):184S–190S. 21. Dennis B, Stamler J, Buzzard M, et al., for the INTERMAP Research Group. J Human Hypert. 2003;17:609–622. 22. Hu FB. Dietary pattern analysis: a new direction in nutritional epidemiology. Curr Opinion Lipidol. 2002;13:3–9. 23. Schulze MB, Hoffmann K. Methodological approaches to study dietary patterns in relation to risk of coronary heart disease and stroke. Brit J Nutr. 2006;95:860–869. 24. McKinnon RA, Reedy J, Handy SL, Brown Rodgers A. Measuring the food and physical activity environments. Shaping the research agenda. Am J Prev Med. 2009;36(4S):S81–S85.
National Cholesterol Education Program, Public Health Service, National Institutes of Health; 1990. 27. Gidding SS, Lichtenstein AH, Faith MS, et al. Implementing American Heart Association pediatric and adult nutrition guidelines. AQ Scientific Statement from the American Heart Association Nutrition Committee of the Council on Nutrition, Physical Activity and Metabolism, Council on Cardiovascular Diseases in the Young, Council on Arteriosclerosis, Thrombosis, and Vascular Biology, Council on Cardiovascular Nursing, Council on Epidemiology and Prevention, and Council for High Blood Pressure Research. Circulation. 2009;119:1161–1175. 28. Centers for Disease Control and Prevention. Competitive foods and beverages available for purchase in secondary schools––selected sites, United States, 2006. MMWR. 2008;57: 935–938. 29. Franco M, Diez Roux AV, Glass TA, Caballero B, Brancati FL. Neighborhood characteristics and availability of healthy foods in Baltimore. Am J Prev Med. 2008;35(6):561–567. 30. Pollan M. Farmer in chief. The New York Times Magazine. October 12, 2008: 62–92. 31. Hawkes C. Uneven dietary development: linking the policies and processes of globalization with the nutrition transition, obesity and dietrelated chronic diseases. Globalization Health. 2006;2:4. 32. Breifel RR, Johnson CL. Secular trends in dietary intake in the United States. Annu Rev Nutr. 2004;24:401–431.
25. Story M, Giles-Corti B, Lazarus Yaroch A, et al. Work Group IV: Future directions for measures of the food and physical activity environments. Am J Prev Med. 2009;36(4S): S182–S188.
33. US Department of Health and Human Services. Health, United States, 2008 with Special Feature on the Health of Young Adults. Washington, DC: US Department of Health and Human Services. Centers for Disease Control and Prevention. National Center for Health Statistics; 2008.
26. Report of the Expert Panel on Population Strategies for Blood Cholesterol Reduction. NIH publication 90-3046. Bethesda, MD:
34. Centers for Disease Control and Prevention. Prevalence of fruit and vegetable consumption and physical activity by race/ethnicity––
46894_CH08_159_190.qxd
188
2/25/10
3:27 PM
Page 188
CHAPTER 8 DIETARY IMBALANCE
United States, 2005. MMWR. 2007;56: 301–304. 35. Centers for Disease Control and Prevention. Youth risk behavior surveillance––United States, 2007. Surveillance Summaries, June 6, 2008. MMWR. 2008;57(No. SS-04). 36. Malmros H. The relation of nutrition to health. Acta Medica Scand. 1950;246(suppl): 137–153. 37. Keys A. Seven Countries: A Multivariate Analysis of Death and Coronary Heart Disease. Cambridge, MA: Harvard University Press; 1980. 38. Artaud-Wild SM, Connor SL, Sexton G, Connor WE. Differences in coronary mortality can be explained by differences in cholesterol and saturated fat intakes in 40 countries but not in France and Finland. Circulation. 1993; 88:2771–2779. 39. Stamler J. Assessing diets to improve world health: nutritional research on disease causation in populations. Am J Clin Nutr. 1994; 59(suppl):146S–156S. 40. Shekelle RB, Stamler J. Dietary cholesterol and ischaemic heart disease. Lancet. 1989;i: 1177–1179. 41. Stamler J, Shekelle RB. Dietary cholesterol and human coronary heart disease. Arch Path Lab Med. 1988;112:1032–1040. 42. Panel on Macronutrients, Subcommittees on Upper Reference Levels of Nutrients and on Interpretation and Uses of Dietary Reference Intakes, and the Standing Committee on the Scientific Evaluation of Dietary Reference Intakes. Letter Report on Dietary Reference Intakes for Trans Fatty Acids. Drawn from the Report on Dietary Reference Intakes for Energy, Carbohydrate, Fiber, Fat, Fatty Acids, Cholesterol, Protein, and Amino Acids. Washington, DC: Institute of Medicine, Food and Nutrition Board; 2002. 43. Ascherio A, Hennekens CH, Buring JE, Master C, Stampfer MJ, Willett WC. Trans-fatty acids intake and risk of myocardial infarction. Circulation. 1994;89:94–101.
44. Willett WC, Stampfer MJ, Manson JE, et al. Intake of trans fatty acids and risk of coronary heart disease among women. Lancet. 1993; 341:581–585. 45. Kromhout D, Menotti A, Bloemberg B, et al. Dietary saturated and trans fatty acids and cholesterol and 25-year mortality from coronary heart disease: the Seven Countries Study. Prev Med. 1995; 24:308–315. 46. Aro A, Kardinaal AFM, Saliminen I, et al. Adipose tissue, isomeric trans fatty acids and risk of myocardial infarction in nine countries: the EURAMIC study. Lancet. 1995;345: 273–278. 47. Mozaffarian D, Katan MB, Ascherio A, Stampfer MJ, Willett WC. Trans fatty acids and cardiovascular disease. N Engl J Med. 2006;354:1601–1613. 48. Shekelle RB, Stamler J. Fish and coronary heart disease: the epidemiologic evidence. Nutr Metab Cardiovasc Dis. 1993;3:46–51. 49. Stone NJ. Fish consumption, fish oil, lipids and coronary heart disease. Circulation. 1996;94: 2337–2340. 50. Hooper L, Thompson RI, Harrison RA, et al. Risks and benefits of omega 3 fats for mortality, cardiovascular disease, and cancer: systematic review. BMJ. 2006;332(7544):752–760. 51. Mozaffarian D, Rimm EB. Fish intake, contaminants, and human health. Evaluating the risks and the benefits. JAMA. 2006;296:1885–1899. 52. Rimm EB, Ascherio A, Giovannucci E, et al. Vegetable, fruit and cereal fiber intake and risk of coronary heart disease among men. JAMA. 1996;275:447–451. 53. Pietinen P, Rimm EB, Korhonen P, et al. Intake of dietary fiber and risk of coronary heart disease in a cohort of Finnish men: the AlphaTocopherol, Beta-Carotene Cancer Prevention Study. Circulation. 1996;94:2720–2727. 54. Greenland S. A meta-analysis of coffee, myocardial infarction, and coronary death. Epidemiol. 1993;4:366–374.
46894_CH08_159_190.qxd
2/25/10
3:27 PM
Page 189
REFERENCES 189
55. Thelle DS. Coffee, tea and coronary heart disease. Curr Opin Lipidology. 1995;6:25–27. 56. Mukamal KJ, Maclure M, Muller JE, Sherwood JB, Mittleman MA. Caffeinated coffee consumption and mortality after acute myocardial infarction. Am Heart J. 2004;147:999–1004. 57. Mukamal KJ, Maclure M, Muller JE, Sherwood JB, Mittleman MA. Tea consumption and mortality after acute myocardial infarction. Circulation. 2002;105:2476–2481. 58. Danaei G, Ding EL, Mozaffarian D, et al. The preventable causes of death in the United States: comparative risk assessment of dietary, lifestyle, and metabolic risk factors. PLoS Med. 2009;6(4): e1000058. doi:10.1371/journal. pmed.1000058. 59. Trichopoulou A, Costacou T, Bamia C, Trichopoulos D. Adherence to a Mediterranean diet and survival in a Greek population. N Engl J Med. 2003;348:2599–2608. 60. Nordmann AJ, Nordmann A, Briel M, et al. Effects of low-carbohydrate vs low-fat diets on weight loss and cardiovascular risk factors. Arch Intern Med. 2006;166:285–293. 61. Dansinger ML, Gleason JA, Griffith JL, Selker HP, Schaefer EJ. Comparison of the Atkins, Ornish, Weight Watchers, and Zone diets for weight loss and heart disease risk reduction. A randomized trial. JAMA. 2005;293: 43–53. 62. Obarzanek E, Sacks FM, Vollmer WM, et al. Effects on blood lipids of a blood pressurelowering diet: the Dietary Approaches to Stop Hypertension (DASH) Trial. Am J Clin Nutr. 2001;74:80–89. 63. Iqbal R, Anand S, Ounpuu S, et al. Dietary patterns and the risk of acute myocardial infarction in 52 countries. Circulation. 2008;118: 1929–1937. 64. Central Committee for Medical and Community Program of the American Heart Association. Dietary fat and its relation to heart attacks and strokes. Circulation. 1961; 23:133–136.
65. Lichtenstein AH, Appel LJ, Brands M, et al. Diet and lifestyle recommendations revision 2006. A Scientific Statement from the American Heart Association Nutrition Committee. Circulation. 2006;114:82–96. 66. Hu FB, Willett WC. Optimal diets for prevention of coronary heart disease. JAMA. 2002; 288:2569–2578. 67. Williams CL, Hayman LL, Daniels SR, et al. Cardiovascular health in childhood: a statement for health professionals from the Committee on Atherosclerosis, Hypertension, and Obesity in the Young (AHOY) of the Council on Cardiovascular Disease in the Young, American Heart Association. Circulation. 2002;106:143–160. 68. US Preventive Services Task Force. The Guide to Clinical Preventive Services 2006. Recommendations of the U.S. Preventive Services Task Force. Washington, DC: Agency for Healthcare Research and Quality. 2006. www.ahrq.gov/clinic/uspstf/uspstbac.htm. Accessed October 14, 2007. 69. Brunner EJ, Thorogood M, Rees K, Hewitt G. Dietary advice for reducing cardiovascular risk. Cochrane Database of Systematic Reviews 2005, Issue 4. Art. No.: CD002128. doi:10. 1002/14651858.CD002128.pub2. http://www .cochrane.org/reviews/en/ab002128.html. Accessed October 1, 2007. 70. Brunner E, White I, Thorogood M, Bristow A, Curle D, Marmot M. Can dietary interventions change diet and cardiovascular risk factors? A meta-analysis of randomized controlled trials. Am J Public Health. 1997;87: 1415–1422. 71. Luepker RV, Perry CL, McKinlay SM, et al. Outcomes of a field trial to improve children’s dietary patterns and physical activity: the Child and Adolescent Trial for Cardiovascular Health (CATCH). JAMA. 1996;275:768–776. 72. Guide to Community Preventive Services. Promoting good nutrition: school-based programs promoting nutrition and physical activity. http://www.thecommunityguide.org/nutrition/ schoolprograms.html. Last updated March 10, 2009. Accessed June 13, 2009.
46894_CH08_159_190.qxd
190
2/25/10
3:27 PM
Page 190
CHAPTER 8 DIETARY IMBALANCE
73. Alliance for a Healthier Generation. President Clinton and American Heart Association announce joint agreement between Alliance for a Healthier Generation and food industry leaders to set healthy standards for snacking in school. http://www.healthiergeneration.org. Accessed October 6, 2006. 74. French SA, Story M, Jeffery RW. Environmental influences on eating and physical activity. Annu Rev Public Health. 2001;22:309–335. 75. Reddy KS, Katan MB. Diet, nutrition and the prevention of hypertension and cardiovascular diseases. Public Health Nutr. 2004;7(1A): 167–186. 76. The Surgeon General’s Report on Nutrition and Health. NIH publication 88-50210. Bethesda, MD: National Cholesterol Education Program, Public Health Service, National Institutes of Health; 1988.
77. World Health Organization. Global Strategy on Diet, Physical Activity and Health. Geneva (Switzerland): World Health Organization; 2004. 78. Willett WC, Koplan JP, Nugent R, Dusenbury C, Puska P, Gaziano T. Prevention of chronic disease by means of diet and lifestyle changes. In: Jamison DT, Breman JG, Measham AR, et al., eds. Disease Control Priorities in Developing Countries. 2nd ed. Washington, DC: International Bank for Reconstruction and Development/The World Bank; 2006: 833–850.
46894_CH09_191_222.qxd
2/25/10
3:31 PM
Page 191
C H A P T E R
9 Physical Inactivity nary heart disease, although greater intensity or duration is recognized as having still greater benefit and may be necessary for weight control. Policies have been proposed, and some implemented, for improving physical activity in various settings—schools, health care, worksites, and communities—and through action of governments and many sectors of society. Such documents as Physical Activity and Health. A Report of the Surgeon General (US, 1996) and the Global Strategy on Diet, Physical Activity and Health (WHO, 2004) clearly indicate that physical inactivity, along with diet, has become a prominent national and global public health concern. There is reason for optimism that meaningful action can be taken at individual, community or populationwide, and global levels as the health and social benefits of doing so become more fully appreciated.
SUMMARY Physical inactivity is a widely prevalent condition in all modern societies and represents, like contemporary dietary patterns, a radical change from the thousands of years of human development up until the most recent two centuries. As an object of research, physical activity has many features in common with diet, and the methods of assessing usual habits of activity pose parallel challenges. Still, many studies of the health effects of physical activity, the behavioral characteristic, and physical fitness, its physiologic counterpart, have been conducted with a wide variety of methods. The early epidemiologic studies by Morris, by Paffenbarger, and by others focused on occupational activity. Taylor and others shifted focus to “leisuretime physical activity,” outside of work. It seemed increasingly unlikely that work activity could contribute to healthy activity levels, given mechanization, automation, and the disappearance of personal locomotion in relation to work. A large body of evidence now exists on determinants of physical inactivity, mechanisms by which it contributes to causation of cardiovascular and other chronic conditions, and its distribution in the United States and some other populations among both adults and children. Partially because of its pervasive occurrence, it contributes importantly to ischemic heart disease risk globally. Numerous conferences and organizations have addressed the need to assure some minimum level of physical activity. The prevailing recommendations for individuals currently are to complete 30 minutes or more of moderate to vigorous physical activity daily, on most or all days of the week. There is general agreement that this level is sufficient to offer significant protection against coro-
INTRODUCTION “Physical inactivity” denotes the relative dearth of energy expenditure for personal labor, locomotion, or recreation that is widely prevalent in contemporary societies. Coupled with modern dietary habits, it represents an imbalance between energy expenditure and energy intake. Like modern dietary habits, it is seen most clearly in relief against the human evolutionary background. As outlined by Blackburn, physical activity ranked for eons alongside food consumption as one of the two “primal human activities.”1 Each could be viewed as being necessary for the conduct of the other. Anthropological evidence on bone structure indicates smaller stature and heavier musculature in early Homo sapiens, who are also inferred to have had great capacity for meeting demands of strength
191
46894_CH09_191_222.qxd
192
2/25/10
3:31 PM
Page 192
CHAPTER 9 PHYSICAL INACTIVITY
and endurance as well as surviving periods of food scarcity. Through several tens of thousands of years, the conditions of life are thought to have changed little, until the first agricultural revolution, which introduced plant and animal domestication. The contrast is offered between this long-standing human experience and the “relatively great height and mass and relatively less muscularity and bone strength of modern affluent humans who have been exposed to only a few generations of sedentariness in the presence of perpetual excess in calories and nutrients.”1, p 102 Blackburn also noted: “The portrayal suggests that the recent magnitude of mass sedentation and caloric intake of modern man may act as a substantial adaptive stress that may result in mass metabolic maladaptations . . . [that] will surely persist in affluent societies unless there is a general cultural change.”1, p 102 Physical Activity and Health, the topic of the US Surgeon General’s report on the occasion of the centennial Olympiad in 1996, recounted the history of attention to physical activity as a societal or cultural interest.2 Physical activity in various forms appears to have been prescribed, as well as being an integral part of cultural norms, in both the East and West for several thousand years. Health benefits were attributed to exercise in ancient writings in China, India, and Greece. Through the Western Middle Ages and into the modern period, early Greek influences persisted, and by the 18th and 19th centuries several authoritative texts on physical activity had been produced. The view that lack of physical activity had its harmful effects was also explicit in the writing of prominent figures of the time. Even the 19th-century introduction of the concept of physical education still claimed Greek antecedents. Scientific study of physical fitness in relation to anthropometry and cardiorespiratory physiology developed in the latter 19th and early 20th centuries, respectively. The practical importance of this science became apparent in the United States at the time of both the First and Second World Wars. It was recognized then that the physical condition of the population was suboptimal. National programs to improve it were instituted in the mid-1940s. Further reports of poor physical performance of US youth in relation to their European counterparts added impetus and visibility to the issue. Epidemiologic study of physical activity in relation to cardiovascular diseases—especially coronary heart disease—began in the 1950s, and Morris,3 Taylor,4 Paffenbarger,5 and others were among the prominent contributors. Morris has provided an account of the early work describing development of the
first hypotheses for epidemiologic investigation of a putative relation between physical activity and protection from coronary heart disease.3 On the basis of this pioneering work and intervening research, the American College of Sports Medicine developed guidelines for evaluating individual fitness and prescribing remedial programs. By the 1990s, this material was in its second edition.6 National attention to physical activity was given new prominence in the United States with issuance of Physical Activity and Health in 1996.2 The Surgeon General’s report reviewed more than four decades of epidemiologic research on physical activity as well as laboratory and clinical investigations of the physiology of exercise. The report described population patterns and trends of physical activity in the United States and addressed the evolution of policies, recommendations, and strategies for intervention in youth and adulthood. Haskell has recently contributed an addendum to Morris’s original account that updates the story to the current decade.7 Subsequent developments have added greatly to appreciating the public health importance of overcoming physical inactivity—especially prominent is emergence of overweight and obesity as a national, and global, epidemic. Focus on children and youth is reflected in numerous activities, illustrated by a supplemental issue in 2000 of Preventive Medicine on children’s physical activity and nutrition and formation of a new partnership of the American Cancer Society, American Diabetes Association, and American Heart Association (AHA) that supports physical education in schools.8 The American College of Sports Medicine and AHA in 2007 issued new recommendations for physical activity among adults.9 The US Department of Health and Human Services has released 2008 Physical Activity Guidelines for Americans, “the first comprehensive guidelines on physical activity ever to be issued by the Federal government.”10, p i Additionally, the World Health Assembly has endorsed the Global Strategy on Diet, Physical Activity and Health, developed and now being implemented by the World Health Organization (WHO).11
CONCEPTS AND DEFINITIONS Physical inactivity and its opposite, physical activity, are associated with a variety of terms and concepts that have distinct meanings within the field. Terms with proposed standard definitions include physical activity (“bodily movement produced by skeletal muscles that requires energy expenditure”); exercise (“a
46894_CH09_191_222.qxd
2/25/10
3:31 PM
Page 193
CONCEPTS AND DEFINITIONS 193
type of physical activity defined as a planned, structured, and repetitive bodily movement done to improve or maintain one or more components of physical fitness”); physical fitness (“a set of attributes that people have or achieve that relates to the ability to perform physical activity”); and physical inactivity (“a level of activity less than that needed to maintain good health”).2 Physical inactivity is thus unhealthy by definition, and it is this part of the activity spectrum that is most directly of concern, although many other aspects are addressed, in cardiovascular epidemiology. Other dimensions of physical activity important for epidemiologic studies of physical inactivity and cardiovascular diseases include the distinction between occupational (work) activity and nonoccupational or leisure-time physical activity. Early studies, through the 1970s, focused on occupational activity, whereas more recent ones have emphasized nonoccupational activity, often termed “leisure time physical activity,” or LTPA. Gradation of activity is commonly expressed in semiquantitative terms such as light, moderate, or vigorous. Alternatively, it may be measured in units of energy expenditure (kilocalories per minute [kcal/min] or kilojoules per minute [kjoule/min], in which 1 kcal 4.184 kjoule) or expressed in effort required, as measured in terms of
the ratio of work metabolic rate to resting metabolic rate, in units called METs. One MET is defined as the rate of oxygen consumption of an adult seated at rest, which is approximately 3.5 milliliters per minute per kilogram body weight (1 MET 3.5 ml/min/kg). As an example, Table 9-1 shows several types of activities classified as light, moderate, or hard/vigorous, with corresponding measures of energy expenditure or oxygen consumption.12 Patterns of habitual physical activity, implicitly nonoccupational, are described in terms of intensity, duration, and frequency of activity. The ability to characterize and measure these components of physical activity patterns in individuals is critical for understanding the relation between physical activity and measures of health. It is also fundamental to policies and recommendations intended to maintain or restore an optimum pattern of activity.13,14 Answers to several kinds of questions depend on this understanding: How is cardiovascular health influenced by different patterns of activity? What pattern should be recommended for prevention of cardiovascular diseases? Will the pattern thought best for cardiovascular disease prevention be best for other major conditions, such as obesity? The distinction between physical fitness and physical activity warrants further clarification because of
Table 9-1
Examples of Common Physical Activities for Healthy U.S. Adults, by Intensity of Effort Required Light Moderate Hard/Vigorous 3.0 METs or 4 kcal/min) 6.0 METs or 7 kcal/min) ( (3.0–6.0 METs or 4–7 kcal/min) ( Walking, briskly uphill or with a load Walking, briskly (3–4 mph) Walking, slowly (strolling) (1–2 mph) Cycling, stationary ( 50 W)
Cycling for pleasure or transportation ( 10 mph)
Cycling, fast or racing ( 10 mph)
Swimming, slow treading
Swimming, moderate effort
Swimming, fast treading or crawl
Conditioning exercise, light stretching
Conditioning exercise, general calisthenics Racket sports, table tennis
Conditioning exercise, stair ergometer, ski machine Racket sports, singles tennis, racquetball
Golf, power cart
Golf, pulling cart or carrying clubs
Bowling Fishing, sitting Boating, power
Fishing, standing/casting Canoeing, leisurely (2.0–3.9 mph)
Fishing in stream Canoeing, rapidly ( 4 mph)
Home care, carpet sweeping Mowing lawn, riding mower Home repair, carpentry
Home care, general cleaning Mowing lawn, power mower Home repair, painting
Moving furniture Mowing lawn, hand mower
Note: The MET (work metabolic rate/resting metabolic rate) is a multiple of the resting rate of oxygen consumption during physical activity. One MET represents the approximate rate of oxygen consumption of a seated adult at rest, or about 3.5 ml/min/kg. The equivalent energy cost of 1 MET in kcal/min is about 1.2 for a 70-kg person or approximately 1 kcal/kg/hr. W = watts. Source: Reprinted from Journal of the American Medical Association, Vol 273, p 404, 1995.
46894_CH09_191_222.qxd
194
2/25/10
3:31 PM
Page 194
CHAPTER 9 PHYSICAL INACTIVITY
the close relation of both to cardiovascular status. Physical fitness connotes numerous attributes, most broadly the “ability to carry out our daily tasks with vigor and alertness, without undue fatigue, and with ample energy to enjoy leisure-time pursuits and to meet unforeseen emergencies.”2, p 20 To further distinguish performance-related (i.e., athletic) fitness from health-related fitness, the latter has been described as including “cardiorespiratory fitness, muscular strength and endurance, body composition, and flexibility.”2, p 21 A recent development is the concept of “active living,” under an initiative of the Robert Wood Johnson Foundation and its Active Living Research Program.15 “Active living” has been defined as “a way of life that integrates physical activity into everyday routines encompassing both leisure-time physical activity and walking and biking for transportation purposes.”15, p S3 The relation of active living research to public health, policy making, healthy communities, and other aspects of this emerging field are addressed in a special supplement to the American Journal of Preventive Medicine.16
MEASUREMENT Procedures for assessing physical activity (or inactivity) are evaluated for their potential research use in Table 9-2.2 Eliciting information about individual activity is similar to doing so for dietary intakes. Four survey approaches are the task-specific diary; recall questionnaire; quantitative history; and global selfreport. Important features of these methods are indicated in the table and have obvious implications for their practicality in various circumstances. The four approaches are all self-report methods with ranges of time reference from days to 1 year. The first three are limited by the need for respondents’ understanding and ability to record or recall details of activity within a specific time frame. They are not feasible methods for characterizing activity of young children. The last survey method is a simpler self-rating of one’s activity as perceived in relation to that of others. Direct monitoring of activity can be accomplished by several means, also characterized in the table. Each technique has strengths and limitations, and applicability of a given method is generally dictated by the circumstances of investigation. Assessing Physical Fitness and Physical Activity in Population-Based Surveys, published in 1989 by the National Center for Health Statistics, presented details of methods used in the United States.17 It includes examples of survey instruments as well as
methodologic studies of relations among the available measures. Strategies for study of activity and fitness and how these relate to health and disease are also discussed. This comprehensive resource provides insight into the critical evaluation of studies in this area and valuable background information for planning new research. Similarly, a supplement to Medicine and Science in Sports and Exercise, “A Collection of Physical Activity Questionnaires for Health-Related Research,”18 presents actual questionnaires used in the population at large and among the elderly, with examples from several major epidemiologic studies. A brief abstract introduces each questionnaire to identify the activity components assessed, time frame of recall, original mode of administration, and both contact information and a primary literature reference. This is a valuable resource for more detailed information about methods of physical activity assessment. Various questionnaire methods for assessing leisure-time physical activity, from the 1960s through the 1980s, were critically reviewed and provided insight into many studies up to that time as this aspect of physical activity came to dominate the field.19 As late as 2002, WHO found a lack of an internationally agreed definition or measure of physical activity and only scant data on activity related to occupation, transportation, or domestic tasks; the most commonly available data related to leisure-time activity.20 Comparability of data around the world has been quite limited.
DETERMINANTS Factors of modern life that result in population-wide physical inactivity, in contrast with our evolutionary history, are commonly described as: disappearance of energy expenditure for obtaining food, substantial shift from heavy to light or nonphysical work in most daily occupations, reliance on motorized vehicles and elevators rather than personal locomotion for transport, and use of nonoccupational or leisure time for passive pursuits, especially the mass habit of viewing television. At the individual level, behavioral theories are proposed as reviewed in the Surgeon General’s report, but research has not established a clear understanding of factors that distinguish those who are active from the great majority who are not.2 Under the rubric of environmental influences on physical activity, French and others considered television watching, automobile use, limited park and recreation space, and occupational inactivity, in addition to prevalent sedentary activities—use of computers and
Yes Yes No
Adult, elderly Adult, elderly Child, adult, elderly All Child, adult, elderly Infant All Adult, elderly Child, adult, elderly No No No Yes
Yes No
No Yes No No
Adult, elderly Adult All All
No No No No
Yes No
No Yes No
No Yes No No
Yes Yes Yes Yes
Yes No No No
Yes Yes
Yes Yes Yes
No Yes No No
Yes Yes No Yes
Yes No No Yes
Yes Yes
Yes Yes Yes
No Yes Yes Yes
No Yes No Yes
Yes No No Yes
Yes Yes
Yes Yes Yes
Yes Yes Yes Yes
No Yes No Yes
No Yes Yes No
No No
No No No
Yes No No No
Yes No No No
Likely to Influence Behavior
Yes No No Yes
Yes Yes
Yes Yes Yes
? Yes Yes Yes
? Yes Yes Yes
Acceptable to Persons
Yes No No Yes
Yes Yes
Yes Yes Yes
? Yes Yes Yes
Yes Yes Yes Yes
Socially Acceptable
No Yes Yes No
No No
No No No
Yes Yes No No
Yes Yes Yes No
Activity Specific
Source: Reprinted from Physical Activity and Health, A Report of the Surgeon General, US Department of Health and Human Services, 1996.
Note: Most tests that are applicable for adults can be used in adolescents as well. Few tests can be applied to the pediatric age groups. Among infants only, direct calorimetry, accelerometers, heart rate monitoring, and stabilometers can be used with accuracy.
Stabilometers Direct calorimetry Indirect calorimetry Doubly labeled water
Accelerometers Horizontal time monitor
Monitoring Behavioral observation Job classification Heart rate monitor Heart rate and motion sensor Electronic motion sensor Pedometer Gait assessment
Yes Yes Yes Yes
Low Subject Effort Cost
3:31 PM
Adult, elderly Adult, elderly Adult, elderly Adult, elderly
Assessment Procedures and Their Potential Use in Epidemiologic Research Low Applicable Use in Low Low Subject Age Large-Scale $ Time Time Groups Studies Cost Cost Cost
2/25/10
Measurement Tool Surveying Task-specific diary Recall questionnaire Quantitative history Global self-report
Table 9-2
46894_CH09_191_222.qxd Page 195
DETERMINANTS 195
46894_CH09_191_222.qxd
196
2/25/10
3:31 PM
Page 196
CHAPTER 9 PHYSICAL INACTIVITY
labor-saving devices such as garden tools (mowers, blowers, and snow-throwers)—to be the major categories of factors responsible for inactivity.21,22 Behavioral determinants and environmental characteristics associated with physical activity or inactivity were reviewed by Sherwood and Jeffery.23 Among behavioral aspects they addressed motivation, selfefficacy and stage of change, exercise history, body weight and health risk profiles, diet, and stress. Environmental aspects were social support, time, access, attributes of exercise behavior, and injury. The authors concluded that epidemic sedentariness and interventions to reverse it were little understood as of the year 2000. A review focusing on women identified psychological, social-environmental, physiologic, demographic, and health status variables as influential in determining women’s activity patterns.24 Regarding children, the American Academy of Pediatrics reported in 2009 on the role of the built environment in shaping children’s levels of physical activity.25 The Committee defined “built environment” as “spaces such as buildings and streets that are deliberately constructed as well as outdoor spaces that are altered in some way by human activity.”25, p 1591 They distinguished two forms of activity, addressing both in their analysis and recommendations:25, p 1591 Many factors influence a child’s level of physical activity, including individual-level psychosocial factors such as self-efficacy; family factors such as parental support; and largerscale factors such as social norms. Although these are all important contributors, this policy statement is limited to focusing on how the physical design of the community can support opportunities for physical activity. Opportunities for recreational physical activity arise with parks and green spaces. “Utilitarian” physical activity, such as walking or bicycling to school and to other activities, is an equally important part of a child’s daily life. Environments that promote more active lifestyles among children and adolescents will be important to enable them to achieve recommended levels of physical activity. The importance of assuring a high and sustained level of physical activity in children is underscored by evidence that early patterns persist, or “track,” through childhood and adolescence, as shown by Kelder and others.26 An important determinant of physical activity is previous physical activity—hence the value of establishing favorable conditions for physical activity from early in life.
MECHANISMS How physical inactivity affects cardiovascular function so as to increase population rates and individual risks of coronary heart disease may be inferred at one level from the expected benefits of a prescribed program of exercise for myocardial function.27 Table 9-3 outlines these benefits and ranks each on the likelihood of its occurrence for most people under the program indicated. Mechanisms that either increase myocardial oxygen supply or decrease myocardial work and oxygen demand are emphasized, suggesting that the principal effects of physical activity on the heart would tend to reduce the risk of myocardial ischemia or decrease its severity, especially under circumstances of increased workload or diminished blood flow. At another level, the mechanisms by which physical activity could influence measures of occurrence of cardiovascular diseases are through their effects on other risk factors. A substantial body of research concerning such effects is reviewed in the Surgeon General’s report and elsewhere.2,28 In brief, physical activity is associated with reduced adiposity, blood pressure, diabetes, dyslipidemia, and inflammation and with positive effects on insulin sensitivity, glycemic control, fibrinolysis, and endothelial function. An example of the studies contributing such evidence is a survey among 412 male law enforcement officers that assessed physical activity, physical fitness, percentage of body fat, cigarettes smoked per day, Type A behavior score, high density lipoprotein(HDL-) and total cholesterol concentration, and systolic and diastolic blood pressure.29 Both physical fitness and physical activity were scaled as five categories of increasing levels, and group-specific mean values of each risk factor were examined by level of fitness or activity (Table 9-4). Details of the statistical tests are presented in the footnotes to the table. Generally, percentage of body fat, cigarettes per day, Type A behavior score, and total cholesterol concentration were less and HDL-cholesterol concentration was greater in the highest in contrast to the lowest levels of both fitness and activity. Blood pressure differences were present across fitness categories but not activity categories. It is noteworthy that all differences that were found statistically significant were related to physical fitness, not physical activity. Because these are cross-sectional observations, however, they do not necessarily indicate effects of fitness or physical activity but could instead reflect characteristics of persons who choose to be more physically active or more fit. In either case, they suggest that in some populations, the study of car-
46894_CH09_191_222.qxd
2/25/10
3:31 PM
Page 197
DISTRIBUTION 197
Table 9-3
Biological Mechanisms by Which Exercise May Contribute to the Primary or Secondary Prevention of Coronary Heart Disease
Maintain or increase myocardial oxygen supply Delay progression of coronary atherosclerosis (possible) Improve lipoprotein profile (increase HDL-C/LDL-C ratio) (probable) Improve carbohydrate metabolism and increase fibrinolysis (probable) Decrease adiposity (usually) Increase coronary collateral vascularization (unlikely) Increase epicardial artery diameter (possible) Increase coronary blood flow (myocardial perfusion) or distribution (possible) Decrease myocardial work and oxygen demand Decrease heart rate at rest and submaximal exercise (usually) Decrease systolic and mean systemic arterial pressure during submaximal exercise (usually) and at rest (possible) Decrease cardiac output during submaximal exercise (probable) Decrease circulating plasma catecholamine levels (decrease sympathetic tone) at rest (probable) and at submaximal exercise (usually) Increase myocardial function Increase stroke volume at rest and in submaximal and maximal exercise (likely) Increase ejection fraction at rest and during exercise (likely) Increase intrinsic myocardial contractility (possible) Increase myocardial function by decreasing afterload (probable) Increase myocardial hypertrophy (although this might not reduce coronary heart disease risk) (probable) Increase electrical stability of myocardium Decrease regional ischemia or at submaximal exercise (possible) Decrease catecholamines in myocardium at rest (possible) and at submaximal exercise (probable) Increase ventricular fibrillation threshold by reducing levels of cyclic AMP (possible) Note: Likelihood that effect will occur for a person participating in endurance-type training program for 16 weeks or longer at 65% to 80% of functional capacity for 25 minutes or longer per session (300 kcal) for three or more sessions per week is expressed as “unlikely,” “possible,” “likely,” “probable,” or “usually.” HDL-C, high density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; AMP, adenosine monophosphate. Source: Reproduced with permission from WL Haskell, Sedentary Lifestyle as a Risk Factor for Coronary Heart Disease, in TA Pearson et al., eds, Primer in Preventive Cardiology, © 1994, American Heart Association.
diovascular rates or risks in relation to physical activity or physical fitness require data on other risk factors for sound interpretation of the results.
DISTRIBUTION United States The frequencies of self-reported leisure-time physical activity (inactive, some activity, or regular activity— see legend for definitions) among persons aged 18 years and older are presented by age, sex, race, education, income, Hispanic origin, geographic region, and metropolitan or nonmetropolitan place of residence, in Table 9-5.30 The data represent three iterations of the National Health Interview Survey (NHIS), in 1998, 2005, and 2006 (Table 9-5). The overall prevalence of self-reported inactivity was 40.5% in 1998 and 2005 and 39.5% in 2006. Inactivity in 2006 ranged from 34.8% to 59.6%, from the youngest to the oldest age groups.
The prevalence of inactivity in 2006 within demographic subgroups was greater for women, Hispanics, and non-Hispanic Blacks, older males and females, and less-educated and lower-income persons than for other groups. Some of these disparities were striking: by race (48.9% among Black or African American only versus 32.8% among American Indian or Alaska Native only), Hispanic origin and race (53.9% among Mexicans versus 35.3% among nonHispanic Whites), education (62.3% among those with no high school diploma or GED versus 29.2% with some college or more), or income (56.0% for those below 100% of the poverty line versus 33.6% among those at 200% or more of the poverty line). Among geographic regions, absolute levels of inactivity were highest in the South (44.8%) and outside metropolitan statistical areas (46.4%). Inactivity was generally somewhat less prevalent in 2006 than in 1998, though the improvements were in most cases quite small. “Regular leisure-time physical activity” was defined by NHIS as “three or more sessions per week
8.3* (n 11)
9.4* (n 10) 37.7 (n 9) 210.5 (n 11) 128.0 (n 11) 86.0 (n 11)
Type A Scoreb
HDL cholesterolb (mg/dl)
Total cholesterola (mg/dl)
Systolic BPa (mm Hg)
Diastolic BPa (mm Hg) 83.9 (n 129)
123.7 (n 129)
212.6 (n 129)
40.1 (n 100)
1.8 (n 107)
3.0 (n 129)
83.6 (n 94)
122.7 (n 94)
206.5 (n 94)
40.9 (n 71)
2.4 (n 74)
3.1 (n 94)
81.2 (n 136)
119.9 (n 136)
200.0 (n 136)
45.5* (n 110)
1.5 (n 111)
0.0 (n 136)
81.9 (n 45)
121.0 (n 45)
207.9 (n 45)
41.5 (n 34)
–6.2 (n 36)
5.1 (n 45)
84.2 (n 130)
122.5 (n 130)
210.0 (n 130)
43.2 (n 106)
1.6 (n 109)
3.2 (n 130)
82.9 (n 93)
122.3 (n 93)
211.9 (n 93)
43.4 (n 67)
1.1 (n 71)
1.2 (n 93)
82.7 (n 83)
121.8 (n 83)
195.6 (n 83)
43.5 (n 64)
1.5 (n 66)
0.8 (n 83)
81.0 (n 36)
123.4 (n 36)
202.6 (n 36)
48.7 (n 31)
1.0 (n 31)
1.3 (n 36)
Source: Reprinted with permission from Research Quarterly for Exercise and Sport, Vol 64, pp 377–384, © 1993 by the American Alliance for Health, Physical Education, Recreation and Dance, 1900 Association Drive, Reston, VA 20191.
Note: BP, blood pressure; HDL, high-density lipoprotein. a Adjusted for age. b Adjusted for age and previous variables. *p 0.05: Percent fat: 1 2, 3, 4, 5; 2 3, 4, 5; 3 5; 4 5 Cigarettes/day: 1 3, 4, 5; 2 3, 4, 5 Type A score: 1 3, 4, 5; 2 5 HDL cholesterol: 5 1
88.9 (n 17)
131.1 (n 17)
226.8 (n 17)
41.2 (n 11)
6.7* (n 17)
10.0* (n 11)
5 18.6 (n 36)
3:31 PM
Cigarettes/dayb
CAD Risk Factor % fata
Coronary Artery Disease (CAD) Risk Factors Associated with Physical Fitness and Physical Activity Levels Physical Fitness Level Physical Activity Levels 1 2 3 4 5 1 2 3 4 38.9* 34.9* 26.4* 23.3* 18.9 27.8 25.2 23.1 21.0 (n 11) (n 17) (n 129) (n 94) (n 136) (n 45) (n 130) (n 93) (n 83)
198
2/25/10
Table 9-4
46894_CH09_191_222.qxd Page 198
CHAPTER 9 PHYSICAL INACTIVITY
46894_CH09_191_222.qxd
2/25/10
3:31 PM
Page 199
DISTRIBUTION 199
Table 9-5
Leisure-Time Physical Activity Among Adults 18 Years of Age and Over, by Selected Characteristics: United States, 1998, 2005, and 2006
[Data are based on household interviews of a sample of the civilian noninstitutionalized population] Some Leisure-Time Regular Leisure-Time Inactive1 Activity1 Activity1 Characteristic 1998 2005 2006 1998 2005 2006 1998 2005 2006 Percent of adults 18 years and over, age-adjusted2,3 40.5 40.5 39.5 30.0 29.3 29.5 29.5 30.2 31.0 18 years and over, crude3 40.2 40.5 39.5 30.0 29.3 29.6 29.8 30.1 30.9 Age 18–44 years 35.2 35.9 34.9 31.4 30.5 30.4 33.5 33.7 34.6 18–24 years 32.8 33.3 34.8 30.1 29.1 27.1 37.1 37.4 38.1 25–44 years 35.9 36.7 35.0 31.8 31.0 31.6 32.4 32.4 33.4 45–64 years 41.2 41.2 39.7 30.6 29.7 30.8 28.2 29.1 29.5 45–54 years 38.9 39.5 38.2 31.4 30.1 30.7 29.8 30.4 31.1 55–64 years 44.9 43.6 41.9 29.3 29.2 30.9 25.8 27.2 27.2 65 years and over 55.4 53.9 53.4 24.7 24.9 24.5 19.9 21.3 22.0 65–74 years 49.1 47.8 48.0 26.5 27.0 25.8 24.4 25.3 26.2 75 years 63.3 60.6 59.6 22.4 22.6 23.1 14.3 16.8 17.3 Sex2 Male 37.8 39.1 38.5 28.7 29.2 28.4 33.5 31.8 33.1 Female 42.9 41.7 40.3 31.1 29.5 30.7 26.0 28.8 29.0 Sex and age Male: 18–44 years 32.0 34.4 34.2 30.7 30.5 28.8 37.2 35.1 36.9 45–54 years 37.7 40.2 39.0 29.6 29.4 27.8 32.6 30.4 32.7 55–64 years 44.5 43.4 41.1 26.9 28.0 31.1 28.6 28.7 28.2 64–74 years 45.3 44.7 46.9 23.6 27.5 25.0 31.1 27.8 28.2 75 years and over 57.4 54.1 52.1 21.6 24.0 26.6 20.9 21.9 21.4 Female: 18–44 years 38.2 37.3 35.6 32.0 30.5 32.0 29.8 32.2 32.4 45–54 years 39.9 38.8 37.5 33.0 30.8 33.0 27.1 30.3 29.5 55–64 years 45.2 43.8 42.6 31.5 30.3 31.1 23.3 25.9 26.3 64–74 years 52.2 50.4 49.0 28.7 26.5 26.5 19.0 23.1 24.5 75 years and over 67.0 64.8 64.4 22.9 21.7 20.8 10.1 13.6 14.7 Race2,4 White only Black or African American only American Indian or Alaska Native only Asian only Native Hawaiian or Other Pacific Islander only 2 or more races
38.8 52.2 49.2 39.4 ...
38.6 54.7 42.7 41.0 *
38.2 48.9 32.8 39.8 *
30.5 25.2 19.0 35.2 ...
29.9 24.1 29.0 31.3 *
29.9 26.2 37.8 29.7 *
30.7 22.6 31.8 25.4 ...
31.6 21.2 28.3 27.6 *
31.9 24.9 29.5 30.5 *
40.7
34.2
...
30.9
35.8
...
28.4
30.0
Hispanic origin and race2,4 Hispanic or Latino Mexican Not Hispanic or Latino White only Black or African American only
55.5 56.7 38.8 36.7 52.2
56.7 54.9 38.1 35.3 54.6
53.4 53.9 37.3 35.3 49.0
23.4 23.9 30.7 31.3 25.1
23.3 24.3 30.1 30.9 24.3
23.8 24.2 30.4 31.0 26.4
21.1 19.4 30.5 32.0 22.6
20.0 20.8 31.8 33.8 21.1
22.8 22.0 32.3 33.8 24.7
Education5,6 No high school diploma or GED High school diploma or GED Some college or more
64.8 47.6 30.2
62.9 50.2 30.7
62.3 47.5 29.2
19.4 28.7 34.3
21.5 28.1 32.4
21.2 29.0 33.3
15.8 23.7 35.5
15.7 21.6 36.8
16.5 23.5 37.6
See footnotes at end of table. Source: Reprinted from Health, United States, 2008, p 318–319.
continues
46894_CH09_191_222.qxd
200
2/25/10
3:31 PM
Page 200
CHAPTER 9 PHYSICAL INACTIVITY
Table 9-5
Leisure-Time Physical Activity Among Adults 18 Years of Age and Over, by Selected Characteristics: United States, 1998, 2005, and 2006—continued Some Leisure-Time Regular Leisure-Time Inactive1 Activity1 Activity1 Characteristic 1998 2005 2006 1998 2005 2006 1998 2005 2006 Percent of adults Percent of poverty level2,7 Below 100% 59.4 58.2 56.0 20.5 22.3 23.4 20.1 19.5 20.6 100%–less than 200% 52.2 52.8 50.4 26.2 25.5 25.8 21.6 21.7 23.8 200% or more 34.7 34.5 33.6 32.4 31.4 31.6 33.0 34.1 34.8 Hispanic origin and race and percent of poverty level2,4,7 Hispanic or Latino: Below 100% 68.6 65.9 65.3 18.0 21.1 19.2 13.4 13.0 15.5 100%–less than 200% 60.8 62.7 59.4 21.2 20.4 22.3 18.0 16.9 18.4 200% or more 45.6 48.6 44.3 27.6 26.3 26.7 26.8 25.1 29.0 Not Hispanic or Latino: White only: Below 100% 53.7 52.1 50.8 22.5 23.5 25.5 23.8 24.4 23.7 100%–less than 200% 49.0 47.6 46.1 27.6 27.5 26.3 23.4 24.9 27.5 200% or more 32.7 31.3 31.2 32.9 32.3 32.5 34.4 36.3 36.3 Black or African American only: Below 100% 64.3 65.1 58.7 17.4 18.8 21.8 18.3 16.1 19.4 100%–less than 200% 55.6 60.5 56.2 24.4 23.7 24.3 19.9 15.8 19.5 200% or more 46.0 48.1 41.2 28.7 26.6 29.4 25.3 25.3 29.5 Geographic region2 Northeast 39.4 39.0 36.1 31.3 28.2 31.1 29.4 32.7 32.8 Midwest 37.3 34.3 34.7 31.7 34.3 32.7 31.0 31.4 32.6 South 46.9 47.6 44.8 27.1 25.6 27.2 26.0 26.8 28.0 West 33.9 36.9 38.1 31.6 30.6 28.9 34.6 32.5 33.0 Location of residence2 Within MSA8 39.3 39.2 38.0 30.6 29.7 30.2 30.0 31.1 31.8 Outside MSA8 44.7 45.7 46.4 27.5 27.9 26.6 27.8 26.5 26.9 *Estimates are considered unreliable. Data not shown have a relative standard error of greater than 30%. ***Data not available. 1 All questions related to leisure-time physical activity were phrased in terms of current behavior and lack a specific reference period. Respondents were asked about the frequency and duration of vigorous and light/moderate physical activity during leisure time. Adults classified as inactive reported no sessions of light/moderate or vigorous leisure-time activity of at least 10 minutes duration; adults classified with some leisure-time activity reported at least one session of light/moderate or vigorous physical activity of at least 10 minutes duration but did not meet the definition for regular leisure-time activity; adults classified with regular leisure-time activity reported three or more sessions per week of vigorous activity lasting at least 20 minutes or five or more sessions per week of light/moderate activity lasting at least 30 minutes in duration. See Appendix II, Physical activity, leisure-time. 2 Estimates are age-adjusted to the year 2000 standard population using five age groups: 18–44 years, 45–54 years, 55–64 years, 65–74 years, and 75 years and over. Age-adjusted estimates in this table may differ from other age-adjusted estimates based on the same data and presented elsewhere if different age groups are used in the adjustment procedure. See Appendix II, Age adjustment. 3 Includes all other races not shown separately and unknown education level. 4 The race groups, white, black, American Indian or Alaska Native, Asian, Native Hawaiian or Other Pacific Islander, and 2 or more races, include persons of Hispanic and non-Hispanic origin. Persons of Hispanic origin may be of any race. Starting with 1999 data, race-specific estimates are tabulated according to the 1997 Revisions to the Standards for the Classification of Federal Data on Race and Ethnicity and are not strictly comparable with estimates for earlier years. The five single-race categories plus multiple-race categories shown in the table conform to the 1997 Standards. Starting with 1999 data, race-specific estimates are for persons who reported only one racial group; the category 2 or more races includes persons who reported more than one racial group. Prior to 1999, data were tabulated according to the 1977 Standards with four racial groups and the Asian only category included Native Hawaiian or Other Pacific Islander. Estimates for single-race categories prior to 1999 included persons who reported one race or, if they reported more than one race, Identified one race as best representing their race. Starting with 2003 data, race responses of other race and unspecified multiple race were treated as missing, and then race was imputed if these were the only race responses. Almost all persons with a race response of other race were of Hispanic origin. See Appendix II, Hispanic origin; Race. 5 Estimates are for persons 25 years of age and over and are age-adjusted to the year 2000 standard population using five age groups: 25–44 years, 45–54 years, 55–64 years, 65–74 years, and 75 years and over. See Appendix II, Age adjustment. 6 GED stands for General Educational Development high school equivalency diploma. See Appendix II, Education. 7 Percent of poverty level is based on family income and family size and composition using U.S. Census Bureau poverty thresholds. Missing family income data were imputed for 30%–35% of adults 18 years of age and over in 1998–2006. See Appendix II, Family income; Poverty. 8 MSA is metropolitan statistical area. Starting with 2006 data, MSA status is determined using 2000 census data and the 2000 standards for defining MSAs. For data prior to 2006, see Appendix II, Metropolitan statistical area (MSA) for the applicable standards. Notes: Standard errors are available in the spreadsheet version of this table. Available from: http://www.cdc.gov/nchs/hus.htm. Data for additional years are available. See Appendix III. Source: CDC/NCHS. National Health Interview Survey, family core and sample adult questionnaires.
46894_CH09_191_222.qxd
2/25/10
3:31 PM
Page 201
DISTRIBUTION 201
of vigorous activity lasting at least 20 minutes or five or more sessions per week of light/moderate activity lasting at least 30 minutes in duration.”30, p 319 This level of activity was reported in 2006 by less than onethird of respondents overall. It was most prevalent among males aged 18–44 years (36.9%) and persons with some college or more (37.6%). For females, prevalence improved somewhat from 1998 to 2006 but remained less than 30%. Levels of activity in youth were assessed for the first time in a national probability sample survey of 9- to 13-year-olds and their parents through the Youth Media Campaign Longitudinal Survey.31 The prevalence of self-reported participation in organized physical activity (most often baseball/softball, soccer, or basketball, with a coach, instructor, or leader) and in free-time physical activity (most often bicycle riding or basketball, outside the school day), within the preceding 7 days, was determined by sex, race, year
Table 9-6
of age, race/ethnicity, and parental education and income (Table 9-6). Overall, 38.5% of children reported organized activities at school and 77.4% reported free-time activities, the latter somewhat more frequent among boys. Organized activities were strongly associated with race/ethnicity and parental education and income, favoring non-Hispanic Whites and parents with more than high school education and more than $50,000 annual income, in contrast to nonHispanic Blacks, Hispanics, and parents with less education and income—but still did not reach 50% participation at most. Free-time activity varied little across strata of the population. In this survey, parents were asked to identify any of five types of barriers to their children’s participation in physical activity (Table 9-7). These barriers and their overall frequencies were: transportation problems, 25.6%; lack of opportunities in the area, 20.1%; expense, 46.6%; lack of parents’ time, 21.0%; and
Percentage of Children Aged 9–13 Years Who Reported Participation in Organized and Free-Time Physical Activity During the Preceding 7 Days, by Selected Characteristics—Youth Media Campaign Longitudinal Survey, United States, 2002 Participated in Organized Participated in Free-Time Physical Activity During Physical Activity Preceding 7 Days During Preceding 7 Days % (95% CI*) % (95% CI*)
Characteristic Sex Female Male Age (yrs) 9 10 11 12 13 Race/Ethnicity§ Black, non-Hispanic Hispanic White, non-Hispanic Parental education High school High school High school Parental Income $25,000 $25,000–$50,000 $50,000 Total
38.6 38.3
(2.5) (2.9)
74.1† 80.5†
(2.0) (1.7)
36.1 37.5 43.1 37.7 38.1
(4.0) (4.0) (3.6) (4.1) (4.2)
75.8 77.0 78.9 77.5 78.0
(3.1) (2.7) (3.0) (3.5) (4.0)
24.1† 25.9† 46.6†
(3.8) (4.0) (3.0)
74.7 74.6 79.3
(4.6) (3.9) (1.7)
19.4† 28.3† 46.8†
(4.8) (3.4) (2.5)
75.3 75.4 78.7
(5.7) (2.9) (2.0)
23.5† 32.8† 49.1† 38.5
(3.7) (3.4) (2.6) 2.0) (
74.1 78.6 78.3 77.4
(3.1) (2.5) (2.0) 1.2) (
*Confidence interval † Statistically significant difference (p 0.05). § Numbers for other racial/ethnic populations were too small for meaningful analysis. Source: Reprinted from MMWR, Vol 52, 2003, p 786.
46894_CH09_191_222.qxd
202
2/25/10
3:31 PM
Page 202
CHAPTER 9 PHYSICAL INACTIVITY
Table 9-7
Characteristics Sex Female Male
Percentage of Parents of Children Aged 9–13 Years Who Reported Barriers to Their Children’s Participation in Physical Activities, by Barrier and Selected Characteristics—Youth Media Campaign Longitudinal Survey, United States, 2002 Lack of Lack of Transportation Opportunities Lack of Neighborhood Problems in Area Expense Parents’ Time Safety % (95% CI*) % (95% Cl) % (95% Cl) % (95% Cl) % (95% Cl) 26.9 24.4
(2.7) (2.6)
20.8 19.5
(2.3) (2.0)
47.5 45.8
(3.2) (2.7)
22.8† (2.2) 19.2† (2.4)
17.6† 14.6†
(2.3) (1.9)
Age (yrs) 9 10 11 12 13
25.6 26.2 26.1 24.9 25.2
(3.7) (3.5) (4.3) (3.0) (3.1)
20.5 19.2 21.1 20.0 19.8
(3.1) (3.5) (3.1) (3.7) (3.5)
46.3 46.4 46.0 49.0 45.4
(3.3) (3.9) (4.6) (3.6) (4.2)
20.3 21.6 20.7 20.8 21.5
(3.6) (3.4) (3.2) (3.2) (3.1)
16.9 18.0 16.9 15.9 12.4
(2.9) (3.4) (3.6) (3.0) (2.7)
Race/Ethnicity§ Black, non-Hispanic Hispanic White, non-Hispanic
32.6† 36.9† 18.9†
(4.8) (5.8) (2.3)
30.6† 30.8† 13.4†
(5.7) (3.6) (2.1)
54.9† 62.3† 39.5†
(6.2) (5.5) (2.5)
23.3 23.3 19.1
(5.6) (4.7) (2.1)
13.3† 41.2† 8.5†
(3.3) (5.8) (1.5)
Parental education High school High school High school
42.7† 32.3† 19.3†
(7.2) (3.6) (2.0)
36.7† 23.8† 15.4†
(6.2) (3.7) (2.2)
65.9† 54.8† 39.2†
(7.7) (4.3) (2.5)
27.3 20.5 20.0
(6.6) (3.1) (2.4)
42.9† 18.2† 10.2†
(7.3) (3.4) (1.5)
Parental income $25,000 $25,001-$50,000 $50,000
44.5† 28.9† 14.4†
(4.7) (3.9) (2.1)
35.6† 21.9† 11.5†
(4.4) (3.2) (2.3)
70.6† 53.6† 30.8†
(4.6) (3.4) (2.6)
25.6† (3.5) 20.4 (3.1) 19.0† (2.6)
29.4† 17.8† 8.6†
(4.0) (3.1) (1.6)
Total
25.6
1.9) (
20.1
1.7) (
46.6
2.0) (
21.0
1.6) (
16.1
1.4) (
*Confidence interval. † Statistically significant difference (p 0.05). § Numbers for other racial/ethnic populations were too small for meaningful analysis. Source: Reprinted from MMWR, Vol 52, 2003, p 787.
lack of neighborhood safety, 16.1%. These problems were identified somewhat more frequently for girls than for boys in each case, but age levels between 9 and 13 years made little difference. Large disparities were evident, however, by race/ethnicity and parental education and income, most disfavoring Hispanics. Non-Hispanic Blacks were similarly affected except for lack of neighborhood safety, conspicuously more frequent among Hispanics. National health objectives are defined and tracked under Healthy People 2010, which sets targets and monitors movement toward or away from these targets over the course of the decade. The Mid-Course Review was a comparison mainly of data for 2003 with baseline data from 1999.32 Targeted improvements among adults aged 18 and
older were reductions in no leisure-time activity and increases in regular moderate-or-vigorous or vigorous activity, muscular strength and endurance, and flexibility. In each area, small gains were made. For youth in school grades 9–12, some gains were seen in being physically active if participating in physical education class and in decreased frequency of television viewing for more than 2 hours per day. However, participation in moderate physical activity for at least 30 minutes per day 5 or more days a week, targeted to improve from 27% to 35%, declined to approximately 20%. Vigorous activity that promotes cardiorespiratory fitness 3 or more days a week for 20 or more minutes per occasion also declined, as did participation in daily school physical education.
46894_CH09_191_222.qxd
2/25/10
3:31 PM
Page 203
DISTRIBUTION 203
The International Physical Activity Questionnaire As noted above, WHO in 2002 found no reliable basis for estimating levels of physical activity internationally.20 The World Health Survey, conducted in 70 countries in 2002–2003, included as one optional module the validated International Physical Activity Questionnaire (IPAQ) to assess days and duration of vigorous, moderate, and walking activities during the last 7 days.33 This module was included in 51 predominantly low- and middle-income countries, with a total of 212,021 respondents in the final sample. There were slightly more female than male respondents (55.7%), their mean age was 38.0 years, and 49.7% resided in urban areas. Results are shown for men and women, respectively, by country within the six WHO Regions, agestandardized to the WHO standard population (Figures 9-1A and B). Prevalence of physical inac-
tivity varied widely among the countries and was generally more frequent among women than men. It was least prevalent in the Southeast Asian and Western Pacific Regions, being less than 10% in all participating countries except Malaysia. Greater than 10% prevalence among men was found in several European countries, all three Eastern Mediterranean countries, all but one country in the Americas, and more than one-half of the African countries. For women, the prevalence was greater than 10% in nearly all countries except several in the European Region. The report concluded that, overall, approximately 15% of men and 20% of women from these 51 countries were at risk of chronic diseases because of physical inactivity. These proportions are not comparable with those from the United States because of differences in age distribution (younger in the WHO survey) and some
52.6
Mauritania Swaziland South Africa Namibia Congo Chad Mauritius Zimbabwe Senegal Mali Côte d’lvoire Ethiopia Kenya Zambia Ghana Malawi Burkina Faso Comoros
49.1 44.7 31.8 27.3 18.5 17.5 16.6 16.5
African Region
11.9 11.3 9.5 9.0 8.9 8.8 8.4 7.3 1.6 38.3
Dominican Republic Brazil Uruguay Paraguay Ecuador Mexico Guatemala
26.1 24.2
Region of the Americas
20.2 17.9 16.7 3.6 39.5
United Arab Emirates Pakistan Tunisia
Eastern Mediterranean Region
13.5 11.5
Turkey Spain Slovakia Kazakhstan Bosnia and Herzegovina Czech Republic Slovenia Hungary Croatia Georgia Russian Federation Ukraine Estonia
7.4 6.5 5.3 4.2
India Sri Lanka Nepal Bangladesh Myanmar
7.5 7.1 7.0 7.0
29.0 27.5 14.6 13.6 12.7 10.0 9.9 9.3 8.9
European Region
9.4
Southeast Asian Region 16.5
Malaysia Laos China Vietnam Philippines
9.9 9.3 7.7
Western Pacific Region
5.7 0.0
20.0
40.0
60.0
Physical Inactivity (%)
Figure 9-1A Prevalence of Physical Inactivity for Men in 51 Countries, Grouped by WHO Region, World Health Survey, 2002–2003. Age-Adjusted to WHO Standard Population. Source: Reprinted with permission from American Journal of Preventive Medicine, Vol 24, © 2008. R Guthold, T Ono, KL Strong, S Chatterji, and A Morabia.
80.0
46894_CH09_191_222.qxd
204
2/25/10
3:31 PM
Page 204
CHAPTER 9 PHYSICAL INACTIVITY
Mauritania Swaziland Namibia South Africa Congo Senegal Chad Zimbabwe Mali Mauritius Ethiopia Côte d’lvoire Ghana Malawi Zambia Kenya Burkina Faso Comoros
3.8
Dominican Republic Brazil Uruguay Ecuador Paraguay Mexico Guatemala
18.1 17.7 17.3 15.4 14.3 13.0 11.2 9.5
22.1 21.1
17.7
4.2
United Arab Emirates Pakistan Tunisia
4.6 4.2 3.5
Bangladesh India Myanmar Sri Lanka Nepal
8.6 7.7 7.0 6.9 6.4
13.8 11.9
10.2
Malaysia Laos China Vietnam Philippines
9.2 9.1 0.0
32.2
24.5 24.4
21.2
72.0
56.4
African Region
44.2
30.4 28.7 27.7
Region of the Americas
59.0
27.6
18.9
Turkey Spain Bosnia and Herzegovina Slovenia Kazakhstan Georgia Hungary Czech Republic Croatia Slovakia Estonia Russian Federation Ukraine
48.6 47.6
43.5
32.9
17.1
Eastern Mediterranean Region
European Region
27.0
15.6 14.8 14.1
12.5
Southeast Asian Region 23.6
15.1
Western Pacific Region 20.0
40.0
60.0
80.0
Physical inactivity (%)
Figure 9-1B Prevalence of Physical Inactivity for Women in 51 Countries, Grouped by WHO Region, World Health Survey, 2002–2003. Age-Adjusted to WHO Standard Population. Source: Reprinted with permission from American Journal of Preventive Medicine, Vol 24, © 2008. R Guthold, T Ono, KL Strong, S Chatterji, and A Morabia.
methodologic aspects of the surveys, even though similar definitions of physical inactivity were used. However, differences may also reflect the long-term trends of increasing inactivity with greater urbanization and economic development, resulting in higher true prevalence in the United States than in the majority of surveyed countries.
CARDIOVASCULAR-RELATED EFFECTS Population Differences The contribution of physical activity to prediction of differences in coronary heart disease or total mortality between populations was one focus of the Seven Countries Study of 16 cohorts of men aged 40–59 years at entry between 1958 and 1964.34 Keys described the physical activity of the men in most cohorts in that study as primarily occupational, with little significant activity at leisure: “Except for the Finnish and American cohorts, the idea of exercise for its own sake was considered a little mad by men be-
yond the age of 40; that view prevails today in many rural areas.”34, p 197 Based mainly on occupation, but adjusted for substantial outside activity, physical activity was categorized into three levels: 1 (sedentary), 2 (moderate), and 3 (heavy work, very active). The frequency distributions of activity levels for the cohorts, excluding men with any evidence of cardiovascular disease, are shown in Table 9-8. The variation in proportions of men who were sedentary is striking, even excluding the Belgrade professors, ranging from 3.9% to 49.7% sedentary (Tanushimaru, Japan, and US railroad workers, respectively). Heavy work was absent in two cohorts but accounted for 60–80% of men in most other cohorts; these were predominantly moderately to very active men. Baseline differences in characteristics of sedentary versus the most active men in some cohorts were older age, higher blood pressure, higher resting pulse rate, higher serum cholesterol concentration, and lower income, although some of these relations were found in only a few cohorts. The relation of baseline physical activity level to death from all causes and
46894_CH09_191_222.qxd
2/25/10
3:31 PM
Page 205
CARDIOVASCULAR-RELATED EFFECTS 205
Table 9-8
Distribution by Physical Activity Class at Entry in Men Free of Cardiovascular Disease, Seven Countries Study Activity Class (%) Activity Class (%) 1 2 3 Cohort 1 2 3 49.7 35.3 0.0 Montegiorgio 5.8 25.8 68.4 7.9 11.6 80.5 Zutphen 23.7 65.7 10.6 17.1 9.2 73.7 Crete 6.6 30.9 62.5 3.9 36.6 59.5 Corfu 31.0 37.8 31.2 9.6 12.6 73.8 Rome railroad 22.0 40.4 37.6 7.5 12.5 80.0 Velika Krsna 8.5 24.5 67.0 7.6 17.1 75.3 Zrenjanin 35.2 35.2 29.6 10.5 18.8 70.7 Belgrade 99.4 0.6 0.0
Cohort US railroada Dalmatia Slavonia Tanushimaru East Finland West Finland Ushibuka Crevalcore
Note: Activity class 1 = sedentary; 2 = moderate activity; 3 = heavy work, very active. a 15.1% not readily classifiable between activity classes 1 and 2. Source: Reprinted with permission of the publisher from Seven Countries by A Keys, Cambridge, Mass: Harvard University Press, © 1980 by the President and Fellows of Harvard College.
from coronary heart disease was examined as was the relation of baseline activity to “hard” coronary heart disease (acute myocardial infarction or coronary death) and incidence of all coronary heart disease. For all coronary heart disease, fatal or nonfatal,
the relation was as shown in Figure 9-2. The overall regression coefficient, r –0.13, indicates that incidence of coronary disease tended to decrease as the proportion of men in a population who were sedentary increased.
2500
Y = 10–Year Any CHD per 10,000 Men
F
2000
1500
Z
I
1000
r = –0.13 Y = 997 – 2.6X
R 500
0 0
Y J
B
G
20
40
60
80
100
X = % of Men Sedentary
Note: B, Belgrade; F, Finland; G, Greece; I, Rural Italy; J, Japan; R, Rome railroad; Y, Yugoslavia except Belgrade; Z, Zutphen. Figure 9-2 Ten-Year Age Standardized Incidence of Any CHD (Any Diagnosis of Coronary Heart Disease) in Population Samples Versus Percentage of Men in Those Samples, Seven Countries Study. Source: Reprinted with permission of the publisher from Seven Countries by A Keys, Cambridge, Mass: Harvard University Press, © 1980 by the President and Fellows of Harvard College.
46894_CH09_191_222.qxd
206
2/25/10
3:31 PM
Page 206
CHAPTER 9 PHYSICAL INACTIVITY
It is apparent by inspection of the figure that this result is heavily influenced by the extreme positions of both Finland (high incidence, low proportion sedentary) and Belgrade (low incidence, high proportion sedentary). In the remaining subset of countries, higher mortality appears to be related to higher prevalence of sedentary living, but such post hoc selection of the data can offer only very frail inferences and fails to explain the two exceptions. Within some cohorts, but not others, there was a tendency for sedentary men to exhibit higher mortality than men whose physical activity was at level 2 or 3 (not shown). This occurred in rural Italy, among Rome railroad workers, and in Greece. Overall, the results did not demonstrate an important contribution of average levels of physical activity to the wide differences in population rates of death or incident coronary heart disease. Individual Differences Within populations, variation in risks of cardiovascular diseases in relation to physical inactivity, other levels of activity, and physical fitness has been reported from numerous studies, albeit with diverse methods of assessment of activity and fitness. The Surgeon General’s report summarizes 36 studies of physical activity and 7 studies of physical fitness in relation to coronary heart disease, 14 studies of physical activity and stroke, and 6 studies of physical activity and hypertension.2 For each report the study population, definitions of physical activity or fitness and the cardiovascular outcome, main findings, evidence for a dose response, adjustment for confounders, and other comments are presented in table form. The Surgeon General’s report concludes that the epidemiologic literature supports the presence of an inverse association between physical activity or physical fitness and coronary heart disease and hypertension but is unclear with respect to stroke. A more detailed evaluation of studies of physical activity and incidence of coronary heart disease was reported by Powell in 1987, including systematic evaluation of the quality of each study with respect to measures of physical activity and coronary heart disease and epidemiologic methods in general.35 Occupational or nonoccupational cohort studies constituted 36 of the total of 43 studies. Most studies were conducted in the United States, with several in the United Kingdom and others mainly in Europe, all chiefly among men younger than 65 years of age. Results were available for women from only five of these studies. The preponderance of studies that could be scored, in men, indicated an inverse relation between
physical activity and incidence of coronary heart disease. This finding was most frequent among the studies rated as best in quality. It was concluded that an inverse, causal relation was supported by these studies. Further, the magnitude of the relative risk of inactivity was judged to be similar to that for hypertension, elevated blood cholesterol concentration, and smoking. High prevalence of inactivity in the US population suggested that this widespread risk factor makes a large contribution to the incidence of coronary heart disease and therefore warrants public policy to support increased physical activity in the population. Berlin and Colditz used Powell’s ratings and selected additional studies published through the 1980s for a meta-analysis of the relative risk of coronary heart disease within particular groups of studies.36 For the group of nonoccupational studies, all of which were included in Powell’s review, several categories of coronary outcomes were considered, when reported: all CHD, CHD death, MI, MI sudden death, or angina pectoris. Various comparisons were also considered: moderate versus high activity, moderate and sedentary versus high activity, and sedentary versus high activity. Comparison of moderate and sedentary versus high activity yields relative risks from 1.1 to 2.3 for different coronary disease outcomes, most strongly for all coronary disease (RR 1.6, CI 1.3–1.8) and for myocardial infarction plus sudden death (RR 2.3, CI 1.5–3.6). A similar range of results was found for occupational as for nonoccupational studies, with higher relative risk estimates in the higher-quality studies. The authors concluded that physical activity protects against coronary heart disease but added that this effect had not been shown to be independent of other risk factors. The question of independence of physical activity as a risk factor was addressed in the second posttrial follow-up report from the Multiple Risk Factor Intervention Trial (MRFIT), which addressed the question of an independent contribution of physical activity to reduced coronary disease risks in men whose activity level was greater than sedentary.37 More than 12,000 high-risk men were enrolled in this long-term trial of risk-factor reduction. They received no intervention on physical activity but completed the Minnesota Leisure Time Physical Activity questionnaire. Scores were recorded in minutes per day of mostly light- or moderate-intensity activity, less than 25 kjoule/min. Mortality over 10.5 years of follow-up (including 3.5 posttrial years) was compared among groups defined by tertiles of physical activity score, from low (1) to high (3). Figure 9-3
46894_CH09_191_222.qxd
2/25/10
3:31 PM
Page 207
CARDIOVASCULAR-RELATED EFFECTS 207
indicates that for cardiovascular death and coronary heart disease death, but not for cancer death, mortality was highest for men with least activity and did not differ appreciably between intermediate and higher activity groups. Table 9-9 further demonstrates that levels 2 and 3 did not differ in mortality reduction relative to level 1. The lower portion of the table reports results by proportional hazards regression, which adjusts for baseline risk factors (age, diastolic blood pressure, total cholesterol concentration, and number of cigarettes smoked per day). The mortality reduction associated with leisure-time physical activity was essentially independent of these other factors, given that the adjustment resulted in little change in the risk ratios. To relate these results to those calculated in the foregoing meta-analysis, results for level 3 in Table 9-9 can be inverted to compare level 1 with level 3; for example, the result for coronary heart disease mortality adjusted for age alone (upper portion of Table 9-9), shown as a risk ratio of 0.84, would correspond with an increased risk of lowest versus highest activity of 1/0.84, or 1.19, about a 20% increase. This is notably less than the relative risks found in the
meta-analysis, perhaps because of a narrower range of activity between strata in MRFIT compared with that in the meta-analysis. Studies including women are fewer, but a study of cardiorespiratory fitness—measured by maximal exercise treadmill test—in women and men in the Aerobics Center Longitudinal Study provides an example.38 Among women there were fewer participants (7080 women and 25,431 men) and fewer deaths (21 cardiovascular and 89 total in women and 226 and 601 deaths, respectively, in men). Tables 9-10 and 9-11 compare observations on predictors of death between women and men. “Low fitness” characterized the 20% of each age-sex group with the poorest cardiorespiratory fitness. The risk of cardiovascular death for low- versus high-fitness categories was 2.42 (CI 0.99–5.92) for women and 1.70 (CI 1.28–2.25) for men, after adjustment for all other factors shown in the tables. An expected effect of this adjustment would be to underestimate the total effect of fitness because it removes the favorable influence of fitness on the factors included in the adjustment. For both cardiovascular and all-cause mortality, the adjusted relative risk for low fitness was greater for women than for men.
CHD Death–10.5-Year Follow-Up by LTPA Tertile
Cumulative Event Rate per 1,000
Cumulative Event Rate per 1,000
CVD Death–10.5-Year Follow-Up by LTPA Tertile
Time from Randomization in Years
Cancer Deaths–10.5-Year Follow-Up by LTPA Tertile
All Deaths–10.5-Year Follow-Up by LTPA Tertile
Cumulative Event Rate per 1,000
Cumulative Event Rate per 1,000
Time from Randomization in Years
Time from Randomization in Years
Time from Randomization in Years
Figure 9-3 Cumulative 10.5-Year Mortality Rates per 1000 for Cardiovascular Diseases (CVD) (Top Left), Coronary Heart Disease (CHD) (Top Right), Multiple Risk Factor Intervention Trial. Source: Reprinted from AS Leon and J Connett, MRFIT Research Group, American Journal of Epidemiology, Vol 20, p 692, © 1991, The Johns Hopkins University School of Hygiene and Public Health.
46894_CH09_191_222.qxd
208
2/25/10
3:31 PM
Page 208
CHAPTER 9 PHYSICAL INACTIVITY
Table 9-9
Risk Ratios and Major Endpoints (and 95% Confidence Limits) by Percentile of Leisure-Time Physical Activity in Men in the Multiple Risk Factor Intervention Trial Fatal Endpoints Tertile 1 Tertile 2 Tertile 3 Age-Adjusted Risk Ratios Cardiovascular disease 1.00 0.78* (0.63, 0.96) 0.79* (0.68, 1.04) Coronary heart disease 1.00 0.73* (0.57, 0.92) 0.84 (0.62, 1.00) Cancer 1.00 1.15 (0.85, 1.52) 1.00 (0.66, 1.21) All causes 1.00 0.85* (0.73, 0.99) 0.87* (0.74, 1.01) Cardiovascular disease Coronary heart disease Cancer All causes
1.00 1.00 1.00 1.00
Proportional Hazards Regression 0.81 * (0.66, 1.01) 0.75* (0.59, 0.96) 1.22 (0.91, 1.63) 0.89 (0.77, 1.04)
0.89 (0.72, 1.09) 0.82 (0.65, 1.04) 1.06 (0.78, 1.44) 0.92 (0.79, 1.07)
Note: Regression of endpoints by age (years), level of diastolic blood pressure (mm Hg), total cholesterol (mmol/l), number of cigarettes per day, treatment group (Special Intervention or Usual Care), and tertile of physical activity. *p 0.05. Source: Reprinted from AS Leon and J Connett (for MRFIT Research Group), American Journal of Epidemiology, Vol 20, p 693, © 1991, by permission of Oxford University Press.
Several recent observations add to the case for increasing activity levels as a component of cardiovascular disease prevention. In a large, nationally representative cohort, the Health and Retirement Study, the relation of physical activity levels and mortality was investigated among persons in three strata of cardiovascular risk.39 Being physically active had its greatest risk-reducing impact among those at highest risk, although being sedentary was especially concentrated in this stratum of the population. A review of prospective epidemiologic studies suggested a linear dose-response relation between physical activity and coronary heart disease, with benefit seen in women and men, middle-aged and older persons, including men with existing coronary heart disease.40 This review also suggested benefit of moderateintensity activity in reducing risk of stroke. A qualitative review identified nine studies of physical activity and two of cardiorespiratory fitness in older subjects ( 50 years at entry), mainly in men, that were reviewed to determine overall effects on fatal or nonfatal coronary heart disease.41 Eight of 11 studies showed inverse relations between being fit or active and development of coronary heart disease, five of them with statistically significant results. It was concluded that benefits of physical activity may extend to older adult men. Reports of the relation between physical activity and body mass index in adolescents, one from the Youth Behavioral Risk Survey (YRBS) and another from the Trial of Activity for Adolescent Girls (TAAG), found inverse associations between overweight status and physical activity in boys (YRBS) and between gain in percent body fat and minutes of
moderate to vigorous physical activity in longitudinal follow-up in girls (TAAG).42,43 The question of adverse effects of physical activity has also been addressed. Studies have focused on precipitation of sudden death and other fatal and nonfatal coronary events as well as musculoskeletal injuries as potential hazards. Many reports concerning sudden death were reviewed by Kohl and colleagues, who noted that the overall risk of sudden death is less among habitually active than among sedentary persons even though an excess risk is obtained specifically during periods of vigorous activity.44 Evidence on “triggering” of acute coronary events has been reviewed by Mittleman, who addressed physical activity in general and sexual activity in particular, in addition to emotional stress.45 With respect to physical activity, use of an imaginative “casecrossover” design revealed that risk of acute myocardial infarction was greater in the first hour after heavy physical exertion than at other times. However, risk of myocardial infarction associated with heavy exertion was lowest for those with this level of exertion at least five times per week. The increased risk in the first hour after exertion appeared to be largely mitigated by better habits of exercise. For the effect of sexual activity, also, any excess risk of precipitating an acute coronary event appeared to be reduced to a negligible degree by regular physical activity. Further discussion of this issue by Wannamethee and Shaper based in part on experience of the British Regional Heart Study suggested that persons with hypertension might be at special risk of coronary events with vigorous exercise.40 Several supporting observations were noted, but the issue was considered unresolved.
3:31 PM
Source: Reprinted with permission from Journal of the American Medical Association, Vol 276, No 3, p 207, Copyright 1996, American Medical Association.
2/25/10
Note: All comparisons are dichotomies, with the referent category being the low-risk group (relative risk 1) and the high-risk group data shown in the table. Data for the reference categories are not included but can be estimated for each predictor by subtracting the values in the table for the high-risk group from the totals (25,341 men, 211,996 man-years, 601 deaths from all causes, 226 deaths from cardiovascular disease; 7,080 women, 52,982 woman-years, 89 deaths from all causes, 21 from cardiovascular disease). Ellipses indicate “not applicable.” a Adjusted for age and examination year. b Adjusted for age, examination year, and each of the other variables in the table. c Crude rate.
Cardiovascular Disease Mortality and All Cause Mortality Risk Analyses for Selected Mortality Predictors, Women’ Aerobics Center Longitudinal Study, 1970–1989 Cardiovascular Disease All Causes Relative Risk Relative Risk Person-Years Death Rate/ Adjustedb Death of Follow-Up 10,000 (95% Rate/10,000 Adjustedb No. of (% of Person- No. of PersonConfidence No. of Person(95% Confidence Mortality Predictor Subjects Years) Deaths Yearsa Adjusteda Interval) Deaths Yearsa Adjusteda Interval) Low fitness 1,352 13,086 (25) 11 7.7 2.79 2.42 40 28.8 2.32 2.10 (20% least fit) (0.99–5.92) (1.36–3.26) Current or recent 1,321 10,811 (20) 5 6.0 1.73 1.70 27 29.0 2.12 1.99 smoker (0.58–4.97) (1.25–3.17) Systolic blood 416 3,959 (7) 8 7.6 2.06 1.47 15 15.1 0.89 0.76 pressure 140 mm Hg (0.55–3.93) (0.41–1.40) Cholesterol 6.2 mmol/l 1,223 9,034 (17) 8 3.9 0.99 0.74 31 18.9 1.16 1.09 ( 240 mg/dl) (0.28–1.95) (0.68–1.74) Either parent dead of 1,788 13,474 (25) 6 3.2 0.76 0.58 24 12.9 0.71 0.70 coronary heart disease (0.20–1.72) (0.43–1.16) Body mass index 777 5,486 (10) 2 2.0 0.48 0.28 15 19.5 1.18 0.94 27 kg/m2 (0.06–1.26) (0.52–1.69) Fasting glucose 6.7 148 1,202 (2) 3 14.2 3.80 4.10 7 33.3 2.03 1.79 mmol/l ( 120 mg/dl) (1.11–15.2) (0.80–4.00) Abnormal 350 2,816 (5) 10 17.3 5.38 5.02 16 26.2 1.61 1.55 electrocardiogram (1.90–13.3) (0.87–2.77) Chronic illness 958 7,085 (13) 7 6.9 1.98 1.66 18 18.1 1.09 1.05 (0.59–4.64) (0.61–1.82) Totals 7,080 52,982 (100) 21 4.0c ... ... 89 16.8c … …
Table 9-10
46894_CH09_191_222.qxd Page 209
CARDIOVASCULAR-RELATED EFFECTS 209
3:31 PM
Source: Reprinted with permission from Journal of the American Medical Association, Vol 276, No 3, p 207, © 1996, American Medical Association.
2/25/10
Note: All comparisons are dichotomies, with the referent category being the low-risk group (relative risk 1) and the high-risk group data shown in the table. Data for the reference categories are not included but can be estimated for each predictor by subtracting the values in the table for the high-risk group from the totals (25,341 men, 211,996 man-years, 601 deaths from all causes, 226 deaths from cardiovascular disease; 7,080 women, 52,982 woman-years, 89 deaths from all causes, 21 from cardiovascular disease). Ellipses indicate “not applicable.” a Adjusted for age and examination year. b Adjusted for age, examination year, and each of the other variables in the table. c Crude rate.
Cardiovascular Disease Mortality and All-Cause Mortality Risk Analyses for Selected Morality Predictors, Men, Aerobics Center Longitudinal Study, 1970–1989 Cardiovascular Disease All Causes Relative Risk Relative Risk Person-Years Death Rate/ Adjustedb Death of Follow-Up 10,000 (95% Rate/10,000 Adjustedb No. of (% of Person- No. of PersonConfidence No. of Person(95% Confidence Mortality Predictor Subjects Years) Deaths Yearsa Adjusteda Interval) Deaths Yearsa Adjusteda Interval) Low fitness 5,223 54,729 (26) 111 20.0 2.69 1.70 250 45.5 2.03 1.52 (20% least fit) (1.28–2.25) (1.28–1.82) Current or recent 6,730 60,829 (29) 82 16.6 2.01 1.57 222 42.7 1.89 1.65 smoker (1.18–2.10) (1.39–1.97) Systolic blood 2,759 26,398 (12) 87 19.5 2.07 1.34 184 43.6 1.67 1.30 pressure 140 mm Hg (1.00–1.80) (1.08–1.58) Cholesterol 6.2 mmol/l 6,025 51,262 (24) 106 16.5 1.86 1.65 229 37.0 1.45 1.34 ( 240mg/dl) (1.26–2.15) (1.13–1.59) Either parent dead of 6,499 53,440 (25) 84 14.3 1.51 1.18 203 33.1 1.24 1.07 coronary heart disease (0.89–1.57) (1.13–1.59) Body mass index 8,198 65,534 (31) 96 14.9 1.70 1.20 223 34.3 1.33 1.02 27 kg/m2 (0.91–1.58) (0.86–1.22) Fasting glucose 6.7 1,396 13,229 (6) 36 15.4 1.49 0.95 92 44.3 1.63 1.24 mmol/l ( 120 mg/dl) (0.66–1.37) (0.98–1.56) Abnormal 1,866 15,680 (7) 99 36.7 4.28 3.01 158 54.0 2.05 1.64 electrocardiogram (2.24–4.04) (1.34–2.01) Chronic illness 4,802 41,016 (19) 124 26.3 3.80 2.52 242 49.8 2.15 1.63 (1.89–3.36) (1.37–1.95) Totals 25,341 211,996 (100) 226 10.7c ... ... 601 28.3c ... ...
210
Table 9-11
46894_CH09_191_222.qxd Page 210
CHAPTER 9 PHYSICAL INACTIVITY
46894_CH09_191_222.qxd
2/25/10
3:31 PM
Page 211
PREVENTION AND CONTROL 211
These and other reports emphasize the overall benefit of habitual exercise, the hazards of rapid change from sedentary to active status, and the appropriate caution in making this change, especially for adults. Due caution and, if necessary, initial supervision are expected to minimize risks of both acute coronary events and musculoskeletal injury. Public Health Impact The contribution of physical inactivity to excess deaths has been addressed recently by Danaie and others, who estimated the numbers of preventable deaths in the United States that were due to modifiable factors.46 The reference exposure level was taken to be the hypothetical circumstance in which the whole population is “highly active ( 1 hour/week of vigorous activity and at least 1,600 met min/ week).” An explanatory note supported this “theoretical-minimum-risk exposure distribution” on the basis of “multiple prospective studies that report beneficial effects of physical activity continuing above the current recommended levels.”46, pp 3–4 They then considered four categorical levels of activity—highly active, recommended level active, insufficiently active, and inactive. Age-specific relative risks were presented for ischemic heart disease and ischemic stroke for each activity level below “highly active”—for example, for ages 30–69 years, the increased relative risks were 1.15, 1.66, and 1.97 for the three respective categories. (Corresponding risk gradients were shown for breast and colon cancer and diabetes as well.) In the aggregate, physical activity below “highly active” was taken to account for 191,000 preventable deaths in 2005—103,000 among women and 88,000 among women. Tackling this and the other preventable causes of death has great life-saving potential, as discussed under Prevention and Control, as follows. For estimation of the global impact of physical inactivity, the report Global Burden of Disease and Risk Factors, from the Disease Control Priorities in Developing Countries Project, used an analogous approach.47 Here, the theoretical-minimum-risk exposure was considered to be “all persons having at least 2.5 hours per week of moderate-intensity activity or equivalent (4,000 KJ/week).”47, p 243 Separately for ischemic heart disease and for cerebrovascular disease, for mortality, years of life lost, and disabilityadjusted life years (DALYs) lost, contributions of physical activity less than the reference level were estimated. Taking all ages from 15–29 years and older, and both men and women, the overall population attributable fraction for physical inactivity was 19–21% for ischemic heart disease and 6–7% for cerebrovas-
cular disease across these three measures. There was little variation in the contribution by age until a slight decrease at age 80 years. Worldwide, these data translate to 1.4 million deaths and 17.7 million DALYs from ischemic heart disease and 303,000 deaths and 4.7 million DALYs from cerebrovascular disease, all due to being less physically active than 2.5 hours per week at moderate intensity.
PREVENTION AND CONTROL The first report to address prevention of atherosclerosis and related conditions is believed to be A Statement on Arteriosclerosis. Main Cause of “Heart Attacks” and “Strokes,” a 21-page booklet published in 1959 by the National Health Education Committee, Inc. (see Chapter 17, “Recommendations, Guidelines, and Policies”).48 The document was “a simple guide which would give the average man and woman something he or she could do in cooperation with the physician to minimize the hazards of arteriosclerosis, main cause of ‘heart attacks’ and ‘strokes.’”48, p 1 The Statement was based on the scientific literature of the day, which gave support to advice on overweight, blood pressure, cholesterol, “excessive” cigarette smoking, and family history. Physical activity was advocated as a regular habit, with caution against unconditioned overexertion—for example, avoiding the sudden collapse of an otherwise sedentary gentleman while shoveling snow. From 1965 through 1996, 33 recommendations were published in the United States alone, culminating most recently in the 2008 Physical Activity Guidelines for Americans.2,10 These and other current recommendations build on recent experience and advocate measures to be taken at individual, community or population-wide, and global levels. A question underlying recommendations at any level is: How much activity is necessary to have a preventive effect? A 1995 report of a consensus conference on physical activity and public health summarized several studies examining the fitness or activity level necessary for cardiovascular benefit.12 Only men were included in five of the six studies addressed, but women were included in one of them, with results comparable to those for men: On an 8-point scale of increasing fitness or activity, the relative risk of cardiovascular mortality was reduced over the first categories of increased fitness or activity. With higher categories, mortality either continued to be reduced or leveled at about point 4 of 8 on the scale. That pattern was interpreted as evidence of a continuous gradient of benefit over a wide range of fitness or activity levels.
46894_CH09_191_222.qxd
212
2/25/10
3:31 PM
Page 212
CHAPTER 9 PHYSICAL INACTIVITY
That review was updated to represent reports from 2000–2003, as shown in Figure 9-4.49 The graphic is analogous to that in the preceding report, but now represents coronary incidence as well as mortality, and fitness and activity are on a 5-point scale. The evidence is much the same: Coronary risk is less at greater degrees physical fitness or activity, with variation among studies––some showing little or no further decrease in risk once a moderate level is attained and others showing continued decrease at higher levels of fitness or activity. A question posed by Erlichman and others was whether these recommendations and the attendant health education policies for cardiovascular disease prevention were sufficient for other health objectives, in particular avoidance of unhealthy weight gain.50 Their conclusion was that weight stability was more closely associated with vigorous than with moderate activity, equating to need for a Physical Activity Level (PAL, the ratio of total energy expenditure to the basal metabolic rate) 1.8:50, p 285 “The policy implications of an optimum PAL of 1.8 are quite different from those tra-
Lee 2001 Manson 2002 Tanasescu 2003
Crespo 2002 Davey 2000 Cheung 2003
ditionally [in cardiovascular disease prevention] linked to individualized health education messages and require new population strategies to change people’s environments so that these become more conducive to spontaneous and sustained physical activity.” For women, it has also been noted that weight-bearing exercise is an important component of physical activity for prevention and treatment of osteoporosis.51 Individual Measures Regarding individual-level intervention, the Surgeon General’s Report included a review of 13 studies of individual-level intervention approaches among adults.2 The studies were typically small, with about 50 to 350 participants, sometimes included women, and were often conducted in employee groups. The range of study duration was from 10 weeks to 2 years. The interventions principally involved group exercise activities. No individual-level intervention studies in children were described. The Surgeon General’s report concluded as follows:2, p 234
Wagner 2002 Yu 2003 Batty 2002
Wannemethee 2000 Hu 2001
110 100
CVD Incidence or Mortality
90 80 70 60 50 40 30 20 1 Lowest
2
3 Level of Physical Activity/Fitness
4
5 Highest
Figure 9-4 Risk Reduction Estimates of the Relationship Between Physical Activity and Cardiovascular Disease, from Epidemiological Studies Published 2000–2003. Source: Reprinted with permission from Journal of Science and Medicine in Sport, Vol 7(Suppl 1), AE Bauman, p 9.
46894_CH09_191_222.qxd
2/25/10
3:31 PM
Page 213
PREVENTION AND CONTROL 213
The review of adult intervention research literature provides limited evidence that interventions to promote physical activity can be effective in a variety of settings using a variety of strategies. . . . Intervention studies with adults were often conducted over a brief period of time, had little or no follow-up, and focused on the endpoint of specified vigorous physical activity rather than on moderate-intensity physical activity or total amount of activity. Studies used different endpoints . . . making them difficult to compare. . . . Few if any studies compared their results to a standard of effectiveness, such as recommended frequency or duration of moderate or vigorous physical activity. The special case of rehabilitation after myocardial infarction requires mention because it is the one area in which clinical trials of physical activity have been conducted with evaluation of intervention in relation to disease outcomes. Review of 22 randomized trials of rehabilitation with exercise indicates benefit for up to 3 years in reduced total and cardiovascular mortality, fatal reinfarction and, at 1 year, sudden death.52 Nonfatal reinfarction was not reduced in frequency. Interpretation of these results is limited by the concurrent interventions in some studies, so the indicated effect may not be attributable to exercise alone. By 2001, a review of trials of cardiac rehabilitation programs—exercise alone or comprehensive programs—was found to result in a 27% decrease in post-cardiac-event mortality.53 Principles of exercise programs designed for rehabilitation of those surviving an acute coronary event have been presented by the American Heart Association.54 A systematic review of studies of health professional advice to increase physical activity among persons age 16 or older found that such advice was effective, although the duration of studies gave no evidence of effect lasting beyond 1 year.55 Eleven studies with selfreported physical activity as the outcome and seven studies of physical fitness documented improvement, but target levels of activity were not attained significantly more often in the intervention groups. The US Preventive Services Task Force (USPSTF), on the contrary, found the evidence “insufficient to recommend for or against behavioral counseling in primary care settings to promote physical activity.”56 Recommendations for overcoming physical inactivity are presented by multiple organizations and agencies today. Although there are some differences, they generally share elements of recommended intensity, duration, and timing of activity. For example, the 2007 Physical Activity and Public Health: Updated
Recommendations for Adults from the American College of Sports Medicine and the American Heart Association describes aerobic activity for adults as follows (for evidence grades, e.g., (I (A)), see Chapter 19, “Evidence and Decision Making”):57, p 1425 To promote and maintain health, all healthy adults aged 18–65 yr need moderate-intensity aerobic physical activity for a minimum of 30 min on five days each week or vigorous-intensity aerobic activity for a minimum of 20 min on three days each week. [I (A)] Also, combinations of moderate- and vigorous-intensity activity can be performed to meet this recommendation. [IIa (B)] For example, a person can meet the recommendation by walking briskly for 30 min twice during the week and then jogging for 20 min on two other days. Moderate-intensity aerobic activity, which is generally equivalent to a brisk walk and noticeably accelerates the heart rate, can be accumulated toward the 30-min minimum from bouts lasting 10 or more minutes. [I (B)] Vigorous-intensity activity is exemplified by jogging, and causes rapid breathing and a substantial increase in heart rate. This recommended amount of aerobic activity is in addition to routine activities of daily living of light intensity (e.g., self care, cooking, casual walking or chopping) or lasting less than 10 min in duration (e.g., walking around home or office, walking from the parking lot). Guidelines specific for children and adolescents, older adults, women during pregnancy and the postpartum period, adults with disabilities, and people with chronic medical conditions are provided in other sources.10 This approach is intended to make physical activity an easy and natural part of daily living rather than a discrete, prescribed task—and it should be available to everyone. To whom, then, do the recommendations apply? In the arena of cardiovascular disease prevention, individual-level guidelines are usually considered to apply to “high-risk” individuals, the relatively small subgroup at the extreme of a risk distribution. This makes it meaningful to distinguish, as proposed by Rose (see Chapter 18, “Strategies of Prevention”), between individual- and mass or population-level strategies. Morris, in a 1992 review of the field of physical activity and cardiovascular disease prevention, questioned this distinction with respect to physical activity.3 Because physical inactivity is so highly prevalent, those at high risk encompass so large a proportion of the population that “high-risk” and “population-wide” are equivalent. Rose also cited
46894_CH09_191_222.qxd
214
2/25/10
3:31 PM
Page 214
CHAPTER 9 PHYSICAL INACTIVITY
what he termed “the prevention paradox”: Interventions that benefit society most are those that reach the great numbers of people at only somewhat elevated risk, who contribute the greatest proportion of cases relative to the much smaller numbers of people at high risk; yet there is relatively less personal benefit to these individuals because their individual probabilities of experiencing events are small. Here, Morris points to the immediate benefit of physical activity, such that those who take it up “will rapidly feel and function better as a result of the manifold benefits that exercise confers.”3, p 252 Community or Population-Wide Measures In reviewing studies aimed at population-wide intervention in adults, the Surgeon General’s report summarized five studies in targeted communities—these included the physical activity components of three large community intervention trials (the Stanford Five City Program and Minnesota and Pawtucket, RI, Heart Health Programs), three worksite programs, and four broad communications campaigns.2 With respect to the community studies, it was concluded that results were generally disappointing but that “community coalitions, widespread community involvement, and well-organized community efforts appear to be important” in increasing physical activity levels.2, p 229 Worksite programs were still too few to draw conclusions regarding critical program elements. Tentatively, however, it appeared that widespread employee involvement, organizational commitment, and supportive policies and programs were important. It was found that communication strategies had only limited impact, although well-placed cues to action, such as signage near elevators pointing to the stairs, appeared to be effective. Additional studies in special populations focused on minorities, persons at risk for chronic disease, or older persons. Population studies in children were also aimed at increasing physical activity and were either school programs or school and community programs. They ranged from a few weeks to 4 years in duration and targeted age groups from as early as grade 3 to as late as grade 12. The interventions included classroom health education, modification of physical education classes, or both. The largest and most recent of these studies was the Children’s Activity Trial for Cardiovascular Health (CATCH), some of whose results are shown in Figure 9-5. Involving more than 3000 students in 96 schools, the CATCH interventions resulted in change in physical education classes. The percentage of class time occupied by moderate to vigorous activity increased from less than 40% to more than 50% in the intervention schools but re-
mained less than 40% in the control schools. Vigorous activity constituted about 20% of time in intervention schools and 18% in control schools. When the Task Force on Community Preventive Services reviewed evidence on community programs to increase physical activity in 2002, three broad intervention areas were identified, each having specific types of interventions represented by available studies.58,59 The Task Force findings are summarized below: Informational approaches Point-of-decision prompts: recommended Community-wide campaigns: recommended Mass media: insufficient evidence Behavioral and social approaches School-based PE: strongly recommended College-based health education and PE: insufficient evidence Classroom-based health education focused on reducing television viewing and video game playing: insufficient evidence Family-based social support: insufficient evidence Social support interventions in community settings: strongly recommended Individually adapted health behavior change programs: strongly recommended Environmental and policy approaches Creation of or enhanced access to places for physical activity combined with informational outreach activities: strongly recommended The Task Force viewed those interventions rated “recommended” or “strongly recommended” as useful to address related Healthy People 2010 objectives; those rated “insufficient evidence” were not to be considered ineffective, but required further research to be adequately evaluated. Regarding children’s physical activity and nutrition, interventions in several settings—schools, health care, and community—were reviewed in a dedicated issue of Preventive Medicine.8 Support was given to development and implementation or refinement of programs for schools and healthcare settings, and more generally to the idea that community-based programs were feasible and could be effective. Linkage between community-wide and school programs was needed to provide maximum impact of interventions. Theoretical foundations suggested great potential for community programs, but research was needed to add to a very limited evidence base. One major initiative in reaching children—specifically “tweens,” ages 9–13 years—was the VERB™ Campaign, designed to market physical activity to this target audience and described in detail in a sup-
46894_CH09_191_222.qxd
2/25/10
3:31 PM
Page 215
PREVENTION AND CONTROL 215
60
Intervention
Physical Activity, % of Lesson (Adjusted Mean ⴞ 1 SE)
50 Control
40 Moderate-to-Vigorous
30
Intervention 20 Control Vigorous 10
0 1
2
4
3
5
6
Semester Note: Observed at six time points, 1991 through 1994. The CATCH intervention was introduced during semester 2, increased the percentage of time spent in moderateto-vigorous and vigorous activity as measured by the System for Observing Fitness Instruction Time classroom observation system. Intervention and control curves diverged significantly according to repeated-measures analysis of variance with the class session as the unit of analysis: for moderate-to-vigorous activity, F 2.17, df 5, 1979, P .02; for vigorous activity, F 2.95, df 5, 1979, P .04. Analysis controlled for CATCH site, the location of the lesson, the specialty of the teacher, and random variation among the schools and weeks of observation. SE, standard error.
Figure 9-5 Moderate-to-Vigorous and Vigorous Physical Activity Observed During the Child and Adolescent Trail for Cardiovascular Health (CATCH) Physical Education Classes. Source: Reprinted with permission from RV Luepker et al, Outcomes of a Field Trial to Improve Children’s Dietary Patterns and Physical Activity: The Child and Adolescent Trial for Cardiovascular Health (CATCH), Journal of the American Medical Association, Vol 275, p 772, Copyright 1996, American Medical Association.
plemental issue of American Journal of Preventive Medicine.60 Supported by a special appropriation from Congress to the Centers for Disease Control and Prevention (CDC), the approach was both innovative and highly successful. Components of the program were described as including general marketing, and ethnic-specific advertising through radio, television, and print media, and promotions in communities,
schools and on the Internet.61 The program was implemented from June 2002–June 2004. An interview survey, the Youth Media Campaign Longitudinal Survey, was conducted annually and found a dose– response effect in relation to VERB™ exposure, with proportionate increases in awareness and positive attitudes about physical activity. Awareness was strongly associated with amount of activity reported.
46894_CH09_191_222.qxd
216
2/25/10
3:31 PM
Page 216
CHAPTER 9 PHYSICAL INACTIVITY
It appeared that mass media approaches had moved beyond the limited evidence base evaluated a decade earlier, from insufficient evidence to the current question of implementation and sustainability.62 Requirements for effective and sustainable programs are actively discussed. Some observers have judged the requirements to include “profound structural and environmental changes” such as “safer and better-equipped urban environments . . . to encourage individuals to be active in their everyday lives.”63 An often-cited opportunity is increasing accessibility to public parks, “critical resources of physical activity in minority communities . . . residential proximity is strongly associated with physical activity and park use . . .”64 A major metropolitan area has assessed how objectively measured levels of physical activity relate to the physical environment around the home, as determined by objective measures of land-use mix, residential density, intersection density, and a derived “walkability index.”65 Strategies for Metropolitan Atlanta’s Regional Transportation and Air Quality (SMARTRAQ) sampled regional residents and determined that the top-quartile group practiced more than two times the physical activity of those in the lowest quartile and were more than two times as likely to meet the recommended 30 minutes or more per day of moderate physical activity. The index was significantly correlated with activity levels, after adjustment for demographic factors (age, gender, education, and ethnicity). Others identified as a fundamental issue the “virtual absence of a public health practice infrastructure for the promotion of physical activity at the local level . . .”:66, pp 68,72 Physical activity promotion constitutes a critical role for public health practice, given the increasing prevalence of inactivity and sedentary behavior, the substantial protection against obesity and chronic disease conferred by regular physical activity, the major contribution of sedentariness and obesity to health disparities, and the increasing understanding of the central role that physical activity plays in overall health and quality of life. The public health infrastructure for physical activity promotion, while undeveloped and untested, is not unlike the public health infrastructure for other major health concerns before they were recognized as such. Given the evidence, the time is right to move forward with putting the infrastructure into place. To not do so is to place future generations at grave risk.
Global Strategies International public policy development in the area of physical activity has been reviewed recently by Blair and colleagues.67 Nearly 20 conference or workshop reports of the late 1980s and early- to mid-1990s were excerpted. They indicated widening attention to the need for public policy concerning physical activity in order to achieve public health impact. Subsequently, several reports illustrate further awakening to the issue of physical inactivity, with attention to action needed at the level of government, the whole of society, and regional or global authorities. The WHO Regional Office for Europe issued two companion reports, Physical Activity and Health in Europe: Evidence for Action and Promoting Physical Activity and Active Living in Urban Environments: The Role of Local Governments.68,69 The latter report concluded that:69, p 40 The problems of diminished physical activity and rising obesity need to be addressed urgently, and cities have an important part to play. To make health policies more robust, governments also need to support further research that quantifies the causal links between physical activity, health and changes in the built and social environments as well as evaluations of local policies and programmes that address these issues. It was noted, further, that “Local strategies and plans should aim at promoting physical activity among people of all ages, in all social circumstances and living in different parts of cities, with special attention to equity, deprivation and vulnerability.”69, p x For assessment of the state of physical activity recommendations in Latin America, a systematic review of publications and other reports was undertaken by use of a modified Community Guide to Preventive Services approach.70 More than 1000 reports were identified, of which 19 (16 published, 3 Brazilian theses) provided a basis for full evaluation. Several areas of intervention were represented among the reports, with school-based physical education programs alone supported by strong evidence for effectiveness. Implementation and maintenance of these programs and policies were recommended. The Disease Control Priorities in Developing Countries Project identified physical activity as convincingly associated with reduction of cardiovascular diseases, type 2 diabetes, cancer, fracture, obesity, and metabolic syndrome, and probably with reduction of depression and sexual dysfunction, all in addition
46894_CH09_191_222.qxd
3/5/10
2:32 PM
Page 217
REFERENCES 217
to maintenance of healthy weight.71 Interventions that could be applied were categorized as educational (in schools, worksites, or healthcare settings), transportation policy and environmental design (reducing automobile use, increasing walking and biking, development or redevelopment of cities and suburbs), initiatives at the community level, and economic policies. It was found that cost-effectiveness studies relating to physical activity interventions were unavailable from developing countries, so comparative evaluation of intervention options in this area was not feasible. Surveillance, demonstration programs, and costeffectiveness evaluations of lifestyle interventions in developing countries were recommended. The WHO Global Strategy on Diet, Physical Activity and Health was adopted by the 57th World Health Assembly (WHA), 2004.72 Resolution WHA 57.17 urges, among its several provisions, that Member States promote individual and community health through healthy diets and physical activity; promote lifestyles that include a healthy diet and physical activity; and encourage mobilization and engagement of “all concerned social and economic groups, including scientific, professional, nongovernmental, voluntary, private-sector, civil society, and industry associations . . .”72, pp 16–17 The WHO Director-General is requested, among other provisions, to provide technical advice to Member States, and to mobilize support for them at both global and regional levels, when requested, in implementing the Strategy and monitoring and evaluating implementation.
CURRENT ISSUES The leading issue in the area of prevention policy concerning physical activity is not about its development but its implementation. One obstacle is some continuing debate over recommendations for moderate, cumulative activity in contrast to more intensive, time-concentrated activity. In “How Much Pain for Cardiac Gain?” the journal Science highlighted the disagreement between those who interpret the data (e.g., Figure 9-4) as supporting one or the other of these alternatives.73 As Science noted, the reception to the recommendation in the 1995 consensus report12 was: “For a nation of couch potatoes, the news seemed too good to be true.”30, p 1324 At issue in part is a broader debate about public health recommendations that are intentionally moderate in character. Their aim is to achieve wider, if more gradual, public adoption than would be ex-
pected for more extreme measures, which might easily be ignored as impracticable or too drastic for individual adoption. An argument in support of moderate amounts and intensities of physical activity was presented by Blair and Connelly, who noted that the optimum prescription could not be written on current evidence (in 1996) but that “a major public health objective must be to mobilize the most sedentary 20 to 30 percent of the adult population,” perhaps with a strategy of making gradual changes over time.74, p 203 Evaluation of programs in varied implementation settings is warranted to determine whether weight stabilization is achieved at the moderate level of intensity or, if discussed above, vigorous activity is required for this critical outcome to be attained. Implementation will depend on social and political responses to the science that has been presented. Put simply by Morris, “Exercise is today’s best buy in public health, not only because of the need and potential, but because it is positive and acceptable, has insignificant side-effects, and can be inexpensive.”3, p 252 At the same time he noted, “The return of physical activity as the norm in everyone’s everyday life—the ‘restoration of biological normality’ in Rose’s words—will require cultural change on a scale similar to that which has occurred with smoking.”3, p 253 In addition, an extensive research agenda has been identified in each of the major reports cited here, including information needed from the most elementary prevalence surveys of physical activity in diverse population groups to demonstrations of effective means of mass change in physical activity behavior among both children and adults. Evaluation is needed of long-term effectiveness of such interventions in terms of maintenance of desirable patterns of activity and benefit, as measured by changes in incidence and prevalence of sedentary behavior and concurrent changes in risks and population rates of cardiovascular and other chronic diseases.
REFERENCES 1. Blackburn H. Physical activity and coronary heart disease: a brief update and population view (Part I). J Card Rehabil. 1983;3:101–111. 2. US Department of Health and Human Services. Physical Activity and Health: A Report of the Surgeon General. Atlanta, GA: US Department
46894_CH09_191_222.qxd
218
2/25/10
3:31 PM
Page 218
CHAPTER 9 PHYSICAL INACTIVITY
of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion; 1996. 3. Morris JN. Exercise versus heart attack: history of a hypothesis. In: Marmot M, Elliott P, eds. Coronary Heart Disease Epidemiology: From Aetiology to Public Health. Oxford (England): Oxford Medical Publications; 1992:242–255. 4. Taylor HL, Jacobs DR, Jr, Shucker B, et al. A questionnaire for the assessment of leisure-time physical activities. J Chronic Dis. 1978;31: 741–755.
(Switzerland): World Health Organization; 2004. 12. Pate RR, Pratt M, Blair SN, et al. Physical activity and public health. JAMA. 1995;273: 402–407. 13. Erlichman J, Kerbey AL, James WPT. Physical activity and its impact on health outcomes. Paper 1: the impact of physical activity on cardiovascular disease and all-cause mortality: an historical perspective. Obesity Rev. 2002;3: 257–271.
5. Paffenbarger RS, Hyde RT. Exercise in the prevention of coronary heart disease. Prev Med. 1984;13:3–22.
14. Blair SN, LaMonte MJ, Nichaman MZ. The evolution of physical activity recommendations: how much is enough? Am J Clin Nutr. 2004; 79(suppl):913S–920S.
6. Durstine JL, King AC, Painter PL, et al., eds. ACSM’s Resource Manual for Guidelines for Exercise Testing and Prescription. 2nd ed. Philadelphia, PA: American College of Sports Medicine; Lea & Febiger; 1993.
15. Orleans CT, Leviton LC, Thomas KA, et al. History of the Robert Wood Johnson Foundation’s Active Living Research Program. Origins and strategy. Am J Prev Med. 2009; 36(2S):S1–S9.
7. Haskell WL. Addendum to chapter 19. In: Marmot M, Elliott P, eds. Coronary Heart Disease Epidemiology: From Aetiology to Public Health. 2nd ed. Oxford (England): Oxford University Press; 2005:286–290.
16. Sallis JF, Orleans CT, Buchner DM. Active living research. A six-year report. Am J Prev Med. 2009;36(2S):S1–S77.
8. Mendlein J, Baranowski T, Pratt M. Physical activity and nutrition in children and youth: opportunities for performing assessments and conducting interventions. Prev Med. 2000;31: S150–S153. 9. American Cancer Society, American Diabetes Association, American Heart Association. Physical education in schools—both quality and quantity are important. http://www.american heart.org/downloadable/heart/1204662840069 Policy%20Statement%20on%20Physical%20 Education%20in%20Schools.pdf. 10. US Department of Health and Human Services. 2008 Physical Activity Guidelines for Americans. http://www.health.gov/paguidelines. Accessed April 1, 2009. 11. World Health Organization. Global Strategy on Diet, Physical Activity and Health. Geneva
17. National Center for Health Statistics. Assessing Physical Fitness and Physical Activity in Population-Based Surveys. DHHS publication no. (PHS) 89-1253. Hyattsville, MD: National Center for Health Statistics, Centers for Disease Control, Public Health Service, US Department of Health and Human Services; 1989. 18. Pereira MA, FitzGerald SJ, Gregg EW, et al., eds. A collection of physical activity questionnaires for health-related research. Med Sci Sports Exercise. 1997;29(suppl):S1–S205. 19. Lamb KL, Brodie DA. The assessment of physical activity by leisure-time physical activity questionnaires. Sports Med. 1990;10:159–180. 20. World Health Organization. The World Health Report 2002. Reducing Risks, Promoting Healthy Life. Geneva (Switzerland): World Health Organization; 2002.
46894_CH09_191_222.qxd
2/25/10
3:31 PM
Page 219
REFERENCES 219
21. French SA, Story M, Jeffrey RW. Environmental influences on eating and physical activity. Annu Rev Public Health. 2001;22:309–335. 22. Brownson RC, Boehmer TK, Luke DA. Declining rates of physical activity in the United States: what are the contributors? Annu Rev Public Health. 2005;26:421–443. 23. Sherwood NE, Jeffrey RW. The behavioral determinants of exercise: implications for physical activity interventions. Annu Rev Public Health. 2000;20:21–44. 24. Speck BJ, Harrell JS. Maintaining regular physical activity in women. Evidence to date. J Cardiovasc Nurs. 2003;18:282–291. 25. Committee on Environmental Health. The built environment: designing communities to promote physical activity in children. Pediatrics. 2009;123:1591–1598. 26. Kelder SH, Perry CL, Klepp K-I, Lytle LL. Longitudinal tracking of adolescent smoking, physical activity, and food choice behaviors. Am J Public Health. 1994;84:1121–1126. 27. Haskell WL. Sedentary lifestyle as a risk factor for coronary heart disease. In: Pearson TA, Criqui MH, Luepker RV, Oberman A, Winston M, eds. Primer in Preventive Cardiology. Dallas, TX: American Heart Association; 1994:173–187. 28. Bassuk SS, Manson JE. Physical activity and the prevention of cardiovascular disease. Current Athero Rep. 2003;5:299–307. 29. Young DR, Steinhardt MA. The importance of physical fitness versus physical activity for coronary artery disease risk factors: a crosssectional analysis. Res Q Exercise Sport. 1993;B4:377–384. 30. US Department of Health and Human Services. Health, United States, 2008 with Special Feature on the Health of Young Adults. Washington, DC: US Department of Health and Human Services; Centers for Disease Control and Prevention. National Center for Health Statistics; 2008.
31. Centers for Disease Control and Prevention. Physical activity levels among children aged 9–13 years––United States, 2002. MMWR 2003;52(33):785–788. 32. US Department of Health and Human Services. Healthy People 2010 Midcourse Review. Washington, DC: US Government Printing Office; December 2006. 33. Guthold R, Ono T, Strong KL, Chatterji S, Morabia A. Worldwide variability in physical inactivity. A 51-country survey. Am J Prev Med. 2008;34:486–494. 34. Keys A. Seven Countries: A Multivariate Analysis of Death and Coronary Heart Disease. Cambridge, MA: Harvard University Press; 1980. 35. Powell KE, Thompson PD, Caspersen CJ, Kendrik JS. Physical activity and the incidence of coronary heart disease. Annu Rev Public Health. 1987;8:253–287. 36. Berlin JA, Colditz GA. A meta-analysis of physical activity in the prevention of coronary heart disease. Am J Epidemiol. 1990;132: 612–628. 37. Leon AS, Connett J, MRFIT Research Group. Physical activity and 10.5 year mortality in the Multiple Risk Factor Intervention Trial (MRFIT). Int J Epidemiol. 1991;20: 690–697. 38. Blair SN, Kampert JB, Kohl HW, et al. Influences of cardiorespiratory fitness and other precursors on cardiovascular disease and all-cause mortality in men and women. JAMA. 1996;276:205–210. 39. Richardson CR, Kriska AM, Lantz PM, Hayward RA. Physical activity and mortality across cardiovascular disease risk groups. Med Sci Sports Exerc. 2004;36:1923–1929. 40. Wannamethee SG, Shaper AG. Physical activity and cardiovascular disease. Sem Vasc Med. 2002;2:257–265. 41. Batty GD. Physical activity and coronary heart disease in older adults. A systematic review of
46894_CH09_191_222.qxd
220
2/25/10
3:31 PM
Page 220
CHAPTER 9 PHYSICAL INACTIVITY
epidemiological studies. Eur J Public Health. 2002;12:171–176. 42. Levin S, Lowry R, Brown DR, Dietz WH. Physical activity and body mass index among US adolescents. Youth Risk Behavior Survey, 1999. Arch Pediatr Adolesc Med. 2003;157: 816–820. 43. Stevens J, Murray DM, Baggett CD, et al. Objectively assessed associations between physical activity and body composition in middle-school girls. The Trial of Activity for Adolescent Girls. Am J Epid. 2007;166: 1298–1305. 44. Kohl HW, Powell KE, Gordon NF, et al. Physical activity, physical fitness, and sudden cardiac death. Epidemiol Rev. 1992;14:37–58. 45. Mittleman MA. Triggering of myocardial infarction by physical activity, emotional stress and sexual activity. In: Willich SN, Muller JE, eds. Triggering of Acute Coronary Syndromes. Dordrecht (the Netherlands): Kluwer Academic Publishers; 1996:71–80. 46. Danaei G, Ding EL, Mozaffarian D, et al. The preventable causes of death in the United States: comparative risk assessment of dietary, lifestyle, and metabolic factors. PLoS Med. 6(4): e1000058. doi:10.1371/journal.pmed. 1000058. 47. Ezzati M, Vander Hoorn S, Lopez AD, et al. Comparative quantitation of mortality and burden of disease attributable to selected risk factors. In: AD Lopez, CD Mathers, M Ezzati, DT Jamison, CJL Murray, eds. Global Burden of Disease and Risk Factors. Washington, DC: The International Bank for Reconstruction and Development/The World Bank; 2006: 241–396. 48. White PD, Wright IS, Sprague HB, et al. A Statement on Arteriosclerosis: Main Cause of “Heart Attacks” and “Strokes.” New York: National Health Education Committee, Inc; 1959. 49. Bauman AE. Updating the evidence that physical activity is good for health: an epidemiological review 2000–2003. J Sci Med Sport. 2004; 7(suppl):6–19.
50. Erlichman J, Kerbey AL, James WPT. Physical activity and its impact on health outcomes. Paper 2: prevention of unhealthy weight gain and obesity by physical activity: an analysis of the evidence. Obesity Rev. 2002;3:273–287. 51. Glassberg H, Balady GJ. Exercise and heart disease in women. Why, how, and how much? Cardiol Rev. 1999;7:301–308. 52. Fletcher GF. How to implement physical activity in primary and secondary prevention. A statement for healthcare professionals from the Task Force on Risk Reduction, American Heart Association. Circulation. 1997;96: 355–357. 53. Jolliffe JA, Rees K, Taylor RS, Thompson D, Oldridge N, Ebrahim S. Exercise-based rehabilitation for coronary heart disease. Cochrane Database Syst Rev. 2001;(1):CD001800. doi: 10.1002/14651858.CD001800. 54. O’Connor GT, Buring JE, Yusuf S, et al. An overview of randomized trials of rehabilitation with exercise after myocardial infarction. Circulation. 1989;80:234–244. 55. Hillsdon M, Foster C, Thorogood M. Interventions for promoting physical activity. Cochrane Database Syst Rev. 2005;(1): CD003180. doi:10.1002/14651858. CD003180.pub2. 56. US Department of Health and Human Services. Agency for Healthcare Research and Quality. US Preventive Services Task Force. The Guide to Clinical Preventive Services 2006. Recommendations of the U.S. Preventive Services Task Force. AHRQ Publication No. 06-0588. Washington, DC: Agency for Healthcare Research and Quality; 2006. http://www.ahrq .gov/clinic/uspstf/uspstbac.htm. Accessed October 14, 2007. 57. Haskell WL, Lee I-M, Pate RR, et al. Physical activity and public health: updated recommendation for adults from the American College of Sports Medicine and the American Heart Association. Med Sci Sports Exerc. 2007;39: 1423–1434. 58. Task Force on Community Preventive Services. Recommendations to increase physical activity
46894_CH09_191_222.qxd
2/25/10
3:31 PM
Page 221
REFERENCES 221
in communities. Am J Prev Med. 2002;22(4S): 67–72. 59. Kahn EB, Ramsey LT Brownson RC, et al. The effectiveness of interventions to increase physical activity. A systematic review. Am J Prev Med. 2002;22(4S):73–107. 60. Wong FL, Huhman M, Berkowitz JM, Cavill N, Maibach EW. The VERBJ Campaign. Not About Health, All About Fun: Marketing Physical Activity to Children. Am J Prev Med. 2008;34(6S):S171–S279. 61. Huhman ME, Potter LD, Duke JC, Judkins DR, Heitzler CD, Wong FL. Evaluation of a national physical activity intervention for children. VERBJ Campaign, 2002–2004. Am J Prev Med. 2007;32:38–43.
67. Blair SN, Booth M, Gyarfas I, Iwane H, et al. Development of public policy and physical activity initiatives internationally. Sports Med. 1996;21:157–163. 68. Physical Activity for Health in Europe: Evidence for Action. Copenhagen, WHO Regional Office for Europe; 2006. 69. Promoting Physical Activity and Active Living in Urban Environments. The Role of Local Governments. Copenhagen, WHO Regional Office for Europe; 2006. 70. Hoehner CM, Soares J, Parra Perez D, et al. Physical activity interventions in Latin America. A systematic review. Am J Prev Med. 2008;34:224–233.
63. Costanza MC, Beer-Borst S, Morabia A. Achieving energy balance at the population level through increases in physical activity. Am J Public Health. 2007;97:520–525.
71. Willett WC, Koplan JP, Nugent R, Dusenbury C, Puska P, Gaziano T. Prevention of chronic disease by means of diet and lifestyle changes. In: Jamison DT, Breman JG, Measham AR, et al., eds. Disease Control Priorities in Developing Countries. 2nd ed. Washington, DC: International Bank for Reconstruction and Development/The World Bank; 2006:833–850.
64. Cohen DA, McKenzie TL, Sehgal A, Williamson S, Golinelli D, Lurie N. Contribution of public parks to physical activity. Am J Public Health. 2007;97:509–514.
72. World Health Organization. Global Strategy on Diet, Physical Activity and Health. Geneva (Switzerland): World Health Organization, 2004.
65. Frank LD, Schmid TL, Sallis JF, Chapman J, Saelens BE. Linking objectively measured physical activity with objectively measured urban form. Findings from SMARTRAQ. Am J Prev Med. 2005;28(2S2):117–125.
73. Barinaga M. How much pain for cardiac gain? Science. 1997;276:1324–1327.
62. Banspach SW. The VERBJ Campaign. Am J Prev Med. 2008;34(6S):S275.
66. Yancey AK, Fielding JE, Flores GR, Sallis JF, McCarthy WJ, Breslow L. Creating a robust public health infrastructure for physical activity promotion. Am J Prev Med. 2007;32: 68–78.
74. Blair SN, Connelly JC. How much physical activity should we do? The case for moderate amounts and intensities of physical activity. Res Q Exercise Sport. 1996;67:193–205.
46894_CH09_191_222.qxd
2/25/10
3:31 PM
Page 222
46894_CH10_223_268.qxd
2/25/10
3:31 PM
Page 223
C H A P T E R
10 Obesity described as to their prevalence in the population, their relation to risk of cardiovascular and other chronic diseases, and their responsiveness to interventions for prevention or reduction of adverse levels of BMI. The chief proximal determinant of BMI is imbalance between energy intake and expenditure. The factors that most directly influence these determinants are dietary imbalance and physical inactivity, reviewed in the preceding chapters. Mechanisms by which obesity contributes to cardiovascular and other diseases involve its strong association with other established major risk factors, especially adverse blood lipid profile, high blood pressure, and diabetes. Early studies were mixed in their findings regarding the relation between obesity and, for example, coronary heart disease. It has become clear more recently that analyses taking smoking into account (which is associated with lower BMI), recognizing the pathways of influence of BMI through other risk factors, and based on longer-term follow-up consistently indicate a significant positive relation between obesity and coronary heart disease, as well as other cardiovascular conditions. Prevention of childhood obesity is the primary public health approach to control of the epidemic. Control, usually meaning reduction of already excessive BMI, concerns both children and adults. Strong evidence is generally lacking for specific interventions for achieving prevention and control of obesity. Key strategies—whether individual, community or population-wide, or global in focus—depend foremost on overcoming dietary imbalance and physical inactivity. Implementation of recommendations based on present knowledge, and evaluation of all efforts to achieve prevention and control of obesity, are the dominant current issues in this arena.
SUMMARY Obesity, when not specifically defined, may refer to any of several characteristics of body size, body composition, or appearance. These include overweight, fatness, and fat distribution such as central, truncal, or abdominal obesity. It is thought that physiologic and metabolic mechanisms that function to promote fat storage were once, in preagricultural society, adaptive to needs for energy reserves to ensure successful reproduction and survival. In modern civilizations in which energy consumption generally exceeds energy expenditure, fat storage is no longer adaptive but detrimental. Obesity has come to be defined as excess weight, specifically due to fat mass, with adverse health consequences. Because the prevalence of obesity has increased sharply in recent decades in many populations, in both children and adults, it is widely recognized as a significant public health problem. This is reflected in numerous recent reports from national and international organizations and agencies, with calls for action to prevent and control this epidemic. Given the several dimensions of obesity, various measurements have been advocated to define it. Generally accepted currently is a ratio of body weight to height, calculated as weight in kilograms divided by height in meters, squared: Body Mass Index (BMI) wt(kg)/ht(m)2. Categories of BMI define underweight, normal weight, overweight, and obesity of increasing degrees. High values of BMI do not necessarily represent fatness but could be greater than average lean mass, a limitation leading some authorities to recommend one or another additional measure to gauge fat distribution. So defined, overweight and obesity can be
223
46894_CH10_223_268.qxd
224
2/25/10
3:31 PM
Page 224
CHAPTER 10 OBESITY
INTRODUCTION When in 1985 the National Institutes of Health (NIH) convened an expert panel to develop a consensus report on the health implications of obesity, the first question addressed was, “What is obesity?”1 The panel defined it as “an excess of body fat frequently resulting in a significant impairment of health.”1, p 1073 Noting the impracticality of actually determining the quantity of fat in the body except by techniques limited to research laboratories, the panel recommended use of anthropometric methods based on measurement of height, weight, and skinfold thickness. The measurement of height and weight could be compared with reference values (the Metropolitan Life Insurance Company tables of “desirable weight”) or expressed as the “body mass index,” calculated as the ratio of weight (in kilograms) to the square of height (in meters). The location of body fat was also emphasized, as assessed by the ratio of body circumferences measured at the waist and hip. Higher values of the waisthip ratio (WHR) represent increased abdominal or central, in contrast to peripheral, fat accumulation. The term obesity thus represented several concepts: fatness, in terms of tissue composition of the body; overweight, relative to either reference standards or individual body height; and central adiposity, or fat distribution with a relative excess of abdominal girth and underlying abdominal fat mass. The panel concluded that weight at 20% or more above desirable levels was a degree of obesity constituting “an established health hazard” and that 34 million Americans had body mass index values indicative of this degree of overweight and were in need of treatment. Obesity, like patterns of diet and physical activity, can be viewed in the perspective of human evolution.2 A detailed historical account is provided by Bray.3 In simplest outline, the genesis of the epidemic occurrence of obesity is generally understood as follows: Obesity, like the contemporary diet in which animal fat has displaced fiber, and like the nonphysical pursuits that have largely replaced physical work in many societies, is mainly a modern phenomenon considered not to have been widespread in ordinary circumstances throughout most of human history. Success in reproduction and survival depended on adaptation to several fundamental conditions: the episodic nutritional demands of pregnancy and lactation; seasonal variation in food availability; and periodic famine sometimes occurring in cycles as short as two or three years. These conditions are considered to be determinants of both genetic and cultural evolution through which humans developed mechanisms
for storing fat, an efficient and readily mobilized energy reserve. Obesity may not have been unknown, as Bray presents evidence of awareness of obesity in numerous Paleolithic “Venus figures” found throughout Europe and the Middle East. Current theory would require that this condition remained rare, however, until very recently. Under the highly prevalent conditions of modern dietary imbalance and physical inactivity, these once-adaptive mechanisms now commonly produce the adverse effect of excess fat accumulation. Links between obesity and cardiovascular diseases have been established through epidemiologic, clinical, and laboratory research, as documented in numerous reviews including several edited collections. Three examples are noted here, of which two are specific to cardiovascular diseases and one (Bray et al.) is more comprehensive.4–6 Coupled with understanding of the health consequences of obesity, growing recognition of the magnitude of the obesity epidemic in the United States and globally has greatly elevated public concern about this problem and engaged significant attention by many health-related organizations and agencies. Several of the many recent reports illustrate the point: The Surgeon General’s Call to Action to Prevent and Decrease Overweight and Obesity 2001;7 the American Heart Association’s Prevention Conference VII: Obesity, a Worldwide Epidemic Related to Heart Disease and Stroke8 and the American Medical Association’s National Summit on Obesity: Building a Plan to Reduce Obesity in America,9 both in 2004; the Institute of Medicine’s Preventing Childhood Obesity: Health in the Balance,10 in 2005; and the American Academy of Pediatrics’ Expert Committee Recommendations Regarding the Prevention, Assessment, and Treatment of Child and Adolescent Overweight and Obesity,11 published in 2007. The global dimension is represented by reports from the International Obesity Task Force, Obesity Prevention: The Case for Action,12 in 2002, and the WHO Global Strategy on Diet, Physical Activity and Health,13 in 2004. Obesity, then, has become a prominent aspect of the epidemiology and prevention of cardiovascular diseases.
CONCEPTS AND DEFINITIONS Obesity In general use, “obesity” refers to an appearance of bodily overweight or excessive abdominal girth of varying degree. The substance of the body mass is taken into account as well, given that an evidently
46894_CH10_223_268.qxd
2/25/10
3:31 PM
Page 225
CONCEPTS AND DEFINITIONS 225
muscular person, or a pregnant female who is otherwise lean, would not be considered “obese.” Definition for scientific and clinical purposes is necessary, but options have long been debated. The NIH panel cited above identified excessive body fat and impairment of health as defining characteristics of obesity. Two features of this definition are problematic. First, determining the amount of body fat relative to the optimum amount implies availability of practical methods for determining fat content of the body and appropriate reference standards. Second, incorporation of impairment of health in the definition begs the question of whether obesity of any degree is unhealthy. Bray and others, in a chapter devoted to definition and classification of obesity, stated:14, p 31 “Excessive body mass for stature, and more specifically an excessive body fat content, is a condition of concern because it is in itself socially and physically debilitating and it represents a risk factor for increased morbidity and mortality rate.” Further, “three main adipose tissue features are of particular importance from a health perspective”:14, p 32 First, increased morbidity and higher mortality rates are seen in those with an excessive proportion of body fat or a high body mass relative to stature. Second, the risk profile tends to be more dangerous when the excess fat is mainly stored on the upper body and less on the buttocks and lower limbs, i.e., when fat topography is typically malelike. Finally, recent research has suggested that the most atherogenic fat depot of the human body is within the abdominal cavity around the viscera, particularly the fat depots with small blood vessels draining into the portal vein carrying blood back to the liver. The amount of abdominal visceral fat appears to be critical in determining whether obesity is going to have major or minor health implications for a given individual. These three sets of characteristics—fatness or body mass relative to stature, male-like or android fat distribution, and size of the abdominal fat depot— were then discussed by Bray and others as “three types of human obesity.” They posed the question of whether obesity is a disease and answered, by analogy to high blood pressure or cholesterol, that none of these three risk factors is a disease in itself. Rather, continuously increasing levels confer increasing risks for particular disease outcomes. It is not increased fat, or obesity, itself but its effects on other systems that results in increased morbidity and mortality. Several concepts are germane to understanding obesity as described. These include not only fatness, overweight, and fat distribution but also vulnerable
periods for weight gain, obesity in children and adolescents, and “tracking.” “Tracking” is the tendency for an individual to stay over time in a similar rank among age peers within the group distribution of, for example, weight or height. This is most relevant for predicting future health from observations in childhood or adolescence. Last, the concept of healthy weight is noteworthy as well. Fatness If fatness means the quantity of adipose tissue in the body, this may be expressed either in absolute units (fat mass, in kg) or as the corresponding percentage of the total body mass that is adipose tissue (percent body fat, without units). Several methods for determination of body composition permit estimation of fatness and are reviewed in Human Body Composition.15 Field methods applicable in population studies depend on anthropometry, alone or in combination with determination of bioelectrical impedance. Prediction equations relate these measurements to fat mass as estimated from laboratory investigation by other techniques. No single method is considered as a “gold standard,” and published prediction equations cannot be assumed to be valid for every population. Percent body fat changes markedly during infancy and childhood. For both boys and girls, the change is from about 14 percent at birth to 26 percent at age 6 months. Percent body fat decreases to age 5, when it reaches 15 percent for boys and 17 percent for girls.16 From ages 10 to 18 years, percent body fat changes little in girls but decreases by about 1 percent per year for boys. Patterns of growth in tissue mass differ by sex, with girls adding proportionately equal amounts of fat and fat-free mass, whereas with fat-free mass dominating growth in boys. During adulthood, percent body fat appears to increase slowly in both men and women, perhaps stabilizing at about age 50 at 25 percent for men and 35 percent for women. Overweight Overweight connotes a relative excess of weight. A long-standing reference standard for overweight has been actuarial tables published by the Metropolitan Life Insurance Company on the relation between weight, height, and longevity among insured persons.17 For many years, the “desirable weight” tables based on the 1959 Society of Actuaries data were used. Weights corresponding to the Metropolitan relative weight, or MRW, of 120 are 20 percent above the observed optimum for longevity in the actuarial analyses. This is the criterion for “obese” specified by the NIH consensus panel.
46894_CH10_223_268.qxd
226
2/25/10
3:31 PM
Page 226
CHAPTER 10 OBESITY
includes the mesenteric and omental fat depots, which differ in metabolic characteristics from peripheral fat stores.21 Several terms are commonly used in reference to excess visceral fat: central adiposity, abdominal obesity, and truncal obesity. The Obesity Education Initiative in the United States presents a classification of overweight and obesity based on BMI categories as discussed above and adds corresponding categorical levels of disease risk (for type 2 diabetes, hypertension, and cardiovascular diseases) in relation to abdominal circumference (Table 10-1).22 By definition, increased risk is considered to begin at a BMI of 25.0 kg/m2; it rises progressively with increasing BMI. Critical values of abdominal circumference to mark increased risk differ between men and women to account for differences in average body size by sex. For both men and women, a criterion value is indicated to distinguish greater from lesser disease risk at a given BMI, except that beyond class I obesity, the categories of abdominal circumference do not contribute further to differentiating risk levels.
Another approach assessing overweight is use of a ratio of weight to height. Most common among these is the body mass index (BMI), also known as Quetelet’s index, after its originator in the mid-19th century. BMI is calculated as the ratio of weight in kilograms to the square of height in meters: BMI wt(kg)/ht(m)2 or wt(lbs)/ht(in)2 703). The population distribution of BMI has been categorized by a WHO Study Group as shown in Figure 10-1:18 “Normal” BMI is from 18.5 to 25 kg/m2; grades 1–3 of obesity correspond to values of 25–30, 30–40, and 40 kg/m2 or greater, respectively. A subsequent WHO report designated these categories as “normal” and “grade 1, grade 2, and grade 3 overweight.”19 It was noted that to achieve a population distribution of BMI with only rare occurrence of values of 25 kg/m2 or greater would entail reaching a population mean of 22 kg/m2 Fat Distribution The third concept of obesity concerns the anatomic distribution of body fat, one aspect of body composition. Methods of measurement for total and regional body composition are described in detail by Heymsfield and others.20 Some laboratory techniques provide direct measures of tissue composition in the abdomen and other sites. Field methods are usually limited to anthropometry, by which various anatomic diameters, circumferences, and skinfold thicknesses can be recorded and indices of interest constructed and calculated. Abdominal circumference and the WHR, for example, have been of interest in epidemiologic studies as indirect measures of visceral adipose tissue. This
Vulnerable Periods The concept of “vulnerable,” “critical,” or “sensitive” periods during the life course relates to points or phases in development where influential factors may determine permanent changes, such as weight gain that will persist in later stages of life.23 Protective influences may operate at such points also. Daniels and others note that studies of obesity in adolescence have not usually included childhood measurements preceding the onset of obesity, so that factors in deter-
Chronic Energy Deficiency Grades
3
2
Obesity Grades
1
1
3
2
Normal
16 15
17.5 18.5 20
25
30
35
40
Body Mass Index
Note: Grades 1 and 2 require that energy expenditure be also below 1.4 times the estimated basal metabolic rate, based on the weight of the individuals. Body mass index = mass in kg/(height in meters) 2.
Figure 10-1 Degree of Chronic Energy Deficiency and Obesity in Relation to Body Mass Index. Source: Reprinted with permission from Report of a WHO Study Group, WHO TRS 797, p 71, © 1990, World Health Organization.
46894_CH10_223_268.qxd
2/25/10
3:31 PM
Page 227
CONCEPTS AND DEFINITIONS 227
Table 10-1
Underweight Normal Overweight Obesity Extreme Obesity
Classification of Overweight and Obesity by BMI, Waist Circumference, and Associated Disease Risks Disease Risk* Relative to Normal Weight and Waist Circumference† Men 102 cm (40 in) or Less Women 88 cm (35 in) BMI (kg/m2) Obesity Class Women 88 cm (35 in) or Less Men 102 cm (40 in) 18.5 — — 18.5–24.9 — — 25.0–29.9 Increased High 30.0–34.9 I High Very High 35.0–39.9 II Very High Very High 40.0† III Extremely High Extremely High
*Disease risk for type 2 diabetes, hypertension, and CVD. † Increased waist circumference can also be a marker for increased risk even in persons of normal weight. Source: Reprinted from National Heart, Lung and Blood Institute, Classification of overweight and obesity by BMI, waist circumference, and associated disease risks. Available at: www.nhlbi.nih.gov/health/public/heart/obesity/lose_wt/bmi_dis.htm. Accessed May 7, 2007.
mining incidence are not yet well understood. Knowledge of key turning points in development of body mass and composition would have great potential for prevention. A familiar example of such vulnerable periods is during fetal development when, it has been inferred, some aspects of metabolic and other regulatory processes become permanently “programmed” under the influence of more or less favorable maternal conditions.24 Obesity in Children and Adolescents A further aspect of definition concerns the contrast between adulthood and earlier periods of life. Infancy, childhood, and adolescence are periods of major change in body size and composition, as well as sexual differentiation, with marked individual differences in tempo or pace of development of these characteristics. Because the excess morbidity and mortality that define obesity in adulthood are not yet apparent, there is no equivalent criterion for identifying obesity in childhood and adolescence as a morbidity-related condition. Implications of these issues are discussed by Flegal, who noted as recently as 1993 that there was no generally accepted definition of obesity for this age group.25 Gidding and others wrote in 1996 that obesity was “defined as the presence of excess adipose tissue.”26, p 3384 They indicated that “any child with weight for height above the 75th percentile for age and sex or who has significantly increased his weight for height percentile and who suffers from a morbidity that would be worsened by obesity (e.g., dyslipidemia, diabetes mellitus, or hypertension) should be considered obese.” It was considered important to distinguish obese patients from those whose overweight is attributable to the lean body mass. To do
this in practice, use of subscapular skinfold measurement was recommended. The International Obesity Task Force (IOTF) convened a workshop to determine how best to assess the prevalence of obesity in children and adolescents around the world and published the proceedings in 1999.27 The workshop adopted BMI as the standard measure, with the absolute cutpoints of the BMI distribution used in adults shown in Figure 10-1. For purposes of international comparisons, criteria for overweight and obesity in children and adolescents agree with grade 1 and 2 obesity in adults, such that overweight begins at a BMI of 25 kg/m2 and obesity at 30 kg/m2. It was noted that further research would be of value to address the sensitivity and specificity of BMI values at the 85th and 95th percentiles (in relation to a specific reference population) to identify an increase in percentage of body fat. To relate these absolute criteria to an underlying reference distribution of BMI, Cole and others utilized survey data from Brazil, Great Britain, Hong Kong, the Netherlands, Singapore, and the United States.28 They presented the BMI values by sex and by half-years of age from 2 to 18 years that correspond to the same percentiles as BMI values of 25 and 30 kg/m2 at age 18 years. Flegal and others subsequently compared prevalence estimates of overweight and obesity in US children by use of the CDC-US growth charts, the international reference values described previously, and a third set. Population differences in the age–sex distributions of BMI result in somewhat different estimates of prevalence for specific strata of the population, particularly at younger ages in this comparison. However, prevailing practice as presented in current guidelines is to use BMI alone.11 Recently
46894_CH10_223_268.qxd
228
2/25/10
3:31 PM
Page 228
CHAPTER 10 OBESITY
established terminology refers to levels of BMI from the 85th to 94th percentile as “overweight” and those above the 95th percentile as “obesity.” The reference distributions are provided by the Centers for Disease Control and Prevention (CDC) and are based on national surveys from 1963 and 1980 for ages 6 to 19 years and from 1971 to 1974 for ages 2 to 5 years.10 Because the distribution of BMI in US children and adolescents has shifted upward since the time of these surveys, the prevalence of overweight and of obesity is substantially greater than 10 percent and 5 percent, respectively. Tracking Study of cardiovascular risk factors in children and adolescents has included attention to the predictive value at later ages of measurements recorded at earlier ages. If values associated with disease risk persist from childhood into young adulthood or beyond, or “track,” then early intervention gains importance. To evaluate the tracking of weight, BMI, or other weight-for-height indices, tracking correlations—the correlations between a first and subsequent value of the measure—were compiled from 22 studies.29 As shown in Figure 10-2, observations were first made from age 1–4 years to 15–20 years, and the values compared were from 1–3 years to 21–50 years apart. In general, the strongest tracking correlations were observed for first measures at later ages, and for shorter intervals. However, tracking from ages 9–14 was nearly as strong as from ages 15–20, suggesting that observations during puberty were more predictive than those made earlier. Tracking is the more important because of the association of obesity with other
risk factors, beginning in childhood and adolescence, including hyperinsulinemia/insulin resistance, dyslipidemia, and hypertension.30 Healthy Weight A countervailing viewpoint has been presented that would emphasize fat cell function rather than BMI or other indices of adiposity as a basis for determining the healthy weight for an individual.31 The underlying premise of the argument is that the metabolic and physiologic derangements associated with obesity are attributable to fat cell filling and loss of protective cell functions. A dichotomy within the BMI distribution identifies as obese some persons whose fat cell function remains normal (and other risk factors are absent), whereas for many considered nonobese, fat cell function is nevertheless impaired at lesser degrees of overweight (and other risk factors are present). Modest weight loss is known to improve these functions, even while overweight or obesity and excess fatness persist. Improvement in diet and physical activity may suffice to improve or restore health at a weight that would not be considered healthy when viewed in terms of BMI alone. Proponents of the “fat cell function hypothesis” support the use of statistically based standards for screening but contend that “A focus on good health practices is likely to yield better health results for most than would a focus on weight.”31, p 450S This view can be weighed against the full array of adverse effects of obesity and the value of public health approaches to prevention and control of its consequences. Refinements of evaluation and intervention at the individual level warrant consideration within the broader policy context. The complementary nature
1 Correlation Coefficients
1–3 yr apart
0.8 4–6 yr apart
0.6
11–15 yr apart
7–10 yr apart 16–20 yr apart
0.4
21–50 yr apart
0.2
0 124
528
9214
15220
Initial Age (yr)
Figure 10-2 Tracking Weight, Body Mass Index (BMI), or Weight-for-Height for Various Follow-Up Periods. Source: Reprinted with permission from E Obarzanek in Obesity: Impact on Cardiovascular Disease, GF Fletcher, SM Grundy, LL Hayman eds, E Obarzanek, p 35. © 1999 Futura Publishing Company, Inc.
46894_CH10_223_268.qxd
2/25/10
3:31 PM
Page 229
DETERMINANTS 229
of the population-wide and individual approaches to prevention implied here is discussed further in Chapter 18, “Strategies of Prevention.”
MEASUREMENT Anthropometry and Body Fat Distribution Methods of anthropometry have become well standardized. Reference works for these procedures are readily available, including a manual, by Lohman and others, that illustrates the techniques.32 Mueller and others are among investigators who have evaluated the utility of various anthropometric measures and indices of body fat distribution for epidemiologic studies.33,34 They took as the criterion of utility the correspondence of any one or combination of measures to levels of cardiovascular risk factors assessed in the same individuals. The preferred approach in adults was the ratio of waist-to-thigh circumferences, being somewhat less predictive of risk factors but more reliably measured. In children and adolescents, body fat distribution was uninformative until sexual maturity was attained, and then both skinfold measures and the waist-hip ratio added to BMI in strength of association with risk factors. The waist-hip ratio must be interpreted with some caution, as a high value may reflect deficient gluteal muscle mass at the hip level, rather than excess abdominal fat.21 Body Composition Methods for assessing body composition include densitometry, hydrometry, whole-body counting, neutron activation analysis, dual energy X-ray absorptiometry (DEXA), electrical impedance and total body electrical conductivity, multicomponent molecular level methods, and several imaging techniques. Several of these techniques incorporate anthropometry. Brown reviewed both the methods and the findings from their application, including their determinants and relation of various measures of body composition to health.15 Estimation of percent body fat (as well as fat distribution) by anthropometry, alone or with bioelectrical impedance, is the most widely applied approach to assessment of body composition in epidemiologic research. These specific methods are reviewed by Roche and by Baumgartner.35,36 Potentially, several kinds of anthropometric variables may be combined for this purpose, including lengths, breadths, circumferences, skinfold thicknesses, and others. With or without bioelectrical impedance, estimation of fat mass depends on
application of prediction equations in which the selected measurements have been evaluated statistically against laboratory-based procedures. Attention to several aspects of the measurement situation, including instruments, subject, standardized protocols, and validity of the prediction equation for the population studied are all required to obtain meaningful and interpretable results. Relative Weight Relative weight indices of obesity depend solely on measurement of height and weight, with sex taken into account. These simple measures are not without sources of error, however, and also require standardization of technique, especially for longitudinal studies or population comparisons. Relative weight tables from a standard population, described above, provide the basis in the United States for evaluating weight for height. The Metropolitan Tables in use since 1959 were updated in 1983 on the basis of more recent actuarial experience, with the result that “desirable weights” increased for every height category but the tallest, for both men and women. For the lower height categories, several pounds were added. Some controversy ensued about the desirability, from a public health perspective, of a change in desirable weights, but the revised recommendations stand.37 BMI has been discussed previously. It is also simply determined from measured weight and height. This is only one of a family of potential weight-forheight indices in which the power for height may take various values—for BMI, the exponent of height is 2. The Bogalusa Heart Study in children, for example, used the ponderal index, wt/ht3, in preference to BMI. Mathematical properties of such indices were addressed by Benn, who concluded that, for adults, BMI is likely to be preferable but that ideally this should be verified by testing for independence of the index from height in the population for which it is to be used.38 It has been argued that such ratios present difficulties in analysis and interpretation that can be avoided by using the component terms separately. For example, weight and height2 might be entered into a statistical model rather than BMI.39 This approach might be more informative in studies of children and adolescents, whose weight, height, and body composition develop in complex ways prior to maturity.
DETERMINANTS Determinants of obesity and its various manifestations can be viewed very broadly, on an evolutionary
46894_CH10_223_268.qxd
230
2/25/10
3:31 PM
Page 230
CHAPTER 10 OBESITY
scale or in an ecological framework. These perspectives were addressed in the preceding two chapters on dietary imbalance and physical inactivity, which are fundamental to understanding epidemic overweight and obesity. From a narrower and more immediate perspective, determinants of individual risk of overweight or obesity are listed in Table 10-2.40 Demographic, familial, and personal factors are indicated, with attention to the influence of personal factors at different stages of life. Familial influences include heredity, shared environment, and their interaction. Genetic influences have been studied in relation to different aspects of obesity, such as body mass index and body composition, as summarized from family studies by Bouchard shown in Figure 10-3.41 Body mass index appeared to have a large heritable component, only a small part of which (5%) was genetic. Body composition, measured as percent body fat or fat mass, had similar cultural heritability but substantially greater genetic heritability (25%). How genetic effects might influence obesity was suggested in a model (also proposed by Bouchard) illustrated in Figure 10-4. High levels of energy intake and low levels of energy expenditure might each have several metabolic or physiologic bases, under genetic influence. Several types of interactions are indicated, some of which are viewed as affecting nutrient partitioning, or “the proneness to store the ingested energy in the form of fat (triglycerides) or lean tissue.”41, p 96 The elements of the model are taken to suggest numerous candidate genes for further investigation. In this connection, a genome-wide association study of DNA samples from the Framingham Heart Study identified a genetic variant predisposing to obesity and present in 10 percent of the study popula-
tion,42 African Americans, and children. The finding was replicated in four other populations samples including western Europeans. The SNP identified as rs7566605, when present as a homozygote CC genotype, was found to be associated with a 1 unit average increase in BMI and greater probability of being obese relative to the GC and GG genotypes. Proximity of this variant to a candidate gene, INSIG2, was taken to implicate this gene on the basis of the function of its protein product, which inhibits fatty acid and triglyceride synthesis. Perhaps this genetic variant plays a role in the nutrient partitioning depicted in Figure 10-4. Another recent investigation of genetic factors related to type 2 diabetes identified a variant in the FTO (fat mass and obesity associated) gene that was associated specifically with increased fat mass.43 The risk-associated homozygotic state was found in 16 percent of adults in the studied populations and accounted for a 3 unit increase in BMI over those without the risk allele. In this case, the gene was of unknown function. A life-course view of obesity was considered by discussants in the American Heart Association (AHA) Prevention Conference VII, noted above, as potentially useful in identifying common critical or sensitive periods for developing obesity, especially in early life.44 Figure 10-5 represents the life-course concept, from the preconceptional state to adolescence. Beginning with mothers’ prepregnant BMI (PP BMI), progressing through prenatal development with increased or decreased birthweight (BW), the depicted course extends through adolescence. The potential culmination at this stage is development of increased BMI, adiposity, and central adiposity and consequent type 2 diabetes, insulin resistance, and CVD risk factors. For under-
Table 10-2
Obesity: Determinants and Risk Factors Demographic Factors Personal Factors Age Fetal growth and birth weight Sex Past or current overweight Race Age at adiposity rebound Socioeconomic circumstances Eating habits Geography: country of residence, urbanization, Physical inactivity/sedentary lifestyle industrialization, migration Metabolic characteristics, including diabetes mellitus Familial Factors Heredity polygenes; single gene(s) with major effect Shared environments (cultural inheritance) Interaction between genetic susceptibility and environmental exposure
Neural controls Cigarette smoking Pregnancy, including age at first pregnancy Concurrent illness or disability
Source: Reprinted with permission from FH Epstein, M Higgins, in Obesity, P Björntorp, BN Brodoff eds, p 340, © 1992, Lippincott-Raven Publishers.
46894_CH10_223_268.qxd
2/25/10
3:31 PM
Page 231
DETERMINANTS 231
30% 25%
30% Cultural Transmission Cultural Transmission
5%
Genetic
Genetic
Nontransmissible
Nontransmissible
65% 45% Percent Body Fat and Fat Mass
BMI and Amount of Subcutaneous Fat
Figure 10-3 Total Transmissible Variance and Its Genetic Component for BMI, Subcutaneous Fat (Sum of Six Skinfolds), and Total Body Fat from Underwater Weighing. Source: Reprinted from Goldbourt, Genetic Factors in Coronary Heart Disease, p 191, © 1994, with kind permission from Kluwer Academic Publishers.
Body mass or body energy for height
Nutrient partitioning
Energy intake
Human organism: the biological and behavioral interface
High level of energy intake Hyperphagia High fat intake Appetite poor ly regulated Satiety poor ly regulated Others
Note: AT-LPL, adipose tissue lipoprotein lipase; DHEA, dihydroepiandosterone; IGF-1, insulin-like growth factor.
Favoring fat deposition High skeletal muscle glycolytic metabolism Low skeletal muscle oxidative metabolism Low protein synthesis Low free testosterone Low fitness level High insulin Low IGF-1 High cortisol High AT-LPL Low DHEA Low human growth hormone Age Others
Energy expenditure
Low level of energy expenditure Low sodium ATPase Low sympathetic nervous system activity Low indirect pathway of glucose storage Low protein synthesis Low protein turnover Hypothyroidism Low substrate cycling Low level of activity Others
Figure 10-4 Major Factors Affecting Body Mass Index and Body Composition. Source: Reprinted from Goldbourt, Genetic Factors in Coronary Heart Disease, p 193, © 1994, with kind permission from Kluwer Academic Publishers.
46894_CH10_223_268.qxd
232
2/25/10
3:31 PM
Page 232
CHAPTER 10 OBESITY
Life-Course Approach to Obesity
Maternal Glycemia
PP BMI
Hyperinsulinemia
? Pancreatic development ? Placental factors
Preconceptional
BW
BW
Prenatal
Infant Feeding
Infant Growth
Infancy
Adiposity Rebound
Diet Activity Inactivity
Childhood
Adolescence
BMI
Type 2 Diabetes Mellitus
Adiposity
Insulin Resistance
Central adiposity
CVD Risk Factors
Figure 10-5 Life-Course Approach to Obesity. Source: Reprinted with permission from Circulation, Vol 110, St Jeor et al., p e472, © 2004, American Heart Association.
standing the determinants of critical periods and influences in this progression, an extensive research agenda was proposed, with identification of agedependent risk factors for obesity seen as a key aspect of addressing the global obesity epidemic. Influences from factors at individual, interpersonal, organizational, and policy/governmental levels were reviewed by the Board of Science of the British Medical Association (BMA) in a 2005 report.45 The analysis of relevant factors and potential preventive actions parallel closely the discussions of diet and physical activity in the two preceding chapters. Especially relevant here is the observed lack of a single UK body positioned to take a lead on food and health policy. Formation of a Council of Nutrition and Physical Activity had been proposed for this role, which the BMA suggested should include responsibility for food, nutrition, physical activity, and environment. This body would have sufficient scope to address those determinants of obesity amenable to policy interventions.
A study of environmental determinants extends to obesity the type of neighborhood quality assessment described previously for physical activity (the SMARTRAQ study, Chapter 9).46 Measures of physical and social environment were devised or adapted from previously published work. This study combined examination data (BMI) from an ongoing longterm study (Multi-Ethnic Study of Atherosclerosis, MESA) with a telephone survey of nonparticipants representing the same communities in which MESA participants lived. Neighborhoods scored most favorably by their residents were associated with lower mean BMI among corresponding MESA participants—2.38 units lower for women and 1.2 units lower for men per standard deviation of the physical neighborhood score. When adjusted for diet and physical activity, these differences were attenuated, indicating that part of the BMI effect was mediated through these characteristics. (For men, a favorable social score was associated unexpectedly with greater, not lesser, BMI.)
46894_CH10_223_268.qxd
2/25/10
3:31 PM
Page 233
MECHANISMS 233
MECHANISMS Obesity contributes to the occurrence of atherosclerotic and hypertensive diseases largely, if not wholly, through other cardiovascular risk factors. Kuller, for example, states that “The Null hypothesis, until proven otherwise, is that obesity, except probably at extremes, is not an independent risk factor for CVD but, rather, is an important determinant of cardiovascular risk factors and, hence, obesity indirectly is an important determinant of CVD.”47, p 23 Figure 10-6 presents a simple schematic view of the relation between obesity and the global burden of cardiovascular disease, in terms of disability adjusted life years (DALYs). The figure is based on results of a comparative quantification of global health risks by Ezzati and others, quoted by Rodgers and others.48 The area of each circle is proportionate to the contribution of each factor to risk and overlaps others in which the factors operate jointly. The proportionate contribution of each of the three factors was determined in relation to the prevalence and risks attributable to levels above its theoretical-minimum-risk exposure distribution, which assumes that the optimum range of BMI is below 21 kg/m2, systolic blood
pressure is below 115 mm Hg, and total cholesterol is below 3.8 mmol/L. BMI contributes 15% of the risk, overlapping in nearly half of its effect with systolic blood pressure and cholesterol. Other factors beyond these two are not taken into account in this analysis but would be expected, under Kuller’s hypothesis, to be the remaining intermediate factors to explain the role of BMI. A hypothetical scheme proposed by Björntorp is very much more detailed though not quantitative (Figure 10-7).21 In this scheme, primary risk factors (pathogenetic factors) or “inducers” lead through obesity to secondary risk factors—early symptoms and “disease triggers.” Positive energy balance operates through general obesity to affect blood pressure, certain blood lipids (but not LDL-cholesterol), and insulin and glucose, leading to diabetes. Additional factors operate independently or through neuroendocrine mechanisms and portal adipose tissue—the mesenteric and omental fat depots—to affect LDLcholesterol and Apo-B100, as well as the other secondary factors. With the added adverse effects of smoking as a secondary factor, these conditions produce “cardiocerebrovascular” disease. The two distinct pathways imply that obesity is sufficient to
Cardiovascular disease 100%
Systolic blood pressure over 115 mmHg 45%
Cholesterol over 3.8 mmol/l 28%
Body mass index . 21 kg/m2 15%
Note: Individual and joint contributions of high blood pressure, cholesterol, and body weight to global cardiovascular burden are shown, with the size of each circle proportional to the size of burden (as measured in DALYs) (WHO 2002). The percentages indicate the attributable burden for each risk factor, and the overlap shows disease caused by joint or mediated effects.
Figure 10-6 Global CVD Burden Caused by High Blood Pressure, Cholesterol, and Body Weight. Source: Reprinted with permission from Disease Control Priorities in Developing Countries, Second Edition, A Rodgers et al., p 853, © 2006 The International Bank for Reconstruction and Development/The World Bank.
46894_CH10_223_268.qxd
234
2/25/10
3:31 PM
Page 234
CHAPTER 10 OBESITY
Pathogenic Factors
Early Symptoms
Inducers
Disease Triggers
(Primary Risk Factors)
(Secondary Risk Factors)
Positive Energy Balance
Obesity
Blood Pressure Insulin
Stress
Diabetes
Glucose Arousal Endocrine Aberation
Smoking
Portal Adipose Tissue
VLDL
Alcohol
HDL (neg)
Saturated Fat
LDL
Salt
APO-B100
Cardiocerebrovascular Disease
Smoking
Note: HDL, high-density lipoprotein; LDL, low-density lipoprotein; VLDL, very-low-density lipoprotein.
Figure 10-7 The Role of “Primary” and “Secondary” Risk Factors in the Pathogenesis of Cardiovascular Disease, Stroke, and Diabetes. Source: Reprinted with permission from P Björntorp, “Portal” Adipose Tissue as a Generator of Risk Factors for Cardiovascular Disease and Diabetes, Arteriosclerosis, Vol 10, pp 493–496, © 1990, American Heart Association.
produce diabetes, but the contribution of portal adipose tissue is necessary to produce cardiocerebrovascular disease. Childhood and Adolescence Relationships between overweight or obesity and major cardiovascular risk factors have been investigated extensively in childhood and adolescence, when they begin. A leading example of this area of investigation is the Bogalusa Heart Study, in Louisiana, initiated in the 1970s.49 More than 3000 Black and White 5- to 18-year-old males and females were included. Percent body fat was estimated from the sum of subscapular and triceps skinfold thicknesses. Figures 10-8 and 10-9, for males and females, respectively, show the relation of this measure to total cholesterol concentration (TCHOL), LDL- to HDL-cholesterol ratio (LR-1), VLDL- LDL- to HDL-cholesterol ratio (LR-2), and systolic and diastolic blood pressure (SBP, DBP). The range of values of quintile groups by percent body fat was from less than 10% to 25% or greater for males and from less than 20% to 35% or greater for females. The odds ratio shown for each intersection in the figures represents the odds of being in the highest-quintile category of the risk factor given
the position in the first through fifth quintiles of the distribution of percent body fat. The odds for each analysis were set at 1.00 for the lowest-quintile group by percent body fat. Most striking for males was the sharply increasing odds for systolic blood pressure from the first to the middle three quintiles and again to the fifth quintile of percent body fat. The gradient was more regular but less steep for diastolic blood pressure. For the two blood lipid ratios, the difference was significant only for the highest quintile of percent body fat. The patterns for females differed from those for males in that the odds ratios were less than those for males at the one level of percent body fat where the results could be compared, 20–24.9%. They were also significant for every risk factor among those in the fourth and fifth quintiles of percent body fat (30–34.9% fat). Fatness levels of 25% for males and 30% for females were proposed as standards to define high-risk groups of children and adolescents. Project HeartBeat! was a longitudinal study of the unique trajectories with age of each of four blood lipid components and three blood pressure measures as they develop from age 8 to 18 years, in relation to multiple measures of body size and composition.50
46894_CH10_223_268.qxd
2/25/10
3:31 PM
Page 235
MECHANISMS 235
Figure 10-8 Adjusted Estimates of Relative Odds for Ranking in the Age- and Race-Specific Highest Quintile of Selected Risk Factors, for Males, Bogalusa Heart Study. Source: Reprinted with permission from DP Williams, American Journal of Public Health, Vol 82, No 3, p 361, © 1992, American Public Health Association.
Figure 10-9 Adjusted Estimates of Relative Odds for Ranking in the Age- and Race-Specific Highest Quintile of Selected Risk Factors, for Females, Bogalusa Heart Study. Source: Reprinted with permission from DP Williams, American Journal of Public Health, Vol 82, No 3, p 361, © 1992, American Public Health Association.
46894_CH10_223_268.qxd
236
2/25/10
3:31 PM
Page 236
CHAPTER 10 OBESITY
This study examined nearly 700 children every four months for up to 4 years and demonstrated that several anthropometric indices clustered in one of two groups: One group comprised BMI, fat-free mass index (the lean mass component of BMI), and abdominal circumference; the other, percent body fat, sums of either two or six skinfold measures, and fat mass index (the fat component of BMI). In general, each index of fatness, and the fat component of BMI more than the lean component, was positively associated with higher levels of blood lipid components except HDL-cholesterol, in which the association was inverse. But for blood pressure, only associations with systolic pressure were strong, whereas weaker ones appeared for fourth-phase diastolic pressure and none, after adjustment for energy intake, moderate-to-vigorous physical activity, and sexual maturation, remained with fifth-phase diastolic pressure. These and other findings indicate some complexity in the relation of body size and composition to concurrently developing cardiovascular risk factors. This complexity bears on design and evaluation of interventions on physical activity and nutrition to prevent risk factors through weight control, as well as on timing of risk-factor assessment throughout childhood and adolescence. Daniels and other contributors to an AHA Scientific Statement on overweight in children and adolescents noted additional conditions of concern: type 2 diabetes, obstructive sleep apnea, inflammation, the metabolic syndrome, increased left ventricular mass, and psychosocial difficulties with peers.23 Earlier onset of puberty in association with higher BMI has also been reported.51 Still, as observed by Dietz, “The major health consequences of obesity in adolescents are related to the adverse cardiovascular effects of the disease.”52, p 607 Adulthood Among more than 115,000 women aged 30 to 55 years at entry to the Nurses’ Health Study, body mass index was investigated in relation to other personal characteristics.53 Mean values or percentage frequencies of the following factors increased by increasing quintile group of body mass index: age, weight, hypertension, diabetes mellitus, elevated total cholesterol concentration, and parental history of myocardial infarction. Dietary intakes of fats were remarkably similar across quintile groups, and cholesterol intake varied only in being exceptionally low in the lowest quintile group of body mass index. Smoking history was essentially opposite, in that current smokers were successively less prevalent and those who never
smoked were more prevalent in higher-quintile groups by body mass index. These relations indicated the importance of adjustment for smoking status in analysis of risks in relation to obesity. A summary of cardiovascular risk-factor associations identified for body mass index and waist-hip ratio in adults included high blood pressure, electrophysiological abnormalities of cardiac function, increased blood viscosity, decreased fibrinolytic capacity, and sleep disturbances, in addition to biochemical factors.54 Relations between BMI and prevalence of other risk factors (percent of population) were investigated in the series of US national health surveys from 1960–1962 through 1999–2000 (men and women combined) (Table 10-3).55 In 1999–2000, the gradient of BMI from 25 to 30 kg/m2 was associated with higher prevalence of high cholesterol, systolic or diastolic blood pressure and diabetes, and lower prevalence of smoking. Over the series of surveys, overall prevalence of high cholesterol, blood pressure, and smoking decreased, whereas prevalence of diabetes increased, especially for the obese but in every BMI stratum. The relation of obesity to cardiovascular risk factors involves insulin resistance, a state in which the function of insulin in tissue uptake of glucose is impaired. As described in a review by Sowers, it is specifically visceral obesity that causes insulin resistance, increased blood concentrations of free fatty acids, and release of proinflammatory substances that in turn increase cardiovascular risk.56 Through type 2 diabetes and glycemic disorders, dyslipidemia, hypertension, impaired thrombolysis, endothelial dysfunction, inflammation, and microalbuminuria, insulin resistance arising from visceral adiposity contributes to development of atherosclerosis. Interventions to improve insulin sensitivity, most prominently weight loss interventions, improve these factors adding to evidence of the role of insulin resistance in atherosclerosis. In addition, a number of structural and functional abnormalities of the cardiovascular system are associated with obesity and were reviewed by Poirier and others.57 Whether reversal of these and other abnormalities associated with obesity can be achieved and sustained with weight loss continues to be discussed.58
DISTRIBUTION Prevalence, Trends, and Forecasts Figure 10-10 illustrates prevalence of overweight and obesity for age groups 6–11, 12–19, and 20–74 years, based on BMI measured in the US national health
46894_CH10_223_268.qxd
2/25/10
3:31 PM
Page 237
DISTRIBUTION 237
Table 10-3
Age- and Sex-Adjusted Trends in Cardiovascular Risk Factors in the US Population Aged 20 to 74 years* Risk Factors by NHES NHANES I NHANES II NHANES III NHANES Total Change (95% BMI Group† 1960–1962 1971–1975 1976–1980 1988–1994 1999–2000 Confidence Interval) High total cholesterol level ( 240 mg/dl) 25 27.1 22.3 22.1 13.8 15.2 11.9 (15.7 to 8.1) 25.0–29.9 39.2 33.1 31.2 23.3 18.7 20.5 (24.5 to 16.5) 30 38.9 33.1 31.5 23.0 17.9 21.0 (27.1 to 14.9) Overall 33.6 28.2 27.2 19.0 17.0 16.6 (19.8 to 13.4) High blood pressure (systolic 140 mm Hg or diastolic 90 mm Hg) 25 24.8 27.7 20.9 25.0–29.9 31.8 32.0 27.6 30 41.6 46.5 35.6 Overall 30.8 33.1 26.3
10.8 15.0 22.3 14.8
10.5 14.9 23.7 14.9
14.3 (17.5 to 11.1) 16.9 (21.3 to 12.5) 17.9 (23.0 to 12.9) 15.9 (18.9 to 12.9)
Smoking 25 25.0–29.9 30 Overall
41.4 33.1 29.7 36.0
34.0 27.0 23.6 29.3
31.3 24.3 20.2 26.4
13.4 (19.1 to 7.7) 12.3 (17.0 to 7.6) 12.3 (19.0 to 5.6) 12.8 (16.5 to 9.1)
3.4 4.3 12.3 5.3
3.2 6.3 15.0 7.4
4.0 6.3 14.0 8.1
0.6 (1.2 to 2.4) 2.0 (0.2 to 4.2) 1.7 (2.1 to 5.5) 2.8 (1.3 to 4.3)
2.4 3.0 6.3 3.5
2.1 4.6 9.0 4.6
2.8 4.2 10.1 5.0
1.3 (0.1 to 2.5) 2.6 (1.0 to 4.2) 7.2 (5.4 to 9.0) 3.2 (2.1 to 4.1)
... ... ... ...
44.7 36.6 32.5 39.2
Total diabetes (diagnosed/undiagnosed) 25 ... ... 25.0–29.9 ... ... 30 ... ... Overall ... ... Diagnosed diabetes 25 25.0–29.9 30 Overall
1.5 1.6 2.9 1.8
2.6 2.8 5.9 3.4
Abbreviations: NHANES, National Health and Nutrition Examination Survey; NHES, National Health Examination Survey. Ellipses indicate data not collected. SI conversion: To convert total cholesterol to mmol/L, multiply by 0.0259. *All prevalence estimates are age- and sex-adjusted percentages. Denominators vary for cholesterol (n 47,754), blood pressure (n 47,172), and diabetes (n 48,800) because of missing data and for smoking (n 30,124) because of missing data, smaller age range (25–74 years), and no data collected as part of the NHES † Body mass index (BMI) was calculated as weight in kilograms divided by the square of height in meters. Source: Reprinted with permission from Journal of the American Medical Association, Vol 293, No 15, EW Gregg et al., p 1871, © 2005 American Medical Association.
surveys from the early 1960s through 2003–2004.59 For adults, estimates were age adjusted. Because terminology has changed recently regarding children and adolescents, it is important to recognize that “overweight” here refers to the current category of “obese,” that is, BMI at or above the sex- and agespecific 95th percentile of the 2000 CDC Growth Charts. Categories for adults are as described previously. At ages 6–11 years, overweight (now “obesity”) neared 20% prevalence as of 2003–2004, having increased from less than 5% in 1963–1965. At ages 12–19 years, the same levels were observed throughout the period as for ages 6–11 years. Both age groups below 20 years had a higher prevalence of obesity in 2003–2004 than did adults aged 20–74 years in 1960–1962. Among adults, prevalence of obesity in-
creased more than twofold to exceed 30%, although the intermediate category, overweight, remained essentially constant, implying that desirable weight became much less frequent. More than two-thirds of the adult population were overweight or obese by 2003–2004. Children and Adolescents—United States Gortmaker and others reported in 1987 that pediatric obesity was increasing in the United States, as documented by measurement of triceps skinfold thickness in the national health surveys conducted from 1963–1965 to 1976–1980.60 Taking as reference the 85th percentile of the distribution of triceps skinfold thickness at ages 6–11 or 12–17 years in 1963–1970, the investigators found an overall 54% and 39%
46894_CH10_223_268.qxd
238
2/25/10
3:31 PM
Page 238
CHAPTER 10 OBESITY
Overweight and Obesity
100 90 80
Percent
70
Overweight including obese, 20–74 years
60 50 40
Overweight, but not obese 20–74 years
30 20
Overweight, 6–11 years
Obese, 20–74 years
10 0
03 20
19
99
–2
00
–9 4 88 19
19
–0 4
Overweight, 12–19 years
60 19 – 6 63 2 19 – 65 66 – 19 70 71 –7 4 19 76 –8 0
0
Year Notes: Estimates for adults are age adjusted. For adults: overweight including obese is defined as a body mass index (BMI) greater than or equal to 25, overweight but not obese as a BMI greater than or equal to 30. For children: overweight is defined as a BMI at or above the sex- and age-specific 95th percentile BMI cut points from the 2000 CDC Growth Charts: United States. Obese is not defined for children. See data table for data points graphed, standard errors, and additional notes.
Figure 10-10 Overweight and Obesity by Age: United States, 1960–2004. Source: Reprinted from Chartbook on Trends in the Health of Americans. Health, United States, 2006.
increase, respectively, in obesity. Superobesity, defined at the 95th percentile, increased in prevalence by 96% and 64%, respectively. The authors noted:60, p 539 “The increases in the prevalence of obesity and superobesity in the US population have ominous implications for the prevalence of a variety of associated disorders. Obesity is already the leading cause of sustained hypertension in adolescents and children.” Several intervening reports have described prevalence of child and adolescent obesity on the basis of the most current NHANES cycle or trends over multiple cycles. Health, United States, 2008 provides data through 2003–2006 for the two age groups, now 6–11 and 12–19 years, for both sexes together and sex-specific by race/ethnicity and (from 1988–1994 forward) by percent of poverty level (Table 10-4).59 As published, the table refers to “overweight,” but the data relate to the 95th percentile of BMI and in current terms report prevalence of “obesity.” As of 2003–2006, overall prevalence was 17.0%, more than four times that in the earliest available survey. Prevalence was somewhat greater for boys than girls. Racial/ethnic patterns differed between boys and girls, with prevalence strikingly greater among Mexican boys and Black or African American girls, at both age levels. These disparities are evident in data from
1976–1980 and have been consistent throughout the 30-year span of comparable data. Freedman and others called attention to the racial and ethnic differences in these trends (through 1999–2002) and the importance of interventions recognizing racial/ethnic differences.61 In the data for 2003–2006, a steep gradient of decreasing prevalence with decreasing poverty is apparent at age 6–11 years but is much less striking at age 12–19 years. Historically, the relation between poverty and obesity has changed, and in different ways for the two age groups. For ages 6 to 11, in 1988–1994, obesity was near-equally frequent across levels of poverty; increases followed at all levels but especially for the lowest level, so as to reach a nearly twofold difference from the most favorable category. For ages 12 to 19, in 1988–1994, obesity was twice as frequent in the poorest as at the more advantaged economic level, but the increase was greatest at the more advantaged level, which nearly reached the prevalence of the poorest category, which experienced a lesser increase. These opposite trends by age indicate different influences on change in obesity by age stratum and may be important to consider in devising approaches to prevention and control of child and adolescent obesity.
2/25/10 3:31 PM
continues
Overweight, Obesity, and Healthy Weight Among Persons 20 Years of Age and Over, by Sex, Age, Race and Hispanic Origin, and Percent of Poverty Level: United States, 1960–1962 Through 2003–2006 [Data are based on measured height and weight of a sample of the civilian noninsututionalized population] Sex, Age, Race and Hispanic Origin,1 Overweight (Includes Obesity)2 and Percent of Poverty Level 1960–1962 1971–1974 1976–19803 1988–1994 1999–2002 2003–2006 4 20–74 years, age-adjusted Percent of Population 44.8 47.7 47.4 56.0 65.2 66.9 Both sexes5 Male 49.5 54.7 52.9 61.0 68.8 72.6 Female 40.2 41.1 42.0 51.2 61.7 61.2 Not Hispanic or Latino: White only, male ... ... 53.8 61.6 69.5 72.1 White only, female ... ... 38.7 47.2 57.0 57.4 Black or African American only, male ... ... 51.3 58.2 62.0 72.0 Black or African American only, female ... ... 62.6 68.5 77.6 80.5 Mexican male ... ... 61.6 69.4 74.1 77.3 Mexican female ... ... 61.7 69.6 71.4 74.4 Percent of poverty level:6 Below 100% ... 49.3 50.0 59.8 65.2 66.0 100%–less than 200% ... 50.9 49.0 58.2 68.0 66.6 200% or more ... 46.7 46.6 54.5 64.9 67.0 20 years and over, age-adjusted4 Both sexes5 ... ... ... 56.0 65.1 66.7 Male ... ... ... 60.9 68.8 72.1 Female ... ... ... 51.4 61.6 61.3 Not Hispanic or Latino: White only, male ... ... ... 61.6 69.4 71.8 White only, female ... ... ... 47.5 57.2 57.9 Black or African American only, male ... ... ... 57.8 62.6 71.6 Black or African American only, female ... ... ... 68.2 77.2 79.8 Mexican male ... ... ... 68.9 73.2 75.8 Mexican female ... ... ... 68.9 71.2 73.9 Percent of poverty level:6 Below 100% ... ... ... 59.6 64.7 65.7 100%–less than 200% ... ... ... 58.0 67.3 66.5 200% or more ... ... ... 54.8 65.1 66.8 13.3 14.6 15.1 23.3 31.1 34.1 Both sexes5 Male 10.7 12.2 12.8 20.6 28.1 33.1 Female 15.7 16.8 17.1 26.0 34.0 35.2 Not Hispanic or Latino: White only, male ... ... 12.4 20.7 28.7 33.0 White only, female ... ... 15.4 23.3 31.3 32.5 Black or African American only, male ... ... 16.5 21.3 27.9 36.3 Black or African American only, female ... ... 31.0 39.1 49.4 54.3
Table 10-4
46894_CH10_223_268.qxd Page 239
DISTRIBUTION 239
3:31 PM
Sources: CDC/NCHS, National Health and Nutrition Examination Survey, Hispanic Health and Nutrition Examination Survey (1982–1984), and National Health Examination Survey (1960–1962). Data from Health, United States, 2008, pp 320–-323.
240
2/25/10
. . . Data not available. 1 Persons of Mexican origin may be of any race. Starting with 1999 data, race-specific estimates are tabulated according to the 1997 Revisions to the Standards for the Classification of Federal Data on Race and Ethnicity and are not strictly comparable with estimates for earlier years. The two non-Hispanic race categories shown in the table conform to the 1997 Standards. Starting with 1999 data, race-specific estimates for persons who reported only one racial group. Prior to data year 1999, estimates were tabulated according to the 1977 Standards. Estimates for single-race categories prior to 1999 included persons who reported one race or, if they reported more than one race, identified one race as best representing their race. See Appendix II, Hispanic origin; Race. 2 Body mass index (BMI) greater than or equal to 25 kilograms/meter2. See Appendix II, Body mass index. 3 Data for Mexicans are for 1982–1984. See Appendix I, National Health and Nutrition Examination Survey (NHANES). 4 Age-adjusted to the 2000 standard population using five age groups: 20–34 years, 35–44 years, 45–54 years, 55–64 years, and 65 years and over (65–74 years for estimates for 20–74 years). Age-adjusted estimates in this table may differ from other age-adjusted estimates based on the same data and presented elsewhere if different age groups are used in the adjustment procedure. See Appendix II, Age adjustment. 5 Includes persons of all races and Hispanic origins, not just those shown separately. 6 Percent of poverty level is based on family income and family size. Persons with unknown percent of poverty level are excluded (5% in 2003–2006). See Appendix II, Family income; Poverty. 7 Body mass index (BMI) greater than or equal to 30 kilograms/meter2. 8 BMI of 18.5 to less than 25 kilograms/meter2. Notes: Percents do not sum to 100 because the percentage of persons with BMI less than 18.5 kilograms/meter2 is not shown and the percentage of persons with obesity is a subset of the percent with overweight. Height was measured without shoes; two pounds were deducted from data for 1960–1962 to allow for weight of clothing. Excludes pregnant women. Standard errors for selected years are available in the spreadsheet version of this table. Available from: http://www.cdc.gov/nchs/hus.htm. Data have been revised and differ from previous editions of Health United States. Data for additional years are available. See Appendix III.
Overweight Obesity, and Healthy Weight Among Persons 20 Years of Age and Over, by Sex, Age, Race and Hispanic Origin, and Percent of Poverty Level: United States, 1960–1962 Through 2003–2006—continues [Data are based on measured height and weight of a sample of the civilian noninsututionalized population] Sex, Age, Race and Hispanic Origin,1 Overweight (Includes Obesity)2 and Percent of Poverty Level 1960–1962 1971–1974 1976–19803 1988–1994 1999–2002 2003–2006 4 Percent of Population 20–74 years, age-adjusted Mexican male ... ... 15.7 24.4 29.0 30.4 Mexican female ... ... 26.6 36.1 38.9 42.6 Percent of poverty level:6 Below 100% ... 20.7 21.9 29.2 36.0 35.9 100%–less than 200% ... 18.4 18.7 26.6 35.4 36.7 200% or more ... 12.4 12.9 21.4 29.2 33.1 20 years and over, age-adjusted4 Both sexes5 ... ... ... 22.9 30.4 33.4 Male ... ... ... 20.2 27.5 32.4 Female ... ... ... 25.5 33.2 34.3 Not Hispanic or Latino: White only, male ... ... ... 20.3 28.0 32.4 White only, female ... ... ... 22.9 30.7 31.6 Black or African American only, male ... ... ... 20.9 27.8 35.7 Black or African American only, female ... ... ... 38.3 48.6 53.4 Mexican male ... ... ... 23.8 27.8 29.5 Mexican female ... ... ... 35.2 38.0 41.8 Percent of poverty level:6 Below 100% ... ... ... 28.1 34.7 35.0 100%–less than 200% ... ... ... 26.1 34.1 35.9 200% or more ... ... ... 21.1 28.7 32.3
Table 10-4
46894_CH10_223_268.qxd Page 240
CHAPTER 10 OBESITY
46894_CH10_223_268.qxd
2/25/10
3:31 PM
Page 241
DISTRIBUTION 241
Adults—United States Age-adjusted data for adults, aged 20–74 years, allow similar assessment of prevalence of obesity as of 2003–2006 and trends since 1960–1962 (Table 10-5).59 For 2003–2006, prevalence of obesity in US adults was 34.1%, slightly higher in females than males. By race/ethnicity, Black and Mexican females had the highest prevalence, 54.3% and 42.6%, respectively. Only a narrow gradient was associated with poverty level, ranging from 33.1 to 36.7%, highest in the intermediate category. When the overweight category is added, including obesity, prevalence estimates nearly double. The overall prevalence of overweight and obesity was 66.0%. It was somewhat higher in males than females and was greatest in Black or African American females; prevalence among Mexicans was nearly equal for males and females, being 74.4 and 77.3%, respectively. No gradient with poverty was apparent for overweight and obesity. Prevalence of obesity increased threefold in males and doubled in females from 1960–1962 to 2003–2006, with even more rapid increases in minorities beginning with availability of race/ethnic-specific data in 1976–1980, nearly a decade later. As was seen in older children and adolescents, the relation of poverty level to trends in obesity is noteworthy: In 1971–1974, obesity was markedly greater at the lowest versus the highest category of poverty, 20.7 and 12.4%, respectively. However, in 2003–2006, the corresponding frequencies were 35.9 and 33.1%. Prevalence had increased by 75% at the lowest level and by more than 150% at the highest level. This social gradient disappeared by virtue of accelerated increase in prevalence in the highest socioeconomic category. Population-wide change in the distribution of BMI among US adults is illustrated in Figure 10-11.8 The middle panel of the figure represents the value of the 50th percentile of BMI at a given age, as observed in any of the series of national health surveys from 1960–1962 to 1988–1994. Participants in the earliest survey could have been observed at age 74 years in 1960 if born in 1886, for example. The birth cohort born in years up to 1890 corresponds to the lowest curve, whereas cohorts with birth years after 1920 would not have attained this age within the period of the 1988–1994 NHANES cycle but would have been examined at any of the earlier surveys. The solid segment of each curve represents that part of the model of a continuous relation between age at examination and 50th percentile value of BMI that reflects actual observations, and the dotted segment represents extension from the model to estimate this value at each age for every birth cohort. For each successive cohort, the 50th percentile value is greater at each age than for the preceding ones. This pattern indicates a continu-
ous increase in the 50th percentile of the BMI distribution at every age, beginning with the 1960–1962 survey and continuing to the 1988–1994 cycle. The same general configuration is shown for values of the 10th and 90th percentiles, indicating an upward shift in the whole distribution of BMI over this period. An important observation is that the increases at the 90th percentile are greater than at the 50th or 10th, which corresponds to increasing asymmetry of the BMI distribution, becoming more skewed toward higher extreme values. Change in body mass index over time within individuals was reported from the First National Health and Nutrition Examination Epidemiologic Survey Follow-Up Study, in 1990.62 Although subsequent development of the US obesity epidemic may well have altered the natural history to some degree, its main features may still be relevant. Gain in body mass index was the dominant pattern for both men and women aged 25–44 years. At ages 45–64, gains and losses were about equally represented, whereas losses outweighed gains among those 65–74 years of age at baseline. The younger adults who gained most were overweight at baseline. Women age 25–34 at baseline gained more than women age 35–44. Major weight gain (5 kg or more) was more common among Black than among White women. This analysis calls special attention to weight changes beginning in the 20s. However, the observation that those already overweight at baseline experienced the greatest increases in weight indicates that preventive measures were needed before the 20s. This is a much greater concern currently in view of the great increase in prevalence of overweight and obesity in childhood and adolescence. Forecasts of development of overweight or obesity have been developed at both individual and population levels. The Framingham Heart Study reported for persons of age 50 a residual lifetime risk of developing overweight, including obesity, of 48.0% if not yet overweight or obese, and of 91.6% disregarding presence of overweight or obesity at age 50.63 The UCLA Health Forecasting Project estimated the proportion of the California population aged 18 years and older that did or would fall in each of six BMI strata by half-decade, 1985–2025.64 From 2010 to 2025, the proportion of the population with BMI below overweight would decline from 39 to 27%, whereas overweight and the stages of obesity together would reach 73%. Economic costs were also projected to increase, from $10 billion annually to future costs totaling hundreds of billions. Global Perspective Addressing the worldwide demographics of obesity, York and others cited reports illustrating the problem of both childhood and adult obesity in many coun-
... ... ... 6.2 ... ... ... ... ... ...
... ... ... 4.7 ... ... ... ... ... ...
... ... ...
4.6 10.7 8.8
3.8 6.1 7.7 5.3
5.0 4.8
15.8 11.2 7.9
8.9 16.3 *13.4
11.6 10.7 14.1 9.7
10.5 11.3
19.8 15.1 14.9
12.6 23.5 19.6
14.6 18.7 24.7 15.3
16.0 16.7
19.3 18.4 16.3
14.5 27.7 19.9
17.3 18.5 22.1 16.8
17.6 18.2
Sources: CDC/NCHS, National Health and Nutrition Examination Survey, Hispanic Health and Nutrition Examination Survey (1982–1984), and National Health Examination Survey (1963–1965 and 1966–1970). Data from Health, United States, 2008, p 324.
*Estimates are considered unreliable. Data preceded by an asterisk have a relative standard error of 20%–30%. . . . Data not available. 1 Persons of Mexican origin may be of any race. Starting with 1999 data, race-specific estimates are tabulated according to the 1997 Revisions to the Standards for the Classification of Federal Data on Race and Ethnicity and are not strictly comparable with estimates for persons who reported only one racial group. Prior to data year 1999, estimates were tabulated according to the 1977 Standards. Estimates for single-race categories prior to 1999 included persons who reported one race or, if they reported more than one race, identified one race as best representing their race. See Appendix II, Hispanic origin; Race. 2 Data for 1963–1965 are for children 6–11 years of age; data for 1966–1970 are for adolescents 12–17 years of age, not 12–19 years. 3 Data for Mexicans are for 1982–1984. See Appendix I, National Health and Nutrition Examination Survey (NHANES). 4 Includes persons of all races and Hispanic origins, not just those shown separately. 5 Percent of poverty level is based on family income and family size. Persons with unknown percent of poverty level are excluded (3% in 2003–2006). See Appendix II, Family income; Poverty. Notes: Overweight is defined as body mass index (BMI) at or above the sex- and age-specific 95th percentile BMI cutoff points from the 2000 CDC Growth Charts: United States. Advance data from vital and health statistics; no 314. Hyattsville, MD: National Center for Health Statistics. 2000. Age is at time of examination at the mobile examination center. Crude rates, not age-adjusted rates, are shown. Excludes pregnant girls 1971–1974. Pregnancy status not available for 1963–1965 and 1966–1970. Standard errors for selected years are available in the spreadsheet version of this table. Available from: http://www.cdc.gov/nchs/hus.htm. Data have been revised and differ from previous versions of Health, United States. Data for additional years are available. See Appendix III.
6.1 6.1
3:31 PM
4.6 4.5
2/25/10
12–19 years of age Both sexes4 Boys Not Hispanic or Latino: White only Black or African American only Mexican Girls Not Hispanic or Latino: White only Black or African American only Mexican Percent of poverty level:5 Below 100% 100%–less than 200% 200% or more
Overweight Among Children and Adolescents 6–19 Years of Age, by Age, Sex, Race and Hispanic Origin, and Percent of Poverty Level: United States, 1963–1965 Through 2003–2006 [Data are based on physical examinations of a sample of the civilian noninsututionalized population] Sex, Age, Race and Hispanic Origin,1 1963–1965 and Percent of Poverty Level 1966–19702 1971–1974 1976–19803 1988–1994 1999–2002 2003–2006 6–11 years of age Percent of Population Both sexes4 4.2 4.0 6.5 11.3 15.8 17.0 Boys 4.0 *4.3 6.6 11.6 16.9 18.0 Not Hispanic or Latino: White only ... ... 6.1 10.7 14.0 15.5 Black or African American only ... ... 6.8 12.3 17.0 18.6 Mexican ... ... 13.3 17.5 26.5 27.5 Girls 4.5 *3.6 6.4 11.0 14.7 15.8 Not Hispanic or Latino: White only ... ... 5.2 *9.8 13.1 14.4 Black or African American only ... ... 11.2 17.0 22.8 24.0 Mexican ... ... 9.8 15.3 17.1 19.7 Percent of poverty level:5 Below 100% ... ... ... 11.4 19.1 22.0 100%–less than 200% ... ... ... 11.1 16.4 19.2 200% or more ... ... ... 11.1 14.3 13.5
Table 10-5
46894_CH10_223_268.qxd Page 242
46894_CH10_223_268.qxd
2/25/10
3:31 PM
Page 243
DISTRIBUTION 243
25
BMI (kg/m2)
10th
1960 1950 1940 1930 1920 1910 1900 1890
20
15
10 10
20
30
40 50 Attained Age
60
70
80
30 1960 1950 1940 1930 1920 1910 1900 1890
BMI (kg/m2)
50th 25
20
15 10
20
30
40 50 Attained Age
60
70
80
40 90th 1960 1950 1940 1930 1920 1910 1900 1890
BMI (kg/m2)
35 30 25 20 15 10
20
30
40 50 Attained Age
60
70
80
Figure 10-11 Differences Among Birth Cohorts of US Adults Born from 1890 to 1960 in Levels of BMI Attained at Ages from 18 to 74 Years—NHES Through NHANES III. Source: Reprinted with permission from Circulation, Vol 110, RH Eckel et al., p 2970, © 2004, American Heart Association.
tries.65 Data compiled by the IOTF from 31 European countries indicated the prevalence of obesity and separately of overweight by sex. Obesity reached a reported prevalence of 36.9% among men in Iceland and 40.9% among women in Yugoslavia. The lowest prevalence was in Tajikistan for both men (3.2%) and women (2%). Expectation of an upward shift in the population distribution of BMI was shown to entail a sharp increase in values above the criterion for obesity because of the changing shape of the distribution. Figure 10-12
demonstrates this point from cross-sectional, not secular trend, data. Each curve represents the BMI distribution of approximately 10 pooled populations examined in the INTERSALT Study and ranked according to the overall average BMI for the pool. The lowest average BMI is associated with a distribution that is nearly normal, that is, only slightly skewed toward higher extreme values. Successively higher average BMIs for the remaining four population groups are associated with increasing skewness. A consequence is a marked increase in prevalence of BMI above a given cutpoint, here 30 or 25 kg/m2 This concept is consistent with the picture of Figure 10-11, as discussed above. On the basis of WHO data sources, Yach and others estimated sex-specific prevalence of obesity at age 15 years and older in 2002 and as projected to 2010 worldwide, separately in high-, upper and lower middle-, and low-income regions defined by the World Bank, and in 11 of the largest countries.66 Differences by sex were small in high-income countries but were twofold or greater in lower- and low-income countries, being higher in women. The high- and upper middle-income regions have far greater prevalence (on the order of 20%) than the low-income region (on the order of 3%). The United States and Mexico were highest in prevalence and had the greatest predicted increases, with Brazil expected to have larger increases than any of the remaining countries (Bangladesh, China, India, Indonesia, Japan, Malaysia, Nigeria, and Pakistan). The aggregate burden of overweight in children worldwide was estimated under the Disease Control Priorities Project to be 17.6 million.67 Two country-specific examples illustrate the challenge for low-income countries. An extreme example of rapid weight change among individuals is the experience of the Nauruans, a small Western Pacific island population undergoing very rapid economic development in consequence of phosphate mining.68 This circumstance left a highly sedentary native Micronesian population with extremely high energy intakes—more than 7000 kcal/day for men and more than 5000 for women. Nauruans were found to have experienced a marked increase in population mean body mass index, from 32.3 to 37.1 kg/m2 for men and from 34.4 to 38.3 for women at ages 20–29 years. This change occurred between the mid-1970s and early 1980s, only a 6.5-year period. A second example is the population of Mauritius, an Indian Ocean island nation surveyed in 1987 and 1992 to study changes in prevalence of obesity and body fat distribution.69 Despite national programs for improving nutrition and physical activity, overall prevalence of overweight or obesity (BMI 25 kg/m2) for individuals at ages 20–74 years increased from 26.1 to 35.7 percent among men and 37.9 to 47.7 percent
46894_CH10_223_268.qxd
3:31 PM
Page 244
CHAPTER 10 OBESITY
Probability Density
244
2/25/10
0.10
0.05
0.00 14
16
18
20
22
24 26 28 30 32 34 Body Mass Index (kg/m2)
36
38
40
42
44
Figure 10-12 The Relationship Between the Shift in BMI Distributions and the Increasing Prevalences of Overweight and Obesity. The figure illustrates the 5 quintiles of BMI distributions found by Rose in the intersalt Study of 52 communities worldwide. Source: Reprinted with permission from Circulation, Vol 110, DA York et al., p e464, © 2004 American Heart Association.
among women. Abdominal obesity, as measured by waist-hip ratio, also increased sharply, by about 50 percent from baseline prevalence in both men and women. These changes in obesity and abdominal fat distribution were especially frequent among younger and leaner adults at entry, especially those with low income.
RATES AND RISKS Given the extensive evidence of relations between obesity and factors causing atherosclerotic and hypertensive diseases, it would be expected that population differences in distributions of BMI or other indices of overweight and obesity, body fatness, or fat distribution would be associated with cardiovascular morbidity and mortality. Relevant data can be examined between populations, within populations, and in relation to estimates of overall public health impact. Population Differences In the Seven Countries Study, coronary heart disease events and death from all causes in 10 years’ followup experience provided the basis for analysis.70 Both BMI and the sum of the triceps and subscapular skinfold thicknesses (Figure 10-13a and b, respectively) were used as measures of obesity, in men 40–59 years of age at entry and initially free of detectable coronary heart disease. Between populations, over wide ranges of both measures across all 16 cohorts, overall trends for both coronary and all-cause mortality were negative; that is, increasing measures of obesity were as-
sociated with decreasing mortality. Correlations standardized for age between each measure of obesity and incident coronary heart disease were not statistically significant (r 0.37 for BMI; r 0.35 for sum of skinfolds). However, BMI was correlated weakly (r from 0.19–0.26) with values at entry of systolic and diastolic blood pressure and serum cholesterol concentration, and equally but negatively correlated with smoking history. In part because this finding differed from experience of the life insurance industry, the relation of BMI was investigated within cohorts as well. The lowest coronary heart disease incidence tended to occur at the lowest or an intermediate level of either body mass index or skinfold thickness. Keys concluded, “In none of the areas of this study was overweight or obesity a major risk factor for death or the incidence of coronary heart disease. In most of the areas the probability of death in ten years appeared to be least for the men somewhat over the average in relative weight or fatness.”70, pp 194–195 In short, in men aged 40–59 years and free of detectable coronary heart disease at entry, differences in event rates among 16 European, North American, and Japanese populations over the following 10 years were not explained by differences in BMI or the sum of the triceps and subscapular skinfold thicknesses. The contribution of East Finland (E in the figures) to this conclusion, with its exceptionally high coronary mortality, cannot be gauged precisely. But by inspection of the figures, its exclusion would appear to reverse the sign of the correlation. In addition, mean values of BMI were mainly well below the current threshold for overweight.
46894_CH10_223_268.qxd
2/25/10
3:31 PM
Page 245
RATES AND RISKS 245
Age-Standardized CHD Deaths per 10,000
700
E
600
r 5 2 0.37
500 R
400 N
300
W G S
100
V
T
B I
Z
M
D
U
0
A
K 22
700 Age-Standardized CHD Deaths per 10,000
C
200
23
25
24 Mean Body Mass Index
26
E
600
r 5 20.35
500 R
400 300
N W
100
B
0 14
K 16
Z
S G M D T 18
B
I
C
200
20
22
24 26 28 30 Mean Sum of Skinfolds, mm
32
34
36
38
Figure 10-13 Age-Standardized Ten-Year Coronary Heart Disease Death Rates of the Cohorts Versus Their Mean Body Mass Index (Panel A) and Sum of Skinfold Thickness (Panel B) at Entry, Seven Countries Study. Ten-Year Incidence of Coronary Heart Disease in Relation to Percentage of Dietary Calories from Saturated Fat, Seven Countries Study. Source: Reprinted by permission of the publisher from Seven Countries by A Keys, Cambridge, Mass: Harvard University Press, © 1980 by the President and Fellows of Harvard College.
Individual Risks The US Pooling Project In the US Pooling Project, analysis centered on five cohorts of men (studied in Albany, Chicago—2 cohorts, Framingham, and Tecumseh) aged 40–64 years at
entry who were initially free of detectable coronary heart disease.71 The data were combined into a single pool for analysis, although cohort-specific results were also presented. As a measure of obesity, relative weight was the ratio of measured weight to desirable weight for the measured height, 100. Mean
46894_CH10_223_268.qxd
246
2/25/10
3:31 PM
Page 246
CHAPTER 10 OBESITY
values of relative weight by age group ranged from 115.8–118.6 across 5-year age subgroups. Analysis was based on 8.6 years of follow-up. Among men age 40–44 or 45–49 years at entry, men experiencing first coronary events had higher baseline relative weights, by four to five units, whereas for older men 50–54 and 55–59 years of age at entry, the baseline relative weights were either slightly greater or somewhat less among those experiencing first coronary events than those remaining free of coronary heart disease. Overall, only a modest association was present, limited to men in their 40s at entry. The Framingham Heart Study In subsequent analysis of 26-year follow-up within the Framingham Heart Study, one component of the Pooling Project, the relation between “desirable weight” or Metropolitan Relative Weight and mortality, was further investigated to take smoking history into account.72 At the beginning of the study in 1949, baseline data showed that men whose relative weight was below 100 were predominantly (80%) smokers, whereas in the highest strata of relative weight, smoking was less common (55%). Former smokers were also concentrated among the lowest relative weight men. When analyzed separately, smokers and nonsmokers exhibited different patterns in the relation of relative weight to mortality (Figure 10-14). Except for the lowest relative weight class, below 100, mortality tended to increase with increasing relative weight, although not in a wholly consistent pattern for each age group. Mortality was substantially higher for smokers than nonsmokers among both 30and 40-year-old men at entry, irrespective of relative weight. Mortality among the lowest relative weight group of smokers was greater than among the highest relative weight group of nonsmokers, excepting the 50–62 years age group. The report concluded, first, that the concept of desirable weight is supported by this analysis. Second, the strong confounding between smoking and relative weight in relation to mortality requires stratification on smoking for meaningful evaluation. Additional analysis from the Framingham Heart Study indicated that relative weight was especially predictive of incident cardiovascular disease, at 26year follow-up, in men and women younger than age 50 years at entry but not older.73 After control for high blood pressure or cholesterol, smoking, glucose intolerance, and electrocardiographic evidence of left ventricular hypertrophy, relative weight predicted coronary events, coronary death, and congestive heart failure in men and fatal and nonfatal coronary heart disease, stroke, and congestive heart failure in women.
A 30-year follow-up further confirmed that nonsmoking men who were overweight (relative weight greater than 110%) at entry to the Framingham Study had up to 3.9 times the mortality of men with desirable relative weights (100–109%).74 The Chicago Cohorts The Chicago Heart Association (CHA) Detection Project in Industry studied nearly 40,000 men and women enrolled in follow-up beginning from 1967 to 1973.75 Baseline risk factor assessment included BMI and smoking status, and vital status was determined with an average follow-up of 27 years. Tables 10-6a and b present results regarding coronary mortality for women and men, respectively. Within each of four age strata, results are shown both for quintile groups by baseline BMI and for continuous analysis. Follow-up is stratified between 0–15 and 16–27 years. For each interval of follow-up, the numbers of deaths observed and results under each of three models are presented as hazard ratios (HR). The three models are: I, adjusted for age, ethnicity, and cigarettes/day; II, adjusted for age and ethnicity but not for cigarettes/day; III, adjusted as in model II but adding systolic blood pressure, serum total cholesterol, and diabetes (yes/no). In general, the results across the two tables show that significant associations between BMI and coronary mortality are more often detected with extended follow-up (16–27 versus 0–15 years); associations are attenuated when adjustment for smoking is removed (model II versus model I); and associations are further attenuated when risk factors related to obesity are added to the adjustment (model III versus model II). The authors observed that:75, p 299 “In the CHA study, while associations of BMI with CVD and CHD mortality are significant only for men aged 60–74 in 15year follow-up, six of eight age-gender subcohorts for CHD (16–27 years) and seven of eight for CVD . . . show significant positive associations even with adjustment for obesity-related risk factors.” The Nurses’ Health Study The Nurses’ Health Study, addressed previously, reported in 1990 on experience of a large US cohort of women and showed that baseline smoking prevalence was strongly related to BMI in that population, whose follow-up began in 1976.53 Smoking was more common in the lowest than in the highest quintile group of BMI, as was the case in the data for men in Framingham. When relative risks for coronary events were estimated with and without adjustment for smoking history, the adjusted relative risks were higher with each increment in BMI. Thus for these women, also, the relation between BMI and risk of
46894_CH10_223_268.qxd
2/25/10
3:31 PM
Page 247
RATES AND RISKS 247
Figure 10-14 Proportion of Men Free of Cardiovascular Disease at Baseline Dying in 26 Years for Each Metropolitan Relative Weight Class by 10-Year Age Group and Smoking Status at Entry, Framingham Heart Study. Source: Reprinted from RJ Garrison, Journal of the American Medical Association, Vol 249, p 2201, 1983.
46894_CH10_223_268.qxd
248
2/25/10
3:31 PM
Page 248
CHAPTER 10 OBESITY
Table 10-6a
Quintile Mean Women aged 18–39 19.5b 21.9 23.8 29.0 Continuousd
Hazard Ratios for CHD Mortality by Follow-Up Period by Quintile of BMI: 16,647 Women (Panel A) and 21,732 Men (Panel B) Aged 18-74 at Baseline, Chicago Heart Association Detection Project in Industry Years 0–15 Years 16–27 Deaths HR-Ia HR-II HR-III Deaths HR-Ia HR-II HR-III 4 6 3 10 23
1.00 2.69 1.31 3.82c 1.46c
1.00 2.71 1.25 3.77c 1.48c
1.00 2.59 1.28 3.95c 1.49c
Women aged 40–9 20.0 22.2 23.9 26.2 31.7 Continuous
3 4 8 8 13 36
1.00 1.47 2.86 3.21 5.37c 1.55c
1.00 1.33 2.59 2.66 4.28c 1.47c
1.00 1.29 2.55 2.39 2.89 1.25
12 8 13 23 39 95
1.00 0.71 1.17 2.36c 4.13c 1.75c
1.00 0.65 1.07 1.99 3.39c 1.66c
1.00 0.69 1.06 2.08c 3.15c 1.49c
Women aged 50–9 20.4 22.9 24.8 27.1 32.3 Continuous
18 15 17 12 24 86
1.00 0.90 1.08 0.76 1.69 1.12
1.00 0.80 0.95 0.64 1.33 1.05
1.00 0.73 0.83 0.52 0.90 0.90
41 38 51 47 69 246
1.00 0.94 1.42 1.24 2.10c 1.36c
1.00 0.87 1.30 1.10 1.81c 1.30c
1.00 0.90 1.32 1.11 1.72c 1.21c
Women aged 60–74 20.4 23.2 25.2 27.5 32.2 Continuous
15 16 15 13 23 82
1.00 1.02 0.99 0.93 1.79 1.22
1.00 0.98 0.93 0.86 1.63 1.19
1.00 1.04 0.82 0.85 1.25 1.07
33 28 37 35 28 161
1.00 0.70 1.02 1.07 0.91 1.03
1.00 0.68 0.98 1.00 0.84 1.00
1.00 0.69 0.97 1.01 0.77 0.95
a
Model I includes age, ethnicity, and cigarettes/day; Model II removes cigarettes/day from Model I; and Model III adds SBP, serum TC, and diabetes (yes, no) to model II. b First two quintiles. c 95% confidence interval does not include 1.0. d Hazard ratio for BMI higher by one SD. Source: Reprinted with permission from Coronary Heart Disease Epidemiology from Aetiology to Public Health, M Marmot, P Elliott eds, AR Dyer, J Stamler, P Greenland, pp 297–298, © 2005 Oxford University Press.
coronary heart disease was considerably stronger when smoking history was taken into account. The Women’s Health Study A report from the Women’s Health Study (WHS) addressed the relation of BMI to stroke, including categories of ischemic and hemorrhagic stroke (Table 10-7).76 Similar to the CHA report, the WHS presented differently adjusted models across multiple strata of BMI (see legend). Adjustment for age alone showed a significant trend in the HR for total stroke because of the relation with ischemic stroke. Adding adjustment for smoking and certain other factors strengthened the association in model 1; addition of hypertension as an adjustment in model 2 attenuated
the HRs, which remained significant for ischemic stroke in model 2; additional adjustment for diabetes and high cholesterol in model III reduced the HRs to a level no longer statistically significant. The Framingham Offspring Study of adults with up to 24 years of follow-up determined that BMI was related to first cerebrovascular (and coronary) event, even after adjustment for traditional risk factors, which reduced but did not eliminate BMI as a significant predictor.77 Norwegian Adolescent Follow-Up Study In Norway, during health surveys conducted from 1963 to 1975, 227,000 adolescents had height and weight recorded.78 Follow-up to 2005 provided 8 million person-years of follow-up with nearly 10,000
46894_CH10_223_268.qxd
2/25/10
3:31 PM
Page 249
RATES AND RISKS 249
Table 10-6b
Quintile Mean Men aged 18–39 21.4 24.0 25.7 27.5 31.4 Continuousc
Hazard Ratios for CHD Mortality by Follow-Up Period by Quintile of BMI: 16,647 Women (Panel A) and 21,732 Men (Panel B) Aged 18-74 at Baseline, Chicago Heart Association Detection Project in Industry Years 0–15 Years 16–27 Deaths HR-Ia HR-II HR-III Deaths HR-Ia HR-II HR-III 8 14 11 11 17 61
1.00 1.45 1.04 1.01 1.49 1.21
1.00 1.40 1.00 0.96 1.48 1.21
1.00 1.29 0.84 0.69 0.87 1.01
19 23 37 38 60 177
1.00 1.04 1.57 1.61 2.51b 1.35b
1.00 1.02 1.53 1.54 2.50b 1.35b
1.00 0.88 1.23 1.12 1.55 1.15
Men aged 40–9 22.5 25.2 26.8 28.6 32.3 Continuous
28 36 28 38 35 165
1.00 1.36 1.07 1.38 1.33 1.06
1.00 1.28 0.99 1.33 1.25 1.04
1.00 1.11 0.82 0.99 0.82 0.90
44 56 46 59 72 277
1.00 1.36 1.14 1.40 1.88b 1.24b
1.00 1.27 1.05 1.34 1.75b 1.23b
1.00 1.25 0.97 1.11 1.45 1.13
Men aged 50–9 22.6 25.4 27.1 28.8 32.7 Continuous
51 50 68 47 73 289
1.00 1.04 1.47b 0.99 1.58b 1.16b
1.00 0.96 1.35 0.89 1.47b 1.13b
1.00 0.90 1.21 0.70 1.08 1.01
51 65 64 80 76 336
1.00 1.30 1.37 1.68b 1.79b 1.20b
1.00 1.22 1.30 1.58b 1.70b 1.19b
1.00 1.23 1.28 1.50b 1.56b 1.14b
Men aged 60–74 22.7 25.2 27.0 28.8 32.4 Continuous
31 43 35 32 68 209
1.00 1.33 1.09 1.02 2.40b 1.28b
1.00 1.30 1.04 0.97 2.26b 1.26b
1.00 1.33 1.06 0.97 2.07b 1.21b
31 28 39 48 47 193
1.00 0.87 1.19 1.60b 1.98b 1.32b
1.00 0.86 1.17 1.56 1.92b 1.32b
1.00 0.88 1.19 1.53 1.83b 1.29b
a
Model I includes age, ethnicity, and cigarettes/day; Model II removes cigarettes/day from Model I; and Model III adds SBP, serum TC, and diabetes (yes, no) to model II. b 95% confidence interval does not include 1.0. c Hazard ratio for BMI higher by one SD. Source: Reprinted with permission from Coronary Heart Disease Epidemiology from Aetiology to Public Health, M Marmot, P Elliott eds, AR Dyer, J Stamler, P Greenland, pp 297–298, © 2005 Oxford University Press.
deaths. BMI was stratified by US percentile values to form a reference category from the 25th to 75th percentile and upper strata from the 75th to 84th and 85th percentile and above. The two higher strata, relative to the reference category, experienced a relative risk for ischemic heart disease death of 2.9 for males and 3.7 for females; respective relative risks for sudden death were 2.2 and 2.7. Colon cancer and respiratory disease deaths were also associated with increased BMI. A link between adolescent BMI and adult mortality was clearly indicated by these results. The PDAY Study The Pathobiological Determinants of Atherosclerosis in Youth (PDAY) Study, cited in Chapter 3, “Ather-
osclerosis,” conducted postmortem examinations of persons dying of noncardiovascular causes at ages from 15 to 34 years.79 Risk factors assessed at the time of death were evaluated in relation to the extent and severity of atherosclerotic lesions in the right coronary artery and aorta. Obesity was represented by weight of the panniculus adiposus, or abdominal fat pad. A significant association was found, providing anatomical evidence of a link between visceral fat and atherosclerosis. The INTERHEART Study The INTERHEART Study of more than 12,000 cases of acute myocardial infarction (MI) and more than 14,000 controls in 52 countries included investigation
13 1.00 1.00 1.00 1.00 8 1.00 1.00 1.00 1.00
Ischemic stroke No. of cases (n 347) Age-adjusted Model 1† Model 2‡ Model 3§
Hemorrhagic stroke No. of cases (n 81) Age-adjusted Model 1† Model 2‡ Model 3§ 16 0.43 (0.18–1.00) 0.48 (0.20–1.12) 0.47 (0.20–1.10) 0.47 (0.20–1.10)
76 1.26 (0.70–2.27) 1.35 (0.75–2.43) 1.29 (0.72–2.33) 1.30 (0.72–2.34)
93 0.96 (0.59–1.53) 1.03 (0.64–1.65) 0.99 (0.62–1.59) 0.99 (0.62–1.60)
22 0.70 (0.31–1.57) 0.81 (0.36–1.84) 0.77 (0.34–1.74) 0.78 (0.34–1.75)
59 1.16 (0.63–2.11) 1.22 (0.67–2.23) 1.10 (0.60–2.02) 1.10 (0.60–2.02)
82 1.00 (0.62–1.61) 1.07 (0.66–1.74) 0.98 (0.61–1.59) 0.98 (0.61–1.59)
Source: Reprinted with permission from Circulation, Vol 111, T Kurth et al., p 1994.
13 0.53 (0.22–1.28) 0.63 (0.26–1.52) 0.58 (0.24–1.41) 0.58 (0.24–1.41)
49 1.22 (0.66–2.24) 1.33 (0.72–2.46) 1.15 (0.62–2.12) 1.12 (0.61–2.10)
62 0.96 (0.58–1.57) 1.06 (0.65–1.75) 0.93 (0.56–1.53) 0.92 (0.56–1.51)
10 0.42 (0.17–1.07) 0.50 (0.20–1.29) 0.45 (0.17–1.15) 0.45 (0.17–1.16)
69 1.81 (1.00–3.27) 1.94 (1.07–3.52) 1.56 (0.85–2.84) 1.49 (0.82–2.73)
81 1.32 (0.81–2.13) 1.44 (0.89–2.34) 1.18 (0.73–1.93) 1.14 (0.70–1.87)
7 0.39 (0.14–1.09) 0.46 (0.17–1.30) 0.39 (0.14–1.11) 0.38 (0.13–1.09)
50 1.77 (0.96–3.25) 1.91 (1.03–3.54) 1.41 (0.76–2.63) 1.29 (0.69–2.41)
57 1.25 (0.75–2.06) 1.37 (0.83–2.28) 1.05 (0.63–1.74) 0.97 (0.58–1.61)
5 0.64 (0.21–1.95) 0.76 (0.24–2.38) 0.59 (0.18–1.89) 0.56 (0.17–1.82)
31 2.58 (1.35–4.95) 2.81 (1.45–5.43) 1.81 (0.93–3.54) 1.54 (0.79–3.02)
36 1.84 (1.07–3.17) 2.05 (1.18–3.55) 1.38 (0.79–2.41) 1.19 (0.68–2.10)
35.0 (n 2361)
0.37 0.55 0.26 0.24
0.01 0.01 0.04 0.21
0.01 0.01 0.16 0.49
P*
3:31 PM
*p for trend across categories. † Adjusted for age. smoking status, exercise, alcohol consumption, and postmenopausal hormone use. ‡ Adjusted for all variables in model 1 plus history of hypertension. § Adjusted for all variables in model 2 plus history of diabetes and elevated cholesterol.
21 1.00 1.00 1.00 1.00
30.0–34.9 (n 4770)
2/25/10
Total stroke No. of cases (n 432) Age-adjusted Model 1† Model 2‡ Model 3§
Hazard Ratios for Total, Ischemic, and Hemorrhagic Stroke by BMI Categories BMI Categories, kg/m2 20.0 20.0–22.9 23.0–24.9 25.0–26.9 27.0–29.9 (n 2024) (n 9882) (n 7935) (n 6038) (n 6043)
250
Table 10-7
46894_CH10_223_268.qxd Page 250
CHAPTER 10 OBESITY
46894_CH10_223_268.qxd
2/25/10
3:31 PM
Page 251
RATES AND RISKS 251
of multiple risk factors.80 Obesity was studied primarily by measurement of BMI and WHR. In Figure 10-15, odds ratios for MI by quintile group of BMI (left panel) and WHR (right panel) are shown for each of three models per quintile group, as described in the figure legend. Adjusted only for age, sex, smoking, and region, BMI was consistently associated with MI in quintiles above the first. Additional adjustment for WHR greatly attenuated these associations, although they remained statistically significant for quintiles 3–5. Adjustment for all other INTERHEART risk factors accounted fully for the effect of BMI at quintiles 4 and 5, in statistical terms. The results for WHR differed from those for BMI. The gradient of relative risk with first-level adjustment was strikingly steeper; adjustment for BMI had little effect to attenuate the odds ratios; and adjustment for all other INTERHEART risk factors attenuated but did not fully account for the WHR association for quintiles 3–5. Further analysis evaluated the population attributable risk (PAR) for WHR ( 0.83 for women, 0.9 for men) and for overweight (BMI 25 kg/m2) and obesity (BMI 30 kg/m2) (Table 10-8). Overall, by sex, and by region in the INTERHEART Study, odds ratios
3.0
Adjusted for age, sex, smoking, and region Adjusted for age, sex, smoking, region, and WHR Adjusted for all other INTERHEART risk factors
for WHR were significant. Odds ratios were less, and less consistently significant, for BMI at either level. PARs for WHR ranged from 8.55 (Chinese locations) to 63.6 for Black and White mixed-race persons in South Africa. Those for overweight and for obesity ranged from 9.3 to 38.7 and from 0.76 to 18.6, respectively. The authors noted:80, p 1646 Of the three measures compared, BMI showed the weakest association with myocardial infarction risk in all ethnic groups, with no significant relation in South Asians, Arabs, and mixed-race Africans . . . By contrast, waist-tohip ratio showed a significant association with myocardial infarction in all ethnic groups, and was the strongest marker in six of the eight ethnic groups. Waist circumference was intermediate between waist-to-hip ratio and BMI in its association with myocardial infarction in most ethnic groups apart from Chinese and black Africans, in whom waist circumference was the strongest predictor. Thus, a marker of abdominal obesity was better than BMI as a predictor of myocardial infarction in all ethnic groups.
Adjusted for age, sex, smoking, and region Adjusted for age, sex, smoking, region, and BMI Adjusted for all other INTERHEART risk factors
OR (95% Cl)
2.5
2.0
1.5
1.0
0.75 Controls Cases
Q1 2860 2122
Q2 2936 2235
Q3 2906 2568
Q4 2890 2480
BMI Quintiles
Q5 2906 2651
Q1 2866 1629
Q2 2870 1816
Q3 2865 2105
Q4 2862 2750
Q5 2869 3507
Waist-to-hip Ratio Quintile
Figure 10-15 Association of BMI and Waist-to-Hip Ratio with Myocardial Infarction Risk. Vertical bars 95% CIs. Source: Reprinted with permission from The Lancet, Vol 366, p 1642, 2005.
PAR (95% C1) 2.8 (2.0 to 4.0) 5.4 (3.4 to 8.5) 2.01 (1.2 to 3.4) 5.3 (3.4 to 8.3) 0.71 (0.16 to 3.15) 1.0 (0.16 to 6.3) 4.0 (2.1 to 7.4) –0.80 (–5.41 to 3.81) 4.4 (1.9 to 9.9) 18.6 (9.6 to 32.8) –0.76 (–10.73 to 9.20) 11.9 (2.4 to 42.9)
3:31 PM
Source: Reprinted with permission from The Lancet, Vol 366, p 1646, 2005.
2/25/10
Prev prevalence. PAR population-attributable risk. BMI body-mass index. WHR waist-to-hip ratio. *Upper two-thirds of the distribution. †Overweight. ‡Obese. §Upper two quintiles for WHR had a PAR of 24.3% versus 7.7% for same quintiles for BMI. ¶Odds ratio for 2nd tertile versus 1st is 1.36, and for 3rd versus 1st is 2.24. OR for top two tertiles versus lowest tertile is 1.77 ||Black and white mixedrace in South Africa.
Odds Ratios and Population-Attributable Risk of Myocardial Infarction for Raised Waist-to-Hip Ratio or Body Mass Index High Waist-to-Hip Ratio*†§ BMI 25†§ BMI 30‡§ 0.83 women/ 0.9 men (Overweight) (Obese) Prev Prev Prev Controls OR¶ (95% C1) PAR (95% C1) Controls OR (95% C1) PAR (95% C1) Controls OR (95% C1) Overall 66.7 1.77 33.7 53.7 1.28 10.8 14.6 1.24 (1.67 to 1.88) (31.0 to 36.5) (1.21 to 1.35) (8.6 to 13.6) (1.16 to 1.33) Female 66.8 1.90 35.9 57.3 1.19 9.3 20.2 1.26 (1.69 to 2.14) (30.5 to 41.7) (1.07 to 1.32) (5.1 to 16.3) (1.12 to 1.43) Male 66.7 1.73 32.1 52.4 1.31 10.9 12.6 1.23 (1.62 to 1.85) (29.1 to 35.4) (1.23 to 1.39) (8.4 to 14.1) (1.13 to 1.34) European 68.4 2.23 44.4 63.3 1.46 16.6 20.7 1.32 (1.98 to 2.51) (39.4 to 49.6) (1.31 to 1.61) (11.7 to 23.0) (1.17 to 1.48) Chinese 53.8 1.18 8.55 37.9 1.33 11.6 4.4 1.16 (1.06 to 1.30) (4.6 to 15.4) (1.20 to 1.47) (8.4 to 15.8) (0.91 to 1.47) South Asian 68.2 1.91 36.8 46.0 1.07 –0.69 9.7 1.24 (1.65 to 2.20) (30.5 to 43.5) (0.94 to 1.21) (–6.06 to 4.68) (1.01 to 1.52) Other Asian 57.0 3.63 58.2 36.7 1.54 14.1 5.7 1.84 (2.91 to 4.52) (51.3 to 64.7) (1.27 to 1.86) (8.7 to 22.1) (1.28 to 2.64) Arabic 78.8 1.47 30.9 72.6 0.99 0.73 26.3 1.02 (1.20 to 1.82) (20.6 to 43.4) (0.83 to 1.19) (–11.48 to 12.93) (0.86 to 1.22) Latin America 79.0 2.06 44.3 64.2 1.24 9.8 18.4 1.26 (1.64 to 2.59) (34.1 to 55.1) (1.05 to 1.46) (3.5 to 24.3) (1.04 to 1.52) Black African 66.6 1.94 41.8 60.2 2.33 38.7 22.7 2.23 (1.19 to 3.17) (22.5 to 63.9) (1.49 to 3.66) (21.7 to 59.0) (1.45 to 3.45) Mixed-race African|| 71.0 3.56 63.6 59.3 1.62 18.9 27.8 1.08 (2.27 to 5.58) (49.2 to 76.0) (1.16 to 2.27) (7.1 to 41.5) (0.75 to 1.55) Other 72.8 1.85 49.1 62.2 2.13 34.3 21.1 1.95 (0.75 to 4.60) (16.7 to 82.3) (0.98 to 4.61) (9.4 to 72.4) (0.88 to 4.31)
252
Table 10-8
46894_CH10_223_268.qxd Page 252
CHAPTER 10 OBESITY
46894_CH10_223_268.qxd
2/25/10
3:31 PM
Page 253
PREVENTION AND CONTROL 253
Public Health Impact In the United States, estimates have been published of the effect of “actual causes of death” to the numbers of deaths occurring annually in this country.81,82 Tabulations of these numbers as presented for the years 1990 and 2000 are shown in Table 10-9. The intent of these analyses was to convey the idea that certified causes of death in terms of ICD codes tell less about the factors that can be modified to reduce mortality than the alternative approach taken here. The 400,000 deaths attributed to “poor diet and physical inactivity” for 2000 are described as due to overweight, an assumption that overlooks nutrient- or pattern-specific contributions, such as salt content of the usual American diet or patterns high in fats, dairy products, and refined carbohydrates, in contrast with the proposed “optimal diets” discussed in Chapter 8, “Dietary Imbalance.” The concept of “actual causes” was taken a further step by Danaei and others in addressing “preventable causes of death in the United States” considering dietary, lifestyle, and metabolic risk factors.83 Under the category of “metabolic risk factors” the authors listed BMI relative to a theoreticalminimum-risk exposure distribution of 21 kg/m2 as contributing 216,000 deaths for 2005. This is exclusive of deaths due to other dietary/physical activity factors (sodium, omega-3 fatty acids, and others), in contrast to the previous reports. Continued interest in such estimates adds importance to a series of reports on methodological issues found in the International Journal of Metabolism.84 On a global basis, the Global Burden of Disease and Risk Factors project has estimated the contribution of overweight and obesity to global mortality, years of life lost, and disability-adjusted life years lost due to diabetes mellitus, hypertensive heart disease,
ischemic heart disease, and cerebrovascular disease.85 Table 10-10 illustrates the data for worldwide mortality from these conditions, given that the estimated contributions to other measures of the health burden are very similar. The population-attributable fractions for the four identified conditions are, respectively, 49, 29, 15, and 8%. These contributions are uniformly higher for high- than for low- and middle-income countries. They are generally higher for females than males. The total numbers of deaths so attributed, again respectively, are 469,000; 260,000; 1,055,000; and 438,000—or a total of more than 2,100,000 deaths annually on a global basis. As part of the AHA’s Prevention Conference VII on the global epidemic of obesity, worldwide comorbidities were addressed.86 The theme of this component of the overall report was that:86, p 476 “Obesity is increasing in prevalence throughout the world . . . and with this change, there is a major increase in associated cardiac, metabolic, and other noncommunicable diseases. . . . Visceral or intra-abdominal obesity, in contrast to subcutaneous or lower-body obesity, carries the greatest risk of cardiac and metabolic diseases.”
PREVENTION AND CONTROL Dietary imbalance and physical inactivity, discussed in Chapters 8 and 9 previously, are a virtually universal component of recommendations and guidelines for prevention and control of cardiovascular risk factors, including overweight and obesity. Several leading organizations and agencies have proposed approaches oriented primarily to overweight and obesity, ranging from clinical interventions at the individual level to public policies regarding government
Table 10-9 Actual Causes of Death in the United States in 1990 and 2000 Actual Cause No. (%) in 1990* No. (%) in 2000 Tobacco 400,000 (19) 435,000 (18.1) Poor diet and physical inactivity 300,000 (14) 400,000 (16.6) Alcohol consumption 100,000 (5) 85,000 (3.5) Microbial agents 90,000 (4) 75,000 (3.1) Toxic agents 60,000 (3) 55,000 (2.3) Motor vehicle 25,000 (1) 43,000 (1.8) Firearms 35,000 (2) 29,000 (1.2) Sexual behavior 30,000 (1) 20,000 (0.8) Illicit drug use 20,000 ( 1) 17,000 (0.7) Total 1,060,000 (50) 1,159,000 (48.2) *Data are from McGinnis and Foege. The percentages are for all deaths. Source: Reprinted with permission from Journal of the American Medical Association, AH Mokdad et al., © 2004 American Medical Association.
99 61 80
81 86 30 53
Latin America and the Caribbean
Middle East and North Africa
South Asia
Sub-Saharan Africa
72
70 44
High-income countries
WORLD
53
47
Sub-Saharan Africa
30
81
51
18
South Asia 33
66
Middle East and North Africa
74
74
42
Low- and middle-income countries 41
62 62
37
East Asia and Pacific
Latin America and the Caribbean
61
WORLD
Europe and Central Asia
96
87
High-income countries 82
80
Low- and middle-income countries 57
96
96
80
73
54
Europe and Central Asia
40
65
37
32
9
57
58
61
32
46
77
41
43
13
70
71
73
42
49
68
47
36
30
70
70
75
39
72
91
70
64
56
93
92
95
63
53
78
50
36
14
72
76
83
55
47
76
38
18
8
65
71
73
38
32
54
30
14
9
42
48
52
26
42
61
41
24
9
54
59
68
39
24
44
22
7
3
31
38
41
17
Hypertensive Heart Disease
39
64
34
19
11
52
58
62
33
32
49
30
16
17
37
49
54
20
52
72
47
30
34
60
72
77
35
4
14
26
10
3
4
17
21
28
1
41
69
26
10
14
55
60
72
7
18
28
13
4
0
22
27
32
3
43
60
29
10
0
55
59
67
33
27
42
25
15
7
36
40
46
20
43
72
36
26
12
61
65
68
42
31
36
30
20
15
41
45
53
21
53
68
49
39
27
69
74
78
38
29
38
28
18
11
39
43
50
21
49
70
43
34
19
65
70
74
3:31 PM
East Asia and Pacific
2/25/10
Region
Contributions of Overweight and Obesity to Risks of Ischemic Heart Disease, Cerebrovascular Disease, Hypertensive Heart Disease, and Diabetes PAF of Mortality (%) 30–44 Years 45–59 Years 60–69 Years 70–79 Years 80+ Years Total Male Female Male Female Male Female Male Female Male Female Male Female All Diabetes Mellitus
254
Table 10-10
46894_CH10_223_268.qxd Page 254
CHAPTER 10 OBESITY
28
10 19
South Asia
Sub-Saharan Africa
52
25
25 13
High-income countries
WORLD
5
11
25
10
9
3
19
20
23
7
23
43
20
17
15
26
14
11
9
24
25
30
9
28
44
27
19
16
45
46
51
23
20
36
19
12
4
29
34
42
20
9
20
8
4
2
13
16
19
6
12
23
11
7
2
18
21
27
9
Cerebrovascular Disease
18
32
16
7
4
22
26
30
14
7
7
16
6
2
1
10
13
14
4
12
21
10
3
1
14
18
20
9
10
17
9
5
5
11
16
20
5
17
24
16
7
7
16
24
27
0
3
7
2
1
1
4
5
7
0
5
8
4
1
1
5
6
9
1
5
7
4
1
0
5
7
8
0
7
9
6
1
0
6
8
10
Source: Data from Global Burden of Disease and Risk Factors, AD Lopez et al. eds. © 2006. The International Bank for Reconstruction and Development/The World Bank.
16
13
Sub-Saharan Africa
8
27
15
5
South Asia 9
20
Middle East and North Africa
24
27
8
Low- and middle-income countries 11
21 19
7
East Asia and Pacific
Latin America and the Caribbean
28
WORLD
Europe and Central Asia
49
49
High-income countries 29
28
Low- and middle-income countries 26
17
35
36
39
19
10
7
13
7
4
2
10
13
16
4
15
22
13
7
3
21
22
25
10
9
11
8
6
4
12
15
17
4
15
16
15
9
7
22
23
23
10
8
12
7
5
3
11
14
16
4
15
19
14
8
5
22
23
24
3:31 PM
58
42 43
52
26
Middle East and North Africa
42
Latin America and the Caribbean
23
Europe and Central Asia
2/25/10
East Asia and Pacific
Ischemic Heart Disease
46894_CH10_223_268.qxd Page 255
PREVENTION AND CONTROL 255
46894_CH10_223_268.qxd
256
2/25/10
3:31 PM
Page 256
CHAPTER 10 OBESITY
action at local, national, and global levels. Examples cited previously7–13 include an IOTF report that presented the case for action on a global level.12 Subsequent to this report, the IOTF Prevention Group developed a set of issues and evidence requirements for obesity prevention.87 Five issues were identified as key elements of a framework for policy and program decision making in obesity prevention:12, p 23 (1) building a case for action on obesity; (2) identifying contributing factors and points of intervention; (3) defining the opportunities for action; (4) evaluating potential interventions; and (5) selecting a portfolio of specific policies, programs, and actions. Point (5) was given priority for attention on the basis that efficacy and effectiveness studies were lacking relative to the need for decision-making. The types of evidence and information identified by the Prevention Group as needed and relevant for this purpose are presented and discussed in Chapter 19, “Evidence and Decision Making” (see Table 19-2). Individual, community or population-wide, and global guidelines, recommendations, and policies are addressed as follows in the context of the relative lack of evaluations emphasized by the Prevention Group. Individual Measures Children and Adolescents Recommendations for addressing child and adolescent overweight and obesity were developed by an expert committee convened in the United States by the American Medical Association, Department of Health and Human Services, and Centers for Disease Control and Prevention and subsequently endorsed by 12 major national organizations. A summary report introduces detailed articles that separately present principles, methods, and procedures for assessment, prevention, and treatment.11,88–90 Together, this set of publications in a single supplement to Pediatrics constitutes a textbook for obesity in childhood and adolescence. Prevention focuses on evidence-supported habits: limiting consumption of sugar-sweetened beverages, eating out at restaurants—especially fast food restaurants, portion sizes, and television and other screen time and encouraging fruit and vegetable consumption, daily breakfast, and family meals. Additional advice based on expert opinion supports high intake of calcium and fiber, a diet balanced in macronutrients and limited in energy-dense foods, exclusive
breastfeeding to age 6 months, continuing to age 12 months and beyond with solid food, and moderate to vigorous physical activity for at least 60 minutes daily.11 Treatment is aimed at improvement in health habits for their own value, as well as attainment of BMI levels below the 85th percentile. A staged approach to treatment proposed four levels: “prevention plus,” to adopt healthy habits leading to reduced BMI; “structured weight management,” with added support for achieving the intended behaviors; “comprehensive multidisciplinary intervention,” with intensified behavior change, visit schedule, and specialist involvement; and “tertiary care intervention” for severely obese youths.11 In contrast to the thrust of these recommendations, including annual assessment of weight in all children, the US Preventive Services Task Force “concludes that the evidence is insufficient to recommend for or against routine screening for overweight in children and adolescents as a means to prevent adverse health outcomes.”91, p 179 Whitlock and others reported separately on a background evidence review on behalf of the USPSTF and concluded:92, p e125 BMI measurements of overweight among older adolescents identify those at increased risk of developing adult obesity. Interventions to treat overweight adolescents in clinical settings have not been shown to have clinically significant benefits, and they are not widely available. Screening to categorize overweight among children under age 12 or 13 who are not clearly overweight may not provide reliable risk categorization for adult obesity. Screening in this age group is compromised by the fact that there is little generalizable evidence for primary care interventions. In the interest of fostering needed research, Whitlock and others presented a schematic view of the flow from screening through interventions to desired outcomes and identified a set of “key questions––KQs” linked to this scheme (Figure 10-16). The IOM report, Preventing Childhood Obesity; Health in the Balance, on the contrary, put the decisionmaking dilemma in a different perspective:10, p 3 Because the obesity epidemic is a serious public health problem calling for immediate reductions in obesity prevalence and in its health and social consequences, the committee believed strongly that actions should be based on the best available evidence––as opposed to waiting for the best possible evidence. However, there is an obligation to accumulate appropriate evidence not
46894_CH10_223_268.qxd
2/25/10
3:31 PM
Page 257
PREVENTION AND CONTROL 257
1 5
Increased physical activity Dietary improvement Sedentary behavior Pre-school children (2–5 years) Latency age children (6–11 years) Adolescents (12–18 years)
Screening 2
Interventions Overweight At risk for Overweight
Stabilized or reduced BMI-for-age
4 6
3
Adverse Effects
Adverse Effects
Improved: Glucose tolerance Blood pressure Lipid disorders Physical fitness
7
Decreased childhood morbidity from diabetes mellitus, slipped capital femoral epiphysis, sleep apnea, hypertension Improved childhood functioning Reduced adult morbidity and mortality
Key Questions Arrow 1: Is there direct evidence that screening (and intervention) for overweight in childhood improves age-appropriate behavioral or physiologic measures, or health outcomes? Arrow 2: a. What are appropriate standards for overweight in childhood and what is prevalence of overweight based on these? b. What clinical screening tests for overweight in childhood are reliable and valid in predicting obesity in adulthod? c. What clinical screening tests for overweight in childhood are reliable and valid in predicting poor health outcomes in adulthood? Arrow 3: What are the adverse effects of screening, including labeling? Is screening acceptable to patients? Arrow 4: Do weight control interventions (behavioral counseling, pharmacotherapy, surgery) lead to improved intermediate outcomes, including behavioral, physiologic or weight-related measures? a. What are common behavioral and health system elements of efficacious interventions? b. Are there differences in efficacy between patient subgroups? Arrow 5: Do weight control interventions lead to improved health outcomes, including decreased morbidity, and/or improved functioning (school attendance, self-esteem and other psychosocial indicators)? Arrow 6: What are the adverse effects of interventions? Are interventions acceptable to patients? Arrow 7: Are improvements in intermediate outcomes associated with improved health outcomes? (Only evaluated if there is no direct evidence for KQ1 or KQ5 and if there is sufficient evidence for KQ4)
Figure 10-16 Screening and Interventions for Overweight in Childhood: Analytic Framework and Key Questions. Source: Reprinted with permission from Pediatrics, Vol 116, p e129, EP Whitlock et al, © 2005 by the AAP.
only to justify a course of action but to assess whether it has made a difference. Therefore, evaluation should be a critical component of any implemented intervention or change. Adults Evidence supporting weight reduction in adults includes significant delay or prevention in progression of blood pressure to levels requiring treatment.93 A similar benefit has been shown in a meta-analysis of 70 studies of blood lipid modification by weight reduction, with decrease in total, low-density, and verylow-density lipoprotein cholesterol and triglyceride levels and increased levels of high-density lipoprotein cholesterol.94 With detailed techniques and effective counseling approaches, success can be achieved in some individuals. How long such intervention effects on weight can be maintained remains uncertain.95 In support of efforts to address obesity in adults, the AMA’s Assessment and Management of Adult
Obesity: A Primer for Physicians is an online guide to clinical practice.96 A series of 10 booklets, 4 clinical tools, and 3 patient handouts are designed to facilitate obesity intervention in primary care settings. The primer is introduced with the advice that “It is never too late to start and have a favorable impact on health. Patients of all ages can and will benefit.”96, p vi The Primer draws on the clinical guidelines published in 1998 by the National Heart, Lung and Blood Institute, whose treatment algorithm is shown in Figure 10-17.97 Both guidelines call for screening of adults by use of BMI and, in addition, measurement of waist circumference. The USPSTF “recommends that clinicians screen all adult patients for obesity and offer intensive counseling and behavior interventions to promote sustained weight loss for obese adults.”91 However, evidence was found insufficient to recommend for or against either moderate- or low-intensity counseling and behavioral interventions in these patients or any level of such interventions in overweight adults.
5 Assess risk factors
6
Advise to maintain weight/address other risk factors
13
Yes
Yes
Periodic weight Check
16
Brief reinforcement/ educate on weight management
No
Hx BMI 25?
No
BMI 25 OR waist circumference > 88 cm (F) > 102 cm (M)
15
14
Yes
12
Assess reasons for failure to lose weight
No
Progress being made/goal achieved?
10
9
Clinician and patient devise goals and treatment strategy for weight loss and risk factor control
8
Maintenance counseling: • Dietary therapy • Behavior therapy • Physical activity
Yes
Yes
Yes
11
No
Does patient want to lose weight?
No
BMI 30 OR {[BMI 25 to 29.9 7 OR waist circumference > 88 cm (F) > 102 cm (M)] And 2 risk factors}
Figure 10-17 Treatment Algorithm. Source: Reprinted from Clinical Guidelines on the Identification, Evaluation, and Treatment of Overweight and Obesity in Adults. The Evidence Report. NIH Publication No. 98-4083, September 1998. National Institutes of Health, p 66.
*This algorithm applies only to the assessment for overweight and obesity and subsequent decisions based on that assessment. It does not include any initial overall assessment for cardiovascular risk factors or diseases that are indicated.
• Measure weight, height, and waist circumference • Calculate BMI
BMI measured in past 2 years?
3:31 PM
4
3
Examination Treatment
2/25/10
No
Hx of 25 BMI?
Patient Encounter
258
2
1
46894_CH10_223_268.qxd Page 258
CHAPTER 10 OBESITY
46894_CH10_223_268.qxd
2/25/10
3:31 PM
Page 259
PREVENTION AND CONTROL 259
The WHO Expert Committee report on anthropometry offered specific recommendations for individuals in each BMI category, from normal to grade 3 overweight. The recommendations reflect understanding of causes and consequences of overweight and its relation to risk factors for cardiovascular and other chronic conditions:19, pp 329–330 For individuals with BMI 18.50–24.99: avoid becoming overweight. There are no recommendations for weight loss. For individuals with BMI 25.00–29.99: avoid weight gain. Before recommending any type of intervention, assess other risk factors. If there are additional risk factors (high abdomen:hip ratio, hypertension, hyperlipidaemia, glucose intolerance or NIDDM [non–insulin-dependent diabetes mellitus], strong family history of diabetes mellitus or premature coronary heart disease), recommend a healthy lifestyle that will contribute to improvement of the risk profile: cessation of smoking, increased physical activity, reduced intake of (saturated) fat. Moderate weight loss is recommended but weight loss per se should not be the primary target of intervention. A large proportion of the adult population will usually fall into this category, and most will receive advice on healthy nutrition and physical activity appropriate for the general population. Regular (yearly) weight measurement will be helpful in monitoring weight development, and weight histories should be noted. Individuals who have continued to gain weight (e.g. 5 kg during the previous 2 years) should be identified for weight maintenance programs designed to halt the weight gain. For individuals with BMI 30.00–39.99: the same recommendations as for grade 1 overweight, although the prevalence of risk factors and of overweight-associated disorders that require medical attention is usually markedly higher and moderate weight loss is therefore more urgently recommended. In many populations, the proportion of adults falling into this category is still considerable, and treatment priorities will have to be set on the basis of, among other things, the prevalence of health problems in the community concerned. The higher the prevalence of chronic diseases such as diabetes and CVD [cardiovascular disease], the greater is the need for individuals with BMI 30.00–39.99 to lose weight. In other words, the potential impact of weight modification in preventing these problems is likely to be influenced by the disease
rates in the population. The risks related to grade 2 overweight in adults depend on other, coexisting, risk factors for chronic noncommunicable diseases. Obese individuals with no additional risk factors or conditions that require medical supervision may be referred to self-help organizations. Such organizations are effective if their leaders have sufficient training in the principles of healthy weight loss (a maximum of about 0.5 kg/week) and of balanced nutrition. For individuals with conditions that do require medical supervision, the focus should be on normalizing the risk factors or alleviating health problems (e.g., improving respiratory function or arthritis in weight-bearing joints) rather than on achieving weight loss per se. For individuals with BMI 40: intensive action to reduce weight. The proportion of adults with grade 3 overweight is small; for these individuals, weight loss per se may be the primary target and options such as surgical treatment for obesity should be considered. Community and Population-Wide Measures Because obesity in adolescence or early adulthood tracks into later life, and weight loss is difficult to achieve and maintain once obesity has become established, prevention of obesity in the first place is the logical priority. The difficulty of sustaining weight loss by already overweight or obese individuals— even if short-term weight loss is possible with one or another of the popular diets—remains a significant obstacle to control.98 This circumstance serves as a stimulus to prevention as the much preferred strategy and makes it the primary focus of community and population-wide measures. This implies early intervention to establish or maintain environmental conditions favoring optimum growth and development without lasting excess weight gain. The population-wide measures advocated for improving dietary patterns and increasing physical activity are the principal means proposed for preventing obesity at the population level.99 The 2008 AHA Scientific Statement Population-Based Prevention of Obesity describes approaches “designed to produce large-scale changes in eating behaviors and levels of physical activity to stabilize the distribution of BMI levels around a mean level that minimizes the percent who become overweight and obese, without increasing prevalence at the underweight end of the continuum.”99, p 436 The Statement adopts the concepts of the WHO report Obesity: Preventing and Managing the Global Epidemic of “universal prevention” with
46894_CH10_223_268.qxd
260
2/25/10
3:31 PM
Page 260
CHAPTER 10 OBESITY
population-wide reach irrespective of individual screening and identification; “selective prevention” for groups at high risk by virtue of demographic, health-related, or other characteristics; and “targeted prevention” intended for individuals at high risk.100 A framework is presented that links international factors—globalization of markets, development, and media programs and advertising—with national/regional, community/local, and work/ school/home-based policies and practices (Figure 10-18). Location of multiple influences at one or another of these levels points to avenues of intervention appropriate at each level. Several components of a comprehensive “spectrum of prevention” are identified, each with its rationale and examples related to increasing physical activity. The report also highlights 13 systematic reviews of intervention studies on obesity prevention, the majority addressing children and adolescents. The authors of the AHA report noted the relatively small number of studies and limited evidence of ability to im-
INTERNATIONAL FACTORS
Globalization of markets
The rapid rise in obesity on a population level––associated with changes in the quantities of food available, marketed, and consumed, along with the very low level of obligatory physical activity for most people––makes obesity prevention efforts as a primary focus truly daunting. Furthermore, the inability to specify––at a population or individual level––the exact behaviors expected to result in energy balance considerably adds to the challenge. Avoiding unhealthy weight gain goes beyond the success of individ-
NATIONAL/ REGIONAL
COMMUNITY LOCALITY
WORK/ SCHOOL/ HOME
Transport
Public Transport
Leisure Activity/ Facilities
Urbanization
Public Safety
Labor
Health
Health Care
Development Social security
Media programs & advertising
prove average BMI levels in the groups studied:99, p 447 “The relatively limited breadth of studies identified, mainly school based and mainly individually oriented, indicates an urgent need to explore preventive interventions in other settings and at multiple levels upstream. Ongoing research may broaden the evidence base, but there is an overall impression that this critical area of research has far too little focus.” The report concluded:99, p 451
Media & Culture
Education
Food & Nutrition
Sanitation
System Manufactured/ Imported Food
Agriculture/ Gardens/ Local markets
INDIVIDUAL
POPULATION
Energy Expenditure
Infections
% OBESE
Worksite Food & Activity
Family & Home
OR UNDER Food intake : Nutrient density
WT
School Food & Activity
National perspective
Figure 10-18 Societal Policies and Processes Influencing the Population Prevalence of Obesity. Source: Reprinted with permission from Circulation, Vol 108, S Kumanyika, E Obarzanek, N Stettler et al., p. 15, © 2008 American Heart Association.
46894_CH10_223_268.qxd
2/25/10
3:31 PM
Page 261
PREVENTION AND CONTROL 261
ual efforts to achieve good dietary quality and adequate physical fitness. It requires a broad range of strategies that include environmental and societal efforts. The fourth group reporting on the AHA Prevention Conference VII addressed prevention and treatment.101 Potential actions were identified across multiple settings—home/family, schools, worksites, healthcare settings, and communities and neighborhoods. The latter were characterized as promising, but relatively unexplored. The concluding discussion addressed “sector-based changes”:101, p e487 Although little research exists on the effectiveness of broad-based policies to influence the prevention of obesity, it seems clear that such programs should be developed and extensively evaluated. Policies to be considered should include changes in physical and social environments, financial incentives and tax policies, factors related to the delivery of health care, and school and work site policies. The Task Force on Community Preventive Services reported its Public Health Strategies for Preventing and Controlling Overweight and Obesity in School and Worksite Settings in 2005.102 Of 44 studies in school settings, 10 were found to be suitable for evaluation. Evidence was considered insufficient to determine the effectiveness of any of the interventions. Of 35 studies in worksites, 7 were found adequate for evaluation with comparable outcomes. Combined interventions with physical activity and nutrition components were recommended, but evidence was found to be insufficient to recommend any single-component interventions. Global Strategies The WHO Global Strategy on Diet, Physical Activity and Health was discussed earlier as an explicit strategy for achieving healthy weight as well as promoting health and prevention of chronic diseases.13 Details of the strategy regarding its potential global impact are noted in the preceding two chapters and need not be repeated here. Also working at the global level is the IOTF which, in support of objectives of the WHO strategy, has developed and released the Sydney Principles, concerning marketing of food products to children.103 The Principles were presented and reviewed at the International Congress on Obesity in Sydney, Australia, in September 2006 and posted for comment through April 2007. In accordance with the Principles, as adopted:103, pp 2–3
Actions to reduce commercial promotions to children should: 1. 2. 3. 4.
Support the rights of children. Afford substantial protection to children. Be statutory in nature. Take a wide definition of commercial promotions. 5. Guarantee commercial-free childhood settings. 6. Include cross-border media. 7. Be evaluated, monitored, and enforced. The intended impact is to counter the predominant marketing to children of energy-dense, nutrientpoor foods and the effects of such marketing on children’s food preferences, beliefs, and consumption: “If applied, the Principles should ensure a substantial level of protection for children against exposure to commercial promotions of obesogenic foods and beverages, and make a significant contribution to a multi-strategy approach to reduce childhood obesity across society.”103, p 2 The Disease Control Priorities in Developing Countries Project reviewed together interventions for blood pressure, cholesterol, and body weight from the perspective of needs in low- and middle-income countries.48 Several principles for effective population-level interventions were presented:48, p 856 • clear responsibility for coordinating prevention efforts, with credible agencies with good communication methods carrying out longterm education programs • intersectoral collaboration, with multiple messages sourced from different organizations, including health sector entities, nonhealth government agencies, schools, workplaces, religious organizations, and voluntary agencies • collaboration with the food industry to ensure the availability of reasonably priced healthier food options, with food labeling that presents relevant information in a clear, reliable, and standardized format • realistic multiyear timeframes Regarding the lifestyle and dietary interventions relevant to obesity prevention or weight reduction, it was noted that some degree of weight reduction has been achieved in a number of multi-intervention trials. However, the overall results in terms of weight reduction and maintenance of weight loss have been “relatively poor.” Given interest in this context of assessing the cost-effectiveness of interventions, it was found that no large-scale randomized trials of weight reduction as
46894_CH10_223_268.qxd
262
3/5/10
2:35 PM
Page 262
CHAPTER 10 OBESITY
an isolated intervention had been conducted, providing no basis for suitable analysis. Again, the need for further research was indicated to address fully any potential policy for obesity prevention and weight reduction. The tentative conclusion was that currently available personal interventions for these purposes were unlikely to be cost effective. The importance of population-wide initiatives against societal determinants of high-energy diets and decreasing physical activity was emphasized as the means to reduce risks of cardiovascular diseases and the need for personal interventions in developing countries in the decades ahead.
CURRENT ISSUES In its report on prevention of childhood obesity, the Institute of Medicine presented an action plan with 10 recommendations and specific steps for implementing each of them.10 One year later, a follow-up report presented an assessment of progress—Progress in Preventing Childhood Obesity: How Do We Measure Up?104 The conclusions of that report, although presented in a context specific to prevention of childhood obesity in the United States, seem relevant to the broader issues of prevention and control of overweight and obesity on a global scale:104, p 9 1. The country is beginning to recognize that childhood obesity is a serious public health problem that increases morbidity and mortality and that has substantial economic and social costs. However, the current level of investment by the public and private sectors still does not match the extent of the problem. 2. Government, industry, communities, schools, and families are responding to the childhood obesity epidemic by implementing a variety of policies, programs, and other interventions. All people bring strong values and beliefs to obesity-related issues, and evidence-based approaches are needed to guide the nation’s collective actions in this response. 3. Current data and evidence are inadequate to comprehensively assess progress in preventing childhood obesity across the United States. Although the best available evidence should be used to develop an immediate response to the childhood obesity epidemic, a more robust evidence base should be developed that identifies promising practices so that such interventions can be scaled-up and supported in diverse settings.
4. Evaluation serves to foster collective learning, accountability, responsibility, and costeffectiveness to guide improvements in childhood obesity prevention policies and programs. Multiple sectors and stakeholders should commit adequate resources to conduct evaluations. Surveillance, monitoring, and research are fundamental components of childhood obesity prevention evaluation efforts. 5. Multiple sectors and stakeholders should conduct evaluations of different types and at different levels to assess and stimulate progress over the short term, intermediate term, and long term to reverse the childhood obesity trend and improve the health of the nation’s children and youth. How to deal effectively with the public health challenges presented by overweight and obesity is the current issue. The four recommendations of the 2006 IOM report can be applied to overweight and obesity generally: lead and commit to prevention; evaluate policies and programs; monitor progress; and disseminate promising practices.
REFERENCES 1. National Institutes of Health Consensus Development Panel on the Health Implications of Obesity. Health implications of obesity. Ann Intern Med. 1985;103(suppl 6):1073–1077. 2. Brown PJ. The biocultural evolution of obesity: an anthropological view. In: Björntorp P, Brodoff BN, eds. Obesity. Philadelphia, PA: JB Lippincott Co; 1992:320–329. 3. Bray GA. Historical framework for the development of ideas about obesity. In: Bray GA, Bouchard C, James WPT, eds. Handbook of Obesity. New York: Marcel Dekker, Inc.; 1998:1–29. 4. Bray GA, Bouchard C, James WPT, eds. Handbook of Obesity. New York: Marcel Dekker, Inc.; 1998. 5. Fletcher GF, Grundy SM, Hayman L, eds. Obesity: Impact on Cardiovascular Disease. Armonk, NY: Futura Publishing Company, Inc.; 1999.
46894_CH10_223_268.qxd
2/25/10
3:31 PM
Page 263
REFERENCES 263
6. Robinson MK, Thomas A, eds. Obesity and Cardiovascular Disease. New York: Taylor & Francis; 2006. 7. US Department of Health and Human Services. The Surgeon General’s call to action to prevent and decrease overweight and obesity. Rockville, MD: US Department of Health and Human Services, Public Health Services, Office of the Surgeon General; 2001. 8. Eckel RH, York DA, Rössner S, et al. Prevention Conference VII. Obesity, a worldwide epidemic related to heart disease and stroke. Executive summary. Circulation. 2004;110:2968–2975. 9. American Medical Association. National Summit on Obesity. Building a Plan to Reduce Obesity in America. Executive Summary and Key Recommendations––October 2004. www.ama-assn.org/ama/pub/category/ 12674.html/. Accessed June 15, 2009. 10. Koplan JP, Liverman CT, Kraak VI, eds. Preventing Childhood Obesity: Health in the Balance. Committee on Prevention of Obesity in Children and Youth, Food and Nutrition Board, Board on Health Promotion and Disease Prevention, Institute of Medicine. Washington, DC: The National Academies Press; 2005. 11. Barlow SE and the Expert Committee. Expert Committee recommendations regarding the prevention, assessment, and treatment of child and adolescent overweight and obesity: summary report. Pediatrics. 2007;120(suppl 4): S164–S192. 12. Kumanyika S, Jeffrey RW, Morabia A, Ritenbaugh C, Antipatis VJ. Obesity prevention: the case for action. Int J Obesity. 2002; 26:425–436. 13. World Health Organization. Global Strategy on Diet, Physical Activity and Health. Geneva (Switzerland): World Health Organization; 2004. 14. Bray GA, Bouchard C, James WPT. Definitions and proposed current classification of obesity. In: Bray GA, Bouchard C, James WPT, eds.
Handbook of Obesity. New York: Marcel Dekker, Inc.; 1998:31–40. 15. Roche AF, Heymsfield SB, Lohman TG, eds. Human Body Composition. Champaign, IL: Human Kinetics; 1996. 16. Van Loan MD. Total body composition: birth to old age. In: Roche AF, Heymsfield SB, Lohman TG, eds. Human Body Composition. Champaign, IL: Human Kinetics; 1996: 205–215. 17. Simopoulos AP. Characteristics of obesity. In: Björntorp P, Brodoff BN, eds. Obesity. Philadelphia, PA: JB Lippincott Co; 1992:309–329. 18. World Health Organization Study Group. Diet, Nutrition, and the Prevention of Chronic Diseases. Technical Report Series 797. Geneva (Switzerland): World Health Organization; 1990. 19. World Health Organization Expert Committee. Physical Status: The Use and Interpretation of Anthropometry. Technical Report Series 854. Geneva (Switzerland): World Health Organization; 1995. 20. Heymsfield SB, Allison DB, Wang Z-M, Baumgartner RN, Ross R. Evaluation of total and regional body composition. In: Bray GA, Bouchard C, James WPT, eds. Handbook of Obesity. New York: Marcel Dekker, Inc.; 1998:41–77. 21. Björntorp P. Regional obesity. In: Björntorp P, Brodoff BN, eds. Obesity. Philadelphia, PA: JB Lippincott Co; 1992:579–586. 22. National Heart, Lung and Blood Institute. Classification of overweight and obesity by BMI, waist circumference, and associated disease risks. www.nhlbi.nih.gov/health/public/ heart/obesity/lose_wt/bmi_dis.htm. Accessed May 7, 2007. 23. Daniels SR, Arnett DK, Eckel RH, et al. Overweight in children and adolescents. Pathophysiology, consequences, prevention, and treatment. Circulation. 2005;111: 1999–2012.
46894_CH10_223_268.qxd
264
2/25/10
3:31 PM
Page 264
CHAPTER 10 OBESITY
24. Barker DJP, Martyn CN. Review: maternal and fetal origins of cardiovascular disease. In: Barker DJP, ed. Fetal and Infant Origins of Adult Disease. London: British Medical Journal; 1992:315–322. 25. Flegal KM. Defining obesity in children and adolescents: epidemiologic approaches. Critical Rev Food Sci Nutr. 1993;33:307–312. 26. Gidding SS, Leibel RL, Daniels S, Rosenbaum M, Van Horn L, Marx GR. Understanding obesity in youth. A statement for healthcare professionals from the Committee on Atherosclerosis and Hypertension in the Young of the Council on Cardiovascular Disease in the Young and the Nutrition Committee, American Heart Association. Circulation. 1996;94: 3383–3387. 27. Dietz WH, Bellizzi MC. Introduction: The use of body mass index to assess obesity in children. Am J Clin Nutr. 1999;70(suppl): 123S–125S. 28. Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard definition for child overweight and obesity worldwide: international survey. BMJ. 2000;320:1–6. 29. Obarzanek E. Obesity in children, adolescents, and families. In: Fletcher GF, Grundy SM, Hayman L, eds. Obesity: Impact on Cardiovascular Disease. Armonk, NY: Futura Publishing Company, Inc; 1999:31–53. 30. Berenson GS for the Bogalusa Heart Study Group. Obesity—a critical issue in preventive cardiology: the Bogalusa Heart Study. Prev Cardiol. 2005;8:234–241. 31. Abernathy RP, Black DR. Healthy body weights: an alternative perspective. Am J Clin Nutr. 1996;63(suppl):448S–451S. 32. Lohman TG, Roche AF, Martorell R. Anthropometric Standardization Reference Manual. Champaign, IL: Human Kinetics; 1988. 33. Mueller WH, Wear ML, Hanis CL, et al. Which measure of body fat distribution is best
for epidemiologic research? Am J Epidemiol. 1991;133;858–869. 34. Sangi H, Mueller WH. Which measure of body fat distribution is best for epidemiologic research among adolescents? Am J Epidemiol. 1991;133:870–883. 35. Roche AF. Anthropometry and ultrasound. In: Roche AF, Heymsfield SB, Lohman TG, eds. Human Body Composition. Champaign, IL: Human Kinetics; 1996:167–189. 36. Baumgartner RN. Electrical impedance and total body electrical conductivity. In: Roche AF, Heymsfield SB, Lohman TG, eds. Human Body Composition. Champaign, IL: Human Kinetics; 1996:79–107. 37. Manson JE, Stampfer MJ, Hennekens CH, Willett WC. Body weight and longevity. JAMA. 1987;257:353–358. 38. Benn RT. Some mathematical properties of weight-for-height indices used as measures of adiposity. Br J Prev Social Med. 1971;25: 42–50. 39. Kronmal RA. Spurious correlation and the fallacy of the ratio standard revisited. J Royal Stat Soc. 1993; 156(pt 3):379–392. 40. Epstein FH, Higgins M. Epidemiology of obesity. In: Björntorp P, Brodoff BN, eds. Obesity. Philadelphia, PA: JB Lippincott Co; 1992: 330–342. 41. Bouchard C. Human obesities. In: Goldbourt U, de Faire U, Berg K, eds. Genetic Factors in Coronary Heart Disease. Dordrecht (the Netherlands): Kluwer Academic Publishers; 1994:189–202. 42. Herbert A, Gerry NP, McQueen MB, et al. A common genetic variant is associated with adult and childhood obesity. Science. 2006; 312:279–283. 43. Frayling TM, Timpson NJ, Weedon MN, et al. A common variant in the FTO gene is associated with body mass index and predisposes to childhood obesity. Science. 2007;316:889–894.
46894_CH10_223_268.qxd
2/25/10
3:31 PM
Page 265
REFERENCES 265
44. St. Jeor ST, Hayman LL, Daniels SR, et al. Prevention Conference VII. Obesity, a worldwide epidemic related to heart disease and stroke. Group II: Age-dependent risk factors for obesity and comorbidities. Circulation. 2004;110:e471–e475. 45. British Medical Association Board of Science. Preventing Childhood Obesity. London: British Medical Association; 2005. 46. Mujahid MS, Diez Roux AV, Shen M, et al. Relation between neighborhood environments and obesity in the Multi-Ethnic Study of Atherosclerosis. Am J Epid. 2008;167: 1349–1357. 47. Kuller LH. The epidemiology of obesity in adults in relationship to cardiovascular disease. In: Fletcher GF, Grundy SM, Hayman L, eds. Obesity: Impact on Cardiovascular Disease. Armonk, NY: Futura Publishing Company, Inc.; 1999:3–29. 48. Rodgers A, Lawes CMM, Gaziano T, Vos T. The growing burden of risk from high blood pressure, cholesterol, and bodyweight. In: Jamison DT, Breman JG, Measham AR, et al., eds. Disease Control Priorities in Developing Countries. 2nd ed. Washington, DC: International Bank for Reconstruction and Development/The World Bank; 2006:851–868. 49. Williams DP, Going SB, Lohman TG, et al. Body fatness and risk for elevated blood pressure, total cholesterol, and serum lipoprotein ratios in children and adolescents. Am J Public Health. 1992;82:358–363. 50. Labarthe DR, Dai S, Day RS, Fulton JE, Grunbaum JA; Project HeartBeat! Findings from Project HeartBeat! Their importance for CVD prevention. Am J Prev Med. 2009;37 (1 suppl):S105–S115. 51. Kaplowitz PB, Slora EJ, Wasserman RC, Pedlow SE, Herman-Giddens ME. Earlier onset of puberty in girls: relation to increased body mass index and race. Pediatrics. 2001;108: 347–353.
52. Dietz WH. Childhood obesity. In: Björntorp P, Brodoff BN, eds. Obesity. Philadelphia, PA: JB Lippincott Co; 1992:606–609. 53. Manson JE, Colditz GA, Stampfer MJ, et al. A prospective study of obesity and risk of coronary heart disease in women. N Engl J Med. 1990;322:882–889. 54. Seidell JC. Relationships of total and regional body composition to morbidity and mortality. In: Roche AF, Heymsfield SB, Lohman TG, eds. Human Body Composition. Champaign, IL: Human Kinetics; 1996:345–353. 55. Gregg EW, Cheng YJ, Cadwell BL, et al. Secular trends in cardiovascular disease risk factors according to body mass index in US adults. JAMA. 2005;293:1868–1874. 56. Sowers JR. Obesity as a cardiovascular risk factor. Am J Med. 2003;115:37S–41S. 57. Poirier P, Giles TD, Bray GA, et al. Obesity and cardiovascular disease: pathophysiology, evaluation, and effect of weight loss. An Update of the 1997 American Heart Association Scientific Statement on Obesity and Heart Disease from the Obesity Committee of the Council on Nutrition, Physical Activity, and Metabolism. Circulation. 2006;113:898–918. 58. Hall JE, Crook ED, Jones DW, Wofford MR, Dubbert PM. Mechanisms of obesity-associated cardiovascular and renal disease. Am J Med Sci. 2002;324:127–137. 59. US Department of Health and Human Services. Health, United States, 2008 with Special Feature on the Health of Young Adults. Washington, DC: US Department of Health and Human Services. Centers for Disease Control and Prevention. National Center for Health Statistics; 2008. 60. Gortmaker SL, Dietz Jr WH, Sobol AM, Wehler CA. Increasing pediatric obesity in the United States. Am J Dis Child. 1987;141: 535–540. 61. Freedman DS, Kettel Kahn L, Serdula MK, Ogden CL, Dietz WH. Racial and ethnic
46894_CH10_223_268.qxd
266
2/25/10
3:31 PM
Page 266
CHAPTER 10 OBESITY
differences in secular trends for childhood BMI, weight, and height. Obesity. 2006;14: 301–308. 62. Williamson DF, Kahn HS, Remington PL, Anda RF. The 10-year incidence of overweight and major weight gain in US adults. Arch Intern Med. 1990;150:665–672. 63. Vasan R, Pencina MJ, Cobain M, Freiberg MS, D’Agostino RB. Estimated risks for developing obesity in the Framingham Heart Study. Ann Intern Med. 2005;143:473–480. 64. Fielding JE, Kominski GF, Hayes-Bautista DE, van Meijgaard J. Trends and forecast of health and economic costs of overweight and obesity in California. Issue Brief; July 2007. UCLA Health Forecasting Project. http://www.health forecasting.org. Accessed August 15, 2007. 65. York DA, Rössner S, Caterson I, et al. Prevention Conference VII: obesity, a worldwide epidemic related to heart disease and stroke. Group I: Worldwide demographics of obesity. Circulation. 2004;110:e463–e470. 66. Yach D, Stuckler D, Brownell KD. Epidemiologic and economic consequences of the global epidemics of obesity and diabetes. Nature Med. 2006;1:62–66. 67. Bundy DAP, Shaeffer S, Jukes M, et al. Schoolbased health and nutrition programs. In: Jamison DT, Breman JG, Measham AR, et al., eds. Disease Control Priorities in Developing Countries. 2nd ed. Washington, DC: International Bank for Reconstruction and Development/The World Bank; 2006: 1091–1108. 68. Dowse G, Zimmet P, Collins V, Finch C. Obesity in Pacific populations. In: Björntorp P, Brodoff BN, eds. Obesity. Philadelphia, PA: JB Lippincott Co; 1992:619–639. 69. Hodge AM, Dowsxe GK, Gareeboo H, Tuomilehto J, Alberti KGMM, Zimmet PZ. Incidence, increasing prevalence, and predictors of change in obesity and fat distribution over 5 years in the rapidly developing population of Mauritius. Int J Obesity. 1996;28: 137–146.
70. Keys A. Seven Countries: A Multivariate Analysis of Death and Coronary Heart Disease. Cambridge, MA: Harvard University Press; 1980. 71. The Pooling Project Research Group. Relationship of blood pressure, serum cholesterol, smoking habit, relative weight and ECG abnormalities to incidence of major coronary events: Final Report of the Pooling Project. J Chronic Dis. 1978;31:201–306. 72. Garrison RJ, Feinleib M, Castelli WP, McNamara PM. Cigarette smoking as a confounder of the relationship between relative weight and long-term mortality: the Framingham Heart Study. JAMA. 1983;249: 2199–2203. 73. Hubert HB, Feinleib M, McNamara PM, Castelli WP. Obesity as an independent risk factor for cardiovascular disease: a 26-year follow-up of participants in the Framingham Heart Study. Circulation. 1982;67: 968–977. 74. Garrison RJ, Castelli WP. Weight and thirtyyear mortality of men in the Framingham Study. Ann Intern Med. 1985;103: 1006–1009. 75. Dyer AR, Stamler J, Greenland P. Obesity. In: Marmot M, Elliott P, eds. Coronary Heart Disease Epidemiology: From Aetiology to Public Health. 2nd ed. Oxford (England): Oxford University Press; 2005:291–310. 76. Kurth T, Gaziano JM, Rexrode KM, et al. Prospective study of body mass index and risk of stroke in apparently healthy women. Circulation. 2005;111:1992–1998. 77. Wilson PWF, Bozeman SR, Burton TM, Hoaglin DC, Ben-Joseph R, Pashos CL. Prediction of first events of coronary heart disease and stroke with consideration of adiposity. Circulation. 2008;118:124–130. 78. Bjørge T, Engeland A, Tverdal A, Davey Smith G. Body mass index in adolescence in relation to cause-specific mortality: a follow-up of 230,000 Norwegian adolescents. Am J Epidemiol. 2008;168:30–37.
46894_CH10_223_268.qxd
2/25/10
3:31 PM
Page 267
REFERENCES 267
79. Strong JP, Oalmann MC, Malcom GT, Pathobiological Determinants of Atherosclerosis in Youth (PDAY) Research Group. Atherosclerosis in youth: relationship of risk factors to arterial lesions. In: Filer LJ Jr, Lauer RM, Luepker RV, eds. Prevention of Atherosclerosis and Hypertension Beginning in Youth. Philadelphia, PA: Lea & Febiger; 1994:13–18. 80. Yusuf S, Hawken S, Ôunpuu S, et al. Obesity and the risk of myocardial infarction in 27,000 participants from 52 countries: a case-control study. Lancet. 2005;366:1640–1649. 81. McGinnis JM, Foege WH. Actual causes of death in the United States. JAMA. 1993;270: 2207–2212. 82. Mokdad AH, Marks JS, Stroup DF, Gerberding JL. Actual causes of death in the United States, 2000. JAMA. 2004;291:1238–1245. 83. Danaei G, Ding EL, Mozaffarian D, et al. The preventable causes of death in the United States: comparative risk assessment of dietary, lifestyle, and metabolic factors. PLoS Med 6(4):e1000058. doi:10.1371/journal.pmed. 1000058. 84. Steinberg KK, Dietz WH. Estimating the health burden of obesity: methodologic challenges. Int J Obesity. 2008;32(suppl 3):S1–S66.
88. Krebs NF, Himes JH, Jacobson D, Nicklas TA, Guilday P, Styne D. Assessment of child and adolescent overweight and obesity. Pediatrics. 2007;120(suppl 4):S193–S228. 89. Davis MM, Gance-Cleveland B, Hassink S, Johnson R, Paradis G, Resnicow K. Recommendations for prevention of childhood obesity. Pediatrics. 2007;120(suppl 4): S229–S253. 90. Spear BA, Barlow SE, Ervin C, et al. Recommendations for treatment of child and adolescent overweight and obesity. Pediatrics. 2007; 120(suppl 4):S254–S288. 91. US Department of Health and Human Services. Agency for Healthcare Research and Quality. US Preventive Services Task Force. The Guide to Clinical Preventive Services 2006. Recommendations of the U.S. Preventive Services Task Force. AHRQ Publication No. 06-0588. Washington, DC: Agency for Healthcare Research and Quality; 2006. http://www.ahrq.gov/clinic/uspstf/uspstbac .htm. Accessed October 14, 2007. 92. Whitlock EP, Williams SB, Gold R, Smith PR, Shipman SA. Screening and interventions for childhood overweight: a summary of evidence for the US Preventive Services Task Force. Pediatrics. 2005;116:e125–e144.
85. Ezzati M, Vander Hoorn S, Lopez AD, et al. Comparative quantitation of mortality and burden of disease attributable to selected risk factors. In: Lopez AD, Mathers CD, Ezzati M, Jamison DT, Murray CJL, eds. Global Burden of Disease and Risk Factors. Washington, DC: The International Bank for Reconstruction and Development/The World Bank; 2006:241–396.
93. Cutler JA. Randomized clinical trials of weight reduction in nonhypertensive persons. Ann Epidemiol. 1991;1:363–370.
86. Caterson ID, Hubbard V, Bray GA, et al. Prevention Conference VII: Obesity, a worldwide epidemic related to heart disease and stroke. Group III: Worldwide comorbidities of obesity. Circulation. 2004;110: 2968–2975.
95. Elmer PJ. Obesity and cardiovascular disease: practical approaches for weight loss in clinical practice. In: Pearson TA, Criqui MH, Luepker RV, Oberman A, Winston M, eds. Primer in Preventive Cardiology. Dallas, TX: American Heart Association; 1994:189–204.
87. Swinburn B, Gill T, Kumanyika S. Obesity prevention: a proposed framework for translating evidence into action. Obesity Rev. 2005;6: 23–33.
96. American Medical Association. Assessment and Management of Adult Obesity. Roadmaps for Clinical Practice Series. http://www .ama-assn.org/ama/pub/physician-resources/
94. Dattilo AM, Kris-Etherton PM. Effects of weight reduction on blood lipids and lipoproteins: a meta-analysis. Am J Clin Nutr. 1992;56:320–328.
46894_CH10_223_268.qxd
268
2/25/10
3:31 PM
Page 268
CHAPTER 10 OBESITY
public-health/general-resources. Accessed June 20, 2009. 97. US Department of Health and Human Services. Clinical Guidelines on the Identification, Evaluation, and Treatment of Overweight and Obesity in Adults. The Evidence Report. NIH Publication No. 98-4083. Bethesda, MD: US Department of Health and Human Services, National Institutes of Health, National Heart, Lung and Blood Institute; September 1998. 98. Eckel RH. The dietary approach to obesity. Is it the diet or the disorder? JAMA. 2005;293: 96–97. 99. Kumanyika SK, Obarzanek E, Stettler N, et al. Population-Based Prevention of Obesity: The Need for Comprehensive Promotion of Healthful Eating, Physical Activity, and Energy Balance: A Scientific Statement from American Heart Association Council on Epidemiology and Prevention, Interdisciplinary Committee for Prevention (Formerly the Expert Panel on Population and Prevention Science). Circulation. 2008;118:428–464. 100. World Health Organization. Obesity: Preventing and Managing the Global Epidemic. WHO Technical Report Series No. 894. Geneva (Switzerland): World Health Organization; 2000.
101. Mullis RM, Blair SN, Aronne LJ, et al. Prevention Conference VII. Obesity, a Worldwide Epidemic Related to Heart Disease and Stroke. Group IV: Prevention/ Treatment. Circulation. 2004;110: e484–e488. 102. Centers for Disease Control and Prevention. Public health strategies for preventing and controlling overweight and obesity in school and worksite settings: a report on recommendations of the Task Force on Community Preventive Services. MMWR. 2005;54 (No. RR-10):1–12. 103. International Obesity Task Force. The Sydney Principles: Guiding principles for achieving a substantial level of protection for children against the commercial promotion of foods and beverages. http://www.iotf.org/sydney principles/index.asp. Accessed June 28, 2009. 104. Koplan JP, Liverman CT, Kraak VI, Wisham SL, eds. Progress in Preventing Childhood Obesity: How Do We Measure Up? Washington DC: Food and Nutrition Board, Board on Health Promotion and Disease Prevention, Institute of Medicine. Washington, DC: The National Academies Press; 2006.
46894_CH11_269_310.qxd
2/25/10
3:39 PM
Page 269
C H A P T E R
11 Adverse Blood Lipid Profile proved with systematic intervention, but these appear to be exceptional situations. It is clear that behavioral or societal factors that lead to dietary imbalance, physical inactivity, and obesity, and then to unfavorable blood lipids, call for strategies beyond the immediate public health arena alone to prevent development of adverse lipids in the first place.
SUMMARY “Adverse blood lipid profile” denotes blood concentrations of several types of fatty substances, such as cholesterol and its subfractions, LDL- and HDLcholesterol, that are an essential part of the causation of atherosclerotic cardiovascular diseases. On the basis of well-standardized laboratory methods and studies conducted in many diverse population settings, a very large body of epidemiologic evidence has established the determinants, distribution, and risks of cardiovascular diseases related to blood lipids. Laboratory research on metabolism and transport of these substances in the blood and tissues and clinical research on effects of diet and drugs as potential interventions to improve lipid profiles are other major components of the research. The totality of evidence strongly supports strategies for prevention and control of adverse blood lipids at both the individual or high-risk level and at the community or populationwide level. Not only in the United States and other developed countries, but in low- and middle-income countries as well, the mortality and burden of ischemic heart disease and stroke are largely attributable to high cholesterol and related lipid disturbances. Guidelines and policies for prevention and control of adverse levels of blood lipids are therefore of global as well as local and national importance. Personal interventions with lipid-lowering medications are addressed in detail in national and international guidelines. However, low prevalence of effective control of cholesterol found in national surveys and other sources indicates only limited success of efforts to implement them. In some healthcare systems and workplace programs, control levels have been im-
INTRODUCTION The “blood lipid profile” refers to concentrations of several fatty substances, mainly cholesterol and triglycerides, circulating in the blood. An “adverse” blood lipid profile represents levels of these components that together increase the extent and severity of atherosclerosis and risk of coronary heart disease and ischemic stroke. “High cholesterol” (hypercholesterolemia), unless otherwise specified, usually refers to elevated total cholesterol concentrations, although common reference to “bad” and “good” cholesterol indicates wide recognition of two subfractions of cholesterol, LDL- (low density lipoprotein) cholesterol and HDL- (high density lipoprotein) cholesterol, respectively. Optimum levels are low in the case of LDL-cholesterol and high for HDL-cholesterol. Because these components of total cholesterol relate to risk in opposite ways, “high cholesterol” does not fully characterize blood lipid-related risks, and reference to the “profile” is therefore often preferred. However, over much of the period in which the causal role of blood lipids in atherosclerosis and its consequences was established, “cholesterol” meant total cholesterol, and this is reflected in much of the literature cited here.
269
46894_CH11_269_310.qxd
270
2/25/10
3:39 PM
Page 270
CHAPTER 11 ADVERSE BLOOD LIPID PROFILE
The Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III, ATP III”), published in 2002, provides an extensive review of the blood lipid profile and its components, including the rationale for intervention, with evidence-based recommendations; procedures for detection and evaluation; approaches to lifestyle and medical intervention; and related topics.1 (For the forthcoming update to ATP III, see www.nhlbi.nih .gov/about/ncep.) The progression from early animal experimental research on diet and atherosclerosis to contemporary understanding of the roles of blood lipids at both population and individual levels is documented in reviews by Stamler,2 Keys,3 the National Research Council,4 and others. In the early 1900s, it was discovered that cholesterol was the necessary dietary constituent to produce experimental atherosclerosis in rabbits and other species. The significance of this knowledge for human atherosclerosis was recognized only later. However, as early as 1916, plasma cholesterol concentrations were observed to be lower in the native population of Indonesia than in Dutch immigrants; coronary heart disease was much less frequent in the indigenous population, and this difference in cholesterol levels was considered to be the explanation. This and other observations cited by Keys “seemed to fit into one picture” suggesting the importance of differences in population distributions of blood cholesterol concentration.3, p 1 In the mid- to late 1950s, as background to the Seven Countries Study, cholesterol surveys were conducted in several populations including countries in Europe as well as Japan and the United States. These studies were the foundation of work that would establish the causal relation between blood cholesterol concentration and coronary heart disease. During the 1950s and 1960s, the relation between dietary intake of fat and cholesterol and blood cholesterol concentration was quantified, as described in Chapter 8, “Dietary Imbalance.” Through the 1970s and 1980s, the composition of blood lipids and their transporting proteins (lipoproteins) was investigated further. Mechanisms of lipid transport and metabolism were also being described at a new level of detail. Epidemiologic studies over this period established the inverse relation of the concentration of HDL-cholesterol to risk of coronary heart disease. At the same time, clinical trials were undertaken to test the efficacy of dietary or drug interventions to reduce total cholesterol concentrations. Resulting evidence that incidence of coronary heart disease could
be reduced stimulated increased public health attention to blood cholesterol. A significant outcome of this progression of research on blood lipids has been development of clinical guidelines and public health policies for prevention of atherosclerosis and its consequences. More recently, a major research focus has been on genetics of blood lipids and atherosclerosis, discussed in Chapter 7, “Genes and Environment.”
CONCEPTS AND DEFINITIONS How blood lipids are described at a molecular level is basic to estimating cardiovascular risk, developing guidelines for classification and treatment, and monitoring risk-factor prevention and control. From both clinical and public health perspectives, it is helpful to understand this aspect of blood lipids. This provides background for discussion of the potential place of various lipid components in policy and practice. Lipoprotein Molecules Table 11-1 presents a broad classification of plasma lipoproteins and indicates several of their properties, which include, in the second column, the major lipids associated with each.5 One or more apolipoproteins, shown in the third column, occupy the surface of the lipoprotein molecule. Apolipoproteins give the molecule surface properties that determine its potential interactions with specific enzymes and cell-surface receptors.6 Electrophoretic mobility is described for each lipoprotein class, in the last column. This property is the extent to which molecules of each lipoprotein class migrate across a suitable transport medium under the influence of an electrical potential. This technique of lipid separation, used for many years, was the basis for an earlier classification in which the distance of migration was characterized for example as “Origin” (no migration), or “” (intermediate migration)—therefore low-density lipoprotein or LDL is equivalent to “ lipoprotein” and LDL-cholesterol or LDL-C is equivalent to lipoprotein cholesterol. Often, inconveniently, LDL-cholesterol is referred to simply as “LDL,” which strictly refers only to the lipoprotein itself and not the cholesterol it contains. The listed lipoprotein classes are arrayed in order of decreasing molecular size and increasing density (fourth and fifth columns). Chylomicrons carry dietary, or exogenous, triglycerides and cholesterol from the intestine into the circulation, and 80–95% of their lipid content is triglycerides. Remnants are portions of cholesterol-laden lipoprotein remaining after breakdown of chylomicrons. Very-low-density
46894_CH11_269_310.qxd
2/25/10
3:39 PM
Page 271
CONCEPTS AND DEFINITIONS 271
Table 11-1 Lipoprotein Class Chylomicrons Remnants VLDL IDL LDL HDL2 HDL3
Classification and Properties of Plasma Lipoproteins Major Lipids Dietary triglycerides, cholesteryl esters Dietary cholesteryl esters Endogenous triglycerides Cholesteryl esters, triglycerides Cholesteryl esters Cholesteryl esters Cholesteryl esters
Apolipoproteins A-I, A-II, A-IV, B-48, C-I, C-II, C-III, E B-48, E
Density (g/ml) 0.95
Diameter (Å) 800–5000
Electrophoretic Mobility Origin
1.006
300
Origin
B-100, C-, C-II, C-III, E B-100, E
1.006
300–800
Pre-
1.006–1.019
250–350
Pre-/
B-100 A-I, A-II A-I, A-II
1.019–1.063 1.063–1.125 1.125–1.210
180–280 90–120 50–90
Note: VLDL, very-low-density lipoprotein; IDL, intermediate-density lipoprotein; LDL, low-density lipoprotein; HDL, high-density lipoprotein. Source: Reprinted with permission from PH Jones, J Patsch, and AM Gotto Jr. The Biochemistry of Blood Lipid Regulation and the Assessment of Lipid Abnormalities, in the Heart: Arteries and Veins, RC Schlant and RW Alexander, p 975, © 1994, The McGraw-Hill Companies.
and intermediate-density lipoproteins (VLDL and IDL) both carry endogenous triglycerides as about 55–80% and 20–50% of their lipid content, respectively. IDL carries endogenous cholesterol as about 20–40% of its lipid content. LDL and HDL both carry endogenous cholesterol—40–50% and 15–25% of their lipid content, respectively. But these two lipoprotein classes are associated with different apolipoproteins (third column) and have different metabolic pathways, resulting in opposite roles in atherogenesis. Examples of other molecular components of the blood lipid profile that have been investigated both in the laboratory and in population studies are lipoprotein(a) (Lp(a), or “L P little a”) and apolipoprotein E (apoE, or “apo E”). Lp(a) is a circulating protein whose blood concentration appears to be primarily genetically controlled. Its metabolism and mechanisms of action may include effects on blood coagulation and binding of materials within the atherosclerotic plaque. It is linked with triglyceride metabolism and with cardiovascular risk.7 ApoE is also genetically determined, with three alleles (designated ε2– ε4) accounting for most of the observed genotypes. These apolipoproteins can be associated with severe lipid disorders and risk of coronary artery disease even when blood lipid profiles are within the usual range of variation.8 Blood Lipid Phenotypes A second approach to classification of blood lipids is the so-called Fredrickson classification. This scheme addresses a series of phenotypes, or clinical patterns, characterized by specific combinations of lipoproteins, total cholesterol levels, and triglyceride levels;
different degrees of atherogenicity; and associations with particular genetic disorders (rare disorders with dominant genetic defects) (Table 11-2).5 The Fredrickson types I–V and their equivalent terms, such as “familial hypercholesterolemia” for type II a or b, are encountered in much of the clinical and some of the epidemiologic literature of the past. This classification does not take HDL-cholesterol concentration into account and has, in this respect, been superseded by more recent approaches.1 The place of HDL-cholesterol and of triglycerides in prevention of coronary heart disease was reviewed in a National Institutes of Health consensus conference in the early 1990s.9 Because high triglyceride levels and low levels of HDL-cholesterol are associated, their independent contributions to risk have been difficult to evaluate. Investigators continue to argue the importance of addressing these lipid components. In the case of HDL-cholesterol, multiple mechanisms have been identified by which it can protect against atherosclerosis and coronary artery disease,10 and treatment has been proposed for low HDL-cholesterol, beginning with lifestyle interventions and progressing as needed to specifically HDL-cholesterol-raising drugs.11,12 Regarding elevated triglycerides, the case continues to be made for treatment on the basis that this is an independent risk factor found in more than 30 percent of the United States adult population.13 Except at extreme high levels, triglyceride is proposed to be treated by lifestyle interventions, although drugs are available. Other approaches to defining risk in relation to blood lipids include a category denoted as “non-HDLcholesterol.” This includes all cholesterol except the
46894_CH11_269_310.qxd
272
2/25/10
3:39 PM
Page 272
CHAPTER 11 ADVERSE BLOOD LIPID PROFILE
Table 11-2
Phenotype I IIa
Fredrickson Classification of the Hyperlipidemias Serum Serum Lipoprotein(s) Cholesterol Triglyceride Elevated Level Level Atherogenicity Chylomicrons Normal to ↑ ↑↑↑↑ None seen LDL ↑↑ Normal
IIb
LDL and VLDL
↑↑
↑↑
III IV
IDL VLDL
↑↑ Normal to ↑
↑↑↑ ↑↑
V
VLDL and chylomicrons
Normal to ↑
↑↑↑↑
Associated with Genetic Disorders Familial lipoprotein lipase deficiency Familial hypercholesterolemia LDL receptor abnormal Familial combined hyperlipidemia Polygenic hypercholesterolemia Familial hypercholesterolemia Familial combined hyperlipidemia Familial dysbetalipoproteinemia Familial hypertriglyceridemia Familial combined hyperlipidemia Familial hypertriglyceridemia Familial multiple-lipoprotein-type hyperlipidemia
Note: LDL, low-density lipoprotein; VLDL, very-low-density lipoprotein; IDL, intermediate-density lipoprotein. Relative degrees of atherogenicity are indicated by plus signs. High-density lipoprotein (HDL) cholesterol levels are not considered in the Fredrickson classification. Source: Reprinted with permission from AM Gotto Jr, Lipid and Lipoprotein disorders, in TA Pearson et al., Primer in Preventive Cardiology, © 1994, American Heart Association, and by permission from Southern Medical Journal, Vol 88, No 4, pp 379–391, 1995.
portion included in HDL-cholesterol, from which it is therefore independent.1 It includes the cholesterol associated with apolipoprotein B, excludes that associated with apolipoprotein A, and includes triglycerides (see Table 11-1). Because total and HDLcholesterol can both be measured in nonfasting blood samples, testing is convenient and a simple calculation provides the result. Determining optimum treatment goals and evaluating treatment options separately for non-HDL-cholesterol from those for LDL-cholesterol alone are problematic but may offer greater overall public health impact.14 Use of ratios among blood lipid components has also been advocated, such as total or LDL-cholesterol/HDL-cholesterol, as compared with total or LDL-cholesterol alone.15 Advantage has been shown for the ratio in predicting coronary risk and therapeutic risk reduction.15 Further, measurement of apolipoproteins A and B has been advocated in place of measurement of the corresponding cholesterol concentrations on grounds of more complete assessment of blood lipid contributions to risk.16 Current Classification in the United States and Europe In the United States currently, classification of blood lipid concentrations for adults under ATP III gives priority to LDL-cholesterol, the “primary target of therapy.”17 Five categories of LDL-cholesterol are defined by cut-points in mg/dl: optimal, 100; near optimal/above optimal, 100–129; borderline high, 130–159; high, 160–189; and very high, 190. For total cholesterol, categories are consistent with past definitions: desirable, 200; borderline high,
200–239; and high, 240. HDL-cholesterol categories are also defined in ATP III: low, 40; and high, 60. Reflecting recent attention to HDL-cholesterol, “low” is now defined at a higher—that is, less extreme—cut point than the previous levels of 30 or 35 mg/dl. Treatment algorithms are based on these categories after consideration of other risk factors that may modify therapy. In Europe, recommendations for cardiovascular disease prevention have been developed jointly by several multinational societies, as discussed in Chapter 20, “Recommendations, Guidelines, and Policies.” Rather than retaining risk categories for single factors, the joint recommendations incorporate other risk factors with total cholesterol concentration within a multivariable risk prediction.18 Nonetheless, target treatment levels are specified for total and LDL-cholesterol. With respect to HDL-cholesterol and triglycerides, levels are specified as markers of increased cardiovascular risk, but no treatment goals are defined. The World Health Organization guidelines, with potential global reach, closely resemble the European recommendations.19 They differ, however, in avoiding specific target levels for treatment on the argument that no level has been found below which there is no further benefit to reduction. Therefore, it is reasoned, no target should be set that could limit the effort to reduce cholesterol to the lowest attainable level. For children (age 2–19 years) in the United States, the NCEP guidelines remain unchanged since their publication in 1991.20 “Acceptable,” “borderline,” and “high” categories are defined by total cholesterol concentrations of 170, 170–199, and 200
46894_CH11_269_310.qxd
2/25/10
3:39 PM
Page 273
MEASUREMENT 273
mg/dl, respectively. Corresponding values for LDLcholesterol are 110, 110–129, and 130 mg/dl. No HDL-cholesterol criteria are indicated for this age group. Because levels of each component of the blood lipid profile vary systematically on different trajectories during childhood and adolescence, it has been suggested that uniform criteria for classification throughout this period of life are inappropriate.21 The American Academy of Pediatrics now recommends age-specific percentile standards (50th, 75th, 90th, 95th) separately by sex and age (5–9, 10–14, and 15–19 years) for total, LDL-, and HDLcholesterol and for triglycerides.22 Cut-points by other percentile values have also been proposed, by single year of age from 12 to 20 years.23
MEASUREMENT Concerns about measurement of blood lipids center on accuracy of laboratory methods. For epidemiologic studies, an additional concern is within-person variation and the limited reliability of a single determination for characterizing the lipid profile or its components in an individual. Laboratory Standardization Issues regarding between-laboratory variation in methods and reliability of cholesterol determination led to establishment in 1958 of the Lipid Standardization Program by the (US) Centers for Disease Control and Prevention (CDC) and the National Bureau of Standards. Through this program, a reference method (the Abell-Kendall method) and a detailed quality-control program have been maintained for more than 50 years, and laboratories throughout the United States and around the world can become certified as meeting published standards. Because of the longstanding operation of this program, it has been possible to determine cardiovascular risks attributable to blood lipids; assure reliability of test results for multicenter clinical trials; monitor distributions and trends in cholesterol levels in many populations; and provide the foundation for policies and practices regarding blood lipids. Separate issues arise in connection with desktop cholesterol analyzers, of which numerous models are available. An assessment of these devices by the US General Accounting Office (GAO) found that measurement performance was generally good under controlled laboratory conditions.24 However, their reliability in practice settings is less certain. Their use to process capillary, or finger-stick, rather than venous blood samples entails further difficulties because of
possible sample dilution through poor finger-stick technique. Variability Within Individuals Because of within-individual variability associated with single cholesterol measurements, multiple measurements provide a more reliable characterization of the individual. The ranges of actual values represented by single test results were estimated by the GAO.24 Both analytical variability, based on current goals for measurement performance, and biological variability, due to true individual variability from dietary and other factors, were considered, first separately and then jointly. For example, a test result of 200 mg/dl, which for an adult would represent a borderline-high value, was taken to represent a range of 40 to 60 mg/dl around that value. There are clear implications of this phenomenon for dependence on single measurements of blood cholesterol, as is typical in epidemiologic studies. In some epidemiologic studies, two or more measurements have been recorded, whether in the recruitment phase of a trial or in the early follow-up experience of a long-term cohort study. The strength of association between cholesterol concentration and outcomes increases substantially when values from two occasions of measurement could be included in the analysis—an analytic approach described as adjustment for “regression-dilution bias.” Estimation of LDL Cholesterol Prominence of LDL-cholesterol and increased attention to triglycerides in risk classification broadens these concerns. This is especially so because LDLcholesterol is usually not directly measured but is calculated according to the equation:1, p III-6 LDL-C (total cholesterol) (HDL-C)
(triglycerides/5) The basis of this calculation is that total cholesterol is the sum of LDL-C, HDL-C, and VLDL-C. The latter quantity can be estimated, under appropriate conditions, as one-fifth of the triglyceride level. However, this estimate is unreliable if the triglyceride level is greater than 400 mg/dl; in this circumstance the LDL-cholesterol level cannot be estimated in this way. This poses difficulties because triglycerides increase acutely in response to fat intake, therefore meaningful determination requires a 9- to 12-hour fasting sample. It also carries substantial laboratory error. Further, the assumption that one-fifth of the triglyceride value represents VLDL-cholesterol is only approximate. Although direct measurement of LDLcholesterol has become available, it is not widely used because of issues of quality control.
46894_CH11_269_310.qxd
274
2/25/10
3:39 PM
Page 274
CHAPTER 11 ADVERSE BLOOD LIPID PROFILE
Other Measurement Issues More general considerations about cholesterol measurement in public-screening situations, including technician training, participant education and follow-up, and other aspects, were addressed by the NCEP and the American Heart Association.25 This is a helpful resource for planners of screening projects or other field studies in which cholesterol determinations will be performed. Units of measurement are either milligrams per deciliter (mg/dl) or millimoles/per Liter (mmol/L). Values of cholesterol concentration in mg/dl are equivalent to values in mmol/L obtained by multiplying mg/dl 0.02586; the reverse conversion requires multiplying mmol/L by 38.67. For triglycerides, the corresponding conversion factors are 0.0112 and 89.29.20 Values frequently noted include: Cholesterol mg/dl mmol/L 240 6.2 200 5.2 170 4.4
Triglyceride mg/dl mmol/L 1000 11.3 400 4.5 250 2.8
Finally, blood lipid determinations may be made in two kinds of samples—serum or plasma. Concentrations in plasma are slightly higher than in serum. To determine serum values equivalent to values reported as plasma concentrations requires multiplication by the adjustment factor 1.03; the reverse adjustment factor is 0.971.
DETERMINANTS Age, Sex, and Race/Ethnicity Across the life span, age-specific mean cholesterol concentrations first increase sharply from birth through infancy. They peak in the preteen years and decrease to the mid- to late teens before resuming an extended phase of increasing levels to midadulthood. A final phase of decreasing levels characterizes later adulthood. Differences in levels of total cholesterol by sex vary across these phases. The preteen peak is similar for boys and girls but, because the decrease that follows is greater for boys, and girls experience a greater increase in HDL-cholesterol, girls emerge from the teens with higher total cholesterol than boys.26 In early adulthood men have higher total cholesterol levels than women, but from the mid-50s on women have the higher levels.27 Differences by race/ethnicity in the United States are reported for adults from the National Health and Nutrition Examination Survey for non-Hispanic Whites and Blacks/African Americans, and for Mexicans (see Distribution, below). In 2003–2006,
age-adjusted mean values for persons 20–74 years were highest for Mexican men (203 mg/dl) and nonHispanic White women (also 203 mg/dl). Lowest levels were found for Blacks/African Americans. Persons with the highest levels may be on treatment to lower cholesterol, and mean values are sensitive to influences at the extremes of the distribution. Further, there may be interaction between race/ethnicity and treatment status for high cholesterol. For these reasons, comparisons of mean values of cholesterol by race/ethnicity may be misleading. Family History Family history of coronary heart disease occurring in early to midadulthood was shown by Williams and colleagues to be a simple and useful means for identifying relatives with high cholesterol.28 The MEDical PEDigree, or MED PED, program was noted in Chapter 7, “Genes and Environment,” as successful in detection of familial hypercholesterolemia in 50,000 persons within 3000 families with positive histories of coronary heart disease, from a total of 90,000 Utah families. The practice of evaluating immediate relatives of patients with known coronary heart disease is supported by this and other evidence of familial occurrence of risk factors and cardiovascular diseases, especially at midadult or earlier ages. However, it is generally believed that this practice is seldom implemented. Family history is more likely influential in practice in the area of blood lipid screening in children and adolescents.20 Recommendations from the NCEP Expert Panel on Blood Cholesterol Levels in Children and Adolescents were based on a principle of “selective screening” in which family history was the criterion for selection:29, p 545 The Panel reached consensus that a low density lipoprotein (LDL)-cholesterol value of 130 mg/dl or higher (95th percentile), when associated with family history of cardiovascular disease (CVD) or parental hypercholesterolemia, is sufficiently elevated to warrant further evaluation and probable treatment and followup. The panel deliberately targeted the family unit and the familial aggregation of CVD and/or inherited lipid problems because hypercholesterolemia in a child from such a family is of clinical significance. Children with parents and grandparents who have premature CVD often have high cholesterol levels. Thus cholesterol levels in a child are linked to familial CVD. Children are to be screened, then, if a parent or grandparent at age 55 years or younger has been found by diagnostic or interventional procedures to
46894_CH11_269_310.qxd
2/25/10
3:39 PM
Page 275
MECHANISMS 275
have coronary atherosclerosis, if a parent or grandparent at age 55 years or younger has had a documented myocardial infarction, angina pectoris, peripheral vascular disease, cerebrovascular disease, or sudden cardiac death, or if a parent (age not specified) has been found to have high cholesterol ( 240 mg/dl). Inability to obtain a family history, or presence in the child of other risk factors, could warrant screening on an exceptional basis. The implication that in the absence of a positive family history a child’s cholesterol level is not significant has been challenged, as discussed further under Prevention and Control, as follows. Dietary Imbalance, Physical Inactivity, and Obesity The essential role of diet in determining blood lipid levels is addressed in some detail in Chapter 8, “Dietary Imbalance.” It is appropriate to note again here the prediction equations of Keys and of Hegsted, linking dietary fat and cholesterol intake with the blood total cholesterol concentration:30, p 875 1) Saturated fatty acids increase and are the primary determinants of serum cholesterol, 2) polyunsaturated fatty acids actively lower serum cholesterol, 3) monounsaturated fatty acids have no independent effect on serum cholesterol and, 4) dietary cholesterol increases serum cholesterol and must be considered when the effects of fatty acids are evaluated. More limited data on low-density-lipoprotein cholesterol (LDL-C) show that changes in LDL-C roughly parallel the changes in serum cholesterol but that changes in high-density-lipoprotein cholesterol cannot be satisfactorily predicted from available data. These and other dietary influences on blood lipids and the roles of physical inactivity and obesity are addressed in the immediately preceding chapters.
MECHANISMS Blood Lipid Profile Mechanisms of metabolism and transport interact with dietary fats in the digestive tract, endogenously produced lipids, and circulating lipids in the blood. These mechanisms tend to balance the normal and the pathophysiological phenomena involving blood lipids that operate continuously, throughout the life span. Three sets of such regulatory processes have been described.6 First are those involved in transport of exogenous lipids, via the chylomicron system, from the intestine to the liver and peripheral tissues. Second is
transport of lipids synthesized in the liver, whether from the liver to peripheral tissues, as circulating free fatty acids, or from the circulation back to the liver. Third, reverse cholesterol transport—which depends on HDL metabolism—removes cholesterol from tissues such as the walls of blood vessels. Each of these mechanisms comprises multiple enzymatic reactions and molecular changes, adding great complexity to the intricate balance measured at any moment as the blood lipid profile. Beyond diet and other such factors and internal regulatory mechanisms, a number of specific conditions or disease states are associated with increased blood lipid concentrations.5 “Secondary hyperlipidemia” describes this group of lipid disorders, which are subclassified as to whether cholesterol or triglyceride concentration is primarily affected (Table 11-3). Fewer of these conditions, mainly endocrinologic disorders, cause hypercholesterolemia. The greater number causes hypertriglyceridemia and includes some of the same endocrinologic disorders as well as alcoholism, diabetes, obesity, and certain specific diseases. Several classes of drugs have similar effects, including beta-blockers and diuretics, which are commonly used in treatment of high blood pressure, heart failure, or coronary heart disease. In addition to secondary hyperlipidemia, genetically caused dyslipidemias are illustrated by several examples in Table 11-2.
Atherosclerosis How adverse blood lipid profiles lead to atherosclerosis was addressed briefly in Chapter 3. Figure 11-1 presents a detailed representation based on the review by St. Clair, who noted:31, p 16 “any hypothetical scheme of the pathogenesis of atherosclerosis will be revised as new information becomes available . . . the pathogenic scheme will also be complex, because it must account for the fact that atherosclerosis is a disease of multiple etiologies and is influenced by a variety of environmental and genetic factors.” His scheme remains useful as an overview of the processes that link blood lipids with vascular pathology. The early stages of the process are shown at the left, later ones at the right. The row of elongated cells represents the single-layer endothelium (inner lining) of an arterial wall; the space above is the lumen and circulating blood (not shown), and below is the intimal layer of the arterial wall, with both cellular and extracellular components. The earliest steps shown involve the influx of cholesterol-laden lipoproteins through the endothelial cells and into the intima. This flow is counterbalanced by cholesterol efflux, mediated by HDL. Within the intima, biochemical modifications such as oxidation of LDL may occur,
46894_CH11_269_310.qxd
276
2/25/10
3:39 PM
Page 276
CHAPTER 11 ADVERSE BLOOD LIPID PROFILE
Table 11-3
Selected Causes of Secondary Hyperlipidemia
Related to hypercholesterolemia Hypothyroidism Nephrotic syndrome Chronic liver disease (mainly primary biliary cirrhosis) Dysglobulinemia Cushing’s syndrome Hyperparathyroidism Acute intermittent porphyria
Cushing’s syndrome Glucocorticoid use Beta-blocker use Diuretic use, hypopituitarism Hypothyroidism Pancreatitis Dysglobulinemia Glycogen storage disease
Related to hypertriglyceridemia Alcoholism Diabetes mellitus Obesity Estrogen use Chronic renal failure
Lipodystrophy Acute intermittent porphyria Pregnancy Stress Uremia
Source: Reprinted with permission from AM Gotto Jr, Lipid and lipid disorders, in TA Pearson et al., Primer in Preventive Cardiology, © 1994, American Heart Association.
producing molecular forms that damage endothelial cells. These products also convert macrophages, or “scavenger” blood cells, that have migrated from the circulating blood into the intima, into macrophagederived “foam” cells (so named from their microscopic appearance, because they are filled with bubble-like aggregates of cholesterol esters). Foam cells in turn release several substances that affect endothelial cell function and stimulate growth of arterial smooth-muscle cells. Late in the process, smooth-muscle cells also may become foam cells with high-cholesterol ester content. Connections between
endothelial cells become disrupted, and changes in endothelial cell function occur. Together, these alterations promote adhesion of blood platelets, leading in turn to localized thrombosis, or clot formation. The effect of small thrombi is to increase the size of the growing atherosclerotic plaque. Large thrombi can result in occlusion of the vessel, causing blockage of blood flow to the heart muscle and consequent clinical manifestations of myocardial ischemia or infarction. On the basis of this discussion, it would be expected that a blood lipid profile with high LDL and
Figure 11-1 Schematic of Cellular and Molecular Events in Pathogenesis of Atherosclerosis. Source: Reproduced with permission from RW St. Clair, Biology of atherosclerosis, in TA Pearson et al., Primer in Preventive Cardiology, p 123, © 1994, American Heart Association.
46894_CH11_269_310.qxd
2/25/10
3:39 PM
Page 277
DISTRIBUTION 277
low HDL cholesterol would foster and accelerate development of atherosclerosis. By contrast, when a favorable balance is maintained, with optimum absolute and relative levels of the blood lipid components, the essential cause of atherogenesis is lacking. It appears that in the absence of an adverse blood lipid profile, the other identified risk factors have only weak atherogenic effects. That atherosclerosis can progress at all under this circumstance appears paradoxical. However, evidence was noted earlier that risk of coronary events continues to be reduced with reduction in LDL-cholesterol from levels well below those conventionally thought to confer increased risk. Perhaps atherogenesis is stimulated when factors are present to increase oxidation of LDL or produce other adverse molecular changes, even at relatively low concentrations of LDL-cholesterol. This understanding supports rethinking assumptions in risk estimation and target-setting for treatment based on particular levels of blood lipids.
DISTRIBUTION Prevalence and Trends United States The percentage frequencies of high serum total cholesterol concentration ( 240 mg/dl) for US adults, aged 20–74 years, are shown in Table 11-4.27 Prevalence is in this instance defined by the measured values alone, regardless whether participants reported use of medication or other interventions to reduce cholesterol. (This contrasts with the approach for high blood pressure, as discussed in Chapter 12.) As noted above, age-adjusted data are shown for all persons, for each sex group, and by sex for each of three groups self-identified by race/ethnicity. The most recent data, from the National Health and Nutrition Examination Surveys conducted from 2003 to 2006, indicate overall prevalence of 16.3%. Prevalence was slightly greater for females than for males overall but, among those who were identified as Mexican, prevalence was less for females than for males. Prevalence was least (11.2%) among Black or African American men and greatest (17.9%) among White females. Poverty level, determined by family income and family size, is represented at three levels. The stratum below 100% of the poverty level had greater prevalence (18.2%) than higher strata (16.2 to 16.5%), a level higher than for any single sex–race/ethnicity group. Age-specific data, not shown here, indicate a doubling of prevalence for
males aged 35–44 years (20.5%) in comparison with those aged 20–34 years (9.5%). For females, prevalence at ages 20–34 years (10.3%) was not doubled until ages 45–54 years (19.7%), then increased an additional 50% at ages 55–64 years (30.5%). Trends in both prevalence of high serum total cholesterol levels and mean values from 1960–1962 to 2003–2006 are also indicated in Table 11-4, although data were not available by race/ethnicity prior to 1976–1980. Overall, prevalence decreased sharply—by more than 50%—from 33.3% to 16.3%. From 1976–1980 to 2003–2006, the greatest decreases in prevalence were among Blacks or African Americans and least among Mexicans. Mean values decreased over this period in a pattern similar to that for prevalence. Because treatment of high cholesterol is not taken into account, some of the apparent changes in prevalence and mean values over time, and differences between groups, may reflect differential effects of treatment. These changes are of special interest in relation to national goals for improving cholesterol levels in the population as a whole. Objectives set by the Healthy People 2010 process, within the focus area of heart disease and stroke prevention, were to reduce prevalence of high cholesterol to 17% and mean serum total cholesterol among adults to 199 mg/dl.32 Data from 1988–1994 were taken as the baseline. The prevalence target was already reached by 1999–2002, and the overall target for the mean level was nearly attained (reached for males but not females) by 2003–2006 (Table 11-4). A different approach to assessing changes in cholesterol over time was taken by Goff and colleagues, who conducted a birth cohort analysis based on the full frequency distribution of cholesterol values reported in each of these same national surveys, from 1960–1962 to 1988–1994.33 Figure 11-2 presents the observed (left panel) and estimated (right panel) distributions by age for persons born in successive decades (legend in the right panel). For example, persons born as early as 1920 but before 1930 would have been as young as age 30 in 1960–1962 and as old as age 70 in 1988–1994. For each decade of attained age at the time of a survey, for each cohort, the distribution of total cholesterol could be described by percentile values; the figure illustrates the 50th percentile, or median, values for each of these distributions. The observed data indicate, in general, lower median values of cholesterol at every age among successively more recent birth cohorts. The estimated data represent a model that extends beyond the observed data as though each age at observation were available for each birth cohort.
216 216 217 ------------211 217 217
222 220 224 -------------------
24.4 28.9 28.9
-------------
28.6 27.9 29.1 18.7 20.7 16.4 19.9 18.7 17.7
19.7 18.8 20.5
211 213 216
213 216 211 216 209 209
203 203 206
204 206 201 204 206 204
23.5 19.3 26.5 19.4 29.0 19.6 Mean Serum Cholesterol Level, mg/dl 215 205 213 204 216 205
26.4 29.6 25.5 26.3 20.3 20.5
27.8 26.4 28.8
200 203 203
202 204 195 200 205 198
203 203 202
17.8 18.8 16.5
17.0 17.4 12.5 16.6 17.6 12.7
17.0 16.9 17.0
203 201 200
199 203 193 194 203 199
200 199 201
18.2 16.5 16.2
16.0 17.9 11.2 13.0 17.7 13.8
16.3 15.6 16.9
Sources: CDC/NCHS, National Health and Nutrition Examination Survey, Hispanic Health and Nutrition Examination Survey (1982–1984), and National Health Examination Survey (1960–1962). Data from Health, United States, 2008, Table 72, pp 314–315.
3:39 PM
-------
-------------
33.3 30.6 35.6
Percent of Population with High Serum Total Cholesterol (Greater Than or Equal to 240 mg/dl)
2/25/10
- - - Data not available. 1 Persons of Mexican origin may be of any race. Starting with 1999 data, race-specific estimates are tabulated according to the 1997 Revisions to the Standards for the Classification of Federal Data on Race and Ethnicity and are not strictly comparable with estimates for earlier years. The two non-Hispanic race categories shown in the table conform to the 1997 Standards. Starting with 1999 data, race-specific estimates are for persons who reported only one racial group. Prior to data year 1999, estimates were tabulated according to the 1977 Standards. Estimates for single-race categories prior to 1999 included persons who reported one race or, if they reported more than one race, identified one race as best representing their race. See Appendix II, Hispanic origin; Race. 2 Data for Mexicans are for 1982–1984. See Appendix I, National Health and Nutrition Examination Survey (NHANES). 3 Age-adjusted to the 2000 standard population using five age groups: 20–34 years, 35–44 years, 45–54 years, 55–64 years, and 65 years and over (65–74 years for estimates for 20–74 years). Age-adjusted estimates may differ from other age-adjusted estimates based on the same data and presented elsewhere if different age groups are used in the adjustment procedure. See Appendix II, Age adjustment. 4 Includes persons of all races and Hispanic origins, not just those shown separately. 5 Percent of poverty level is based on family income and family size. Persons with unknown percent of poverty level are excluded (4% in 2003–2006). See Appendix II, Family income; Poverty. Notes: High serum cholesterol is defined as greater than or equal to 240 mg/dl (6.20 mmol/L). Borderline high serum cholesterol is defined as greater than or equal to 200 mg/dl and less than 240 mg/dl. Risk levels have been defined by the Third Report of the National Cholesterol Education Program Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults. National Heart, Lung and Blood Institute, National Institutes of Health. September 2002. (Available from: http://www.nhlbi.nih.gov/guidelines/cholesterol/index.htm and summarized in JAMA 2001;285(19):2486–97). Individuals who take medicine to lower their serum cholesterol levels and whose measured total serum cholesterol levels are below the cut-offs for high and borderline high cholesterol are not defined as having high or borderline high cholesterol, respectively. See Appendix II, Cholesterol, serum. Standard errors for selected years are available in the spreadsheet version of this table. Available from: http://www.cdc.gov/nchs/hus.htm. Data for additional years are available. See Appendix III.
Both sexes Male Female Not Hispanic or Latino: White only, male White only, female Black or African American only, male Black or African American only, female Mexican male Mexican female Percent of poverty level:5 Below 100% 100%–less than 200% 200% or more 20–74 years, age-adjusted3 Both sexes4 Male Female Not Hispanic or Latino: White only, male White only, female Black or African American only, male Black or African American only, female Mexican male Mexican female Percent of poverty level:5 Below 100% 100%–less than 200% 200% or more
4
278
20–74 years, age-adjusted3
Serum Total Cholesterol Levels Among Persons 20–74 Years of Age or Over, by Sex, Age, Race and Hispanic Origin, and Poverty Level: United States, Selected Years 1960–1962 Through 2003–2006 [Data are based on interviews and laboratory work of a sample of the civilian noninstitutionalized population] Sex, Age, Race and Hispanic Origin,1 1960–1962 1971–1974 1976–19802 1988–1994 1999–2002 2003–2006 and Percent of Poverty Level
Table 11-4
46894_CH11_269_310.qxd Page 278
CHAPTER 11 ADVERSE BLOOD LIPID PROFILE
46894_CH11_269_310.qxd
2/25/10
3:39 PM
Page 279
DISTRIBUTION 279
Cholesterol, mg/dL
275
50th Percentile
50th Percentile
250 1890 1900 1910 1920 1930 1940 1950 1960
225 200 175 150 10
20
50 60 30 40 Attained Age, y
70
80
10
20
30
40 50 60 Attained Age, y
70
80
Figure 11-2 Observed (Left Panel) and Estimated (Right Panel) 50th Percentile Curves for Serum Total Cholesterol Concentration for the Age Range 18 Through 74 Years by Birth Cohort. In the panels depicting estimated patterns, the solid lines reflect the ranges for which data were observed and the dotted lines reflect the ranges for which the values were extrapolated from the observed data. The models used to derive these estimates included the following independent variables: age, age2 (quadratic), birth year, and age2 by birth year. Scale intervals are equal in all panels, whereas the ranges differ as appropriate to each percentile shown. To convert cholesterol from milligrams per deciliter to millimoles per liter, multiply by 0.02586. Source: Reprinted with permission from Archives of Internal Medicine, Vol 162, © 2002, American Medical Association.
high cholesterol; 12.0% of the total were on treatment; and 5.4% of the total had a total cholesterol concentration below 200 mg/dl. An update to 2005–2006 reported that the proportion of persons with high cholesterol who had it controlled had increased to 17.1%.35 Medication use was reported by 54.4% of persons told of having high cholesterol,
1.5
Prevalence, %
The resulting picture is one of a marked downward shift in the cholesterol distribution among adults in the United States, over a period of more than three decades. Changes in peak values, at ages in the 50s, were greatest, and those at the earliest adult ages were least. These changes were not subject to effects of drug treatment of high cholesterol, both because people with median-level values were not considered to require treatment and because, in any case, effective treatment was not widely available. Comparable results were found across the whole distribution, with somewhat greater decreases in the 90th and 75th than in the 50th percentile, which in the more recent years probably indicate treatment effects additional to the population-wide shift in distribution. Figure 11-3 presents the whole cholesterol distribution for persons born in 1910 (dashed line) or in 1940 (solid line) who were examined at age 50 years. Again, it is evident that the entire distribution shifted downward over the three-decade interval, illustrated by these two birth cohorts. From a public health perspective, it is of further interest to know the extent to which people with high cholesterol are aware of the condition, are receiving appropriate intervention for it, and have it controlled. Ford and others reported on such data from the 1999–2000 cycle of the National Health and Nutrition Examination Survey (NHANES):34 High cholesterol was defined as a concentration 200 mg/dl or reported use of cholesterol-lowering medication. Prevalence by this definition was 55.7%; 69.5% of this total group had had a prior cholesterol check; 35.0% of the total were aware of having
1910 1940
1.0
0.5
0 100
150 200 250 300 350 Serum Total Cholesterol Concentration, mg/dL
400
Figure 11-3 The Estimated Distributions of Serum Total Cholesterol Concentration for 50-Year-Old Persons Born in 1910 and 1940. The models used to derive these estimates included the following independent variables: age, age2 (quadratic), birth year, age by birth year, and age2 by birth year. The prevalence is the estimated proportion of people with an exact cholesterol concentration. To convert cholesterol from milligrams per deciliter to millimoles per liter, multiply by 0.02586. Source: Reprinted with permission from Archives of Internal Medicine, Vol 162, © 2002, American Medical Association.
46894_CH11_269_310.qxd
280
2/25/10
3:39 PM
Page 280
CHAPTER 11 ADVERSE BLOOD LIPID PROFILE
a marked increase from 39.1% in 1999–2000. Another report, specific to LDL-cholesterol, indicated that nearly 30% of adults were eligible for treatment under ATP III guidelines—16% by “therapeutic lifestyle change,” or TLC, and 13.4% by drugs.36 But only 43.7% of those found with high LDL-cholesterol had been diagnosed previously, and only 77.4% of those—or 33.8% of the total group— reported being told to undertake TLC. Regarding elevated triglycerides, an evaluation of detection and treatment based on NHANES 1999–2004 suggested that recommended lifestyle changes—the principal approach at all but extreme elevations—are implemented infrequently, despite a high prevalence of overweight or obesity, physical inactivity, and smoking among those affected.37 Information regarding distributions of total cholesterol levels in children and adolescents is available from the 2005–2006 cycle of NHANES.38 Mean values were reported by sex and race/ethnicity for two age groups. For age 4 to 11 years, mean values ranged from 160.8 to 166.5 mg/dl, being lowest for Mexican American girls and highest for both Black and White non-Hispanic boys. For age 12–19 years, mean values ranges from 154.5 to 165.0 mg/dl, being lowest for
non-Hispanic White boys and highest for non-Hispanic White girls. Prevalence of total cholesterol levels 200 mg/dl was 9.6% of adolescents aged 12–19 years. For the reasons discussed above regarding age patterns of blood lipids in childhood and adolescence, these observations as reported by broad age groups give only limited insight to the dynamic changes in blood lipids during this period of life. The data are too sparse to permit reliable estimates of mean values by year of age for each sex–race/ethnicity group, as would be more informative.26 However, contrasting even these data for boys reveals a shift in distributions between the two age groups: for non-Hispanic Whites, from a mean value of 166.5 to 154.5 mg/dl; for non-Hispanic Blacks, from 166.5 to 161.7 mg/dl; and for Mexican Americans, from 162.3 to 158.2 mg/dl. Prevalence would have been considerably higher for the younger age group than that found at age 12–19 years, on the basis of a constant cut-point for all ages. Europe and Global Differences between populations in distributions of total cholesterol concentration were recognized several decades ago and were emphasized especially by Keys (Figure 11-4).39 In the absence of central stan-
Figure 11-4 Mean Values of Serum Cholesterol Concentration by Age in Clinically Healthy Men in Selected Populations. Source: Reprinted with permission from A Keys, Serum Cholesterol and the Question of Normal, in Multiple Laboratory Screening, ES Benson and PE Strandjord, eds, p 169, © 1969 Academic Press, Inc.
46894_CH11_269_310.qxd
2/25/10
3:39 PM
Page 281
DISTRIBUTION 281
dardization of laboratory methods at that time, comparison of these absolute values between populations is at best tenuous. Here the differences are very large, however, and even if not precise in terms of absolute differences in cholesterol levels at any given age, the overall distributions are clearly distinct between Japan, at the lower extreme, and Germany, New Zealand, and Minnesota, all at higher levels. In each of these populations, increases with age are apparent from the 20s to the 40s or beyond. Decreasing values at older ages in each of the populations, except those in Japan, are also notable. This pattern could reflect selective mortality of older persons with the highest cholesterol concentrations or cohort effects in which the oldest persons escaped the increased cholesterol levels that were experienced by younger ones. In Minnesota, data were also obtained for women. Their age curve for total cholesterol concentration also fell at the oldest ages, but with a peak in the age curve 15 years later than for men. Knuiman and colleagues studied a wider range of populations, in 1980, including Ghana, Ivory Coast, and Nigeria in Africa; Surinam in South America; Pakistan and the Philippines in the Pacific and Asia; and three groups in Europe (East and West Finland and Hungary).40 With a centralized laboratory to assure comparability, they found total cholesterol concentrations to vary widely among adults age 33–38 and 43–48 years across these populations. For example, total cholesterol concentration ranged from about 116 mg/dl in Nigerian men to 247 mg/dl in East Finnish men. A twofold range in HDL-cholesterol concentrations was also found, and HDL-cholesterol as a proportion of the total varied from 15–18% in Pakistani men to 29–32% in Ghanaians. Body mass index was positively associated with total cholesterol concentration and inversely related to HDL-cholesterol concentration across these populations. The most informative comparison of cholesterol levels, including awareness, treatment, and control, beyond the United States was provided by the WHO MONICA Project.41 Final risk-factor surveys in the 10-year project were conducted between 1989 and 1997, among adults aged 35–64 years, in 32 populations in 19 countries on three continents. Countries included were mainly in Europe, but China, Australia, New Zealand, Canada, and the United States were also represented. Hypercholesterolemia was defined for this analysis as 6.5 mmol/L (250 mg/dl) or reporting use of prescribed lipid-lowering medication within 2 weeks prior to the survey. Other cut-points were also displayed, as shown in Figure 11-5. The populations are ranked, separately for men and women, by decreasing prevalence of hypercholes-
terolemia defined as concentrations 5.0 mmol/L (195 mg/dl). In the figure, the solid bars represent prevalence at 6.5 mmol/L; the dotted bar adds prevalence due to positive history of recent lipidlowering medication use; the cross-hatched bar adds prevalence at 6.2 mmol/L (240 mg/dl); and the stippled bar adds prevalence attributed to levels 5.0 mmol/L. Total prevalence was greater for men than for women in this age range. At 6.5 mmol/L, the range of prevalence was from 3 to 53% among men and from 4 to 40% among women across populations. For each of the 32 populations, the mean serum total cholesterol concentration was reported, as well as the prevalence of values 6.5 mmol/L or reporting medication use; the proportion of the prevalent subgroup reporting treatment with drugs alone, diet alone, both, or neither; and the proportion of the prevalent subgroup defined as controlled, that is, having total serum cholesterol levels below 6.5 mmol/L. From 0 to 41% of those with hypercholesterolemia were treated with drugs alone; from 0 to 46% were treated with diet alone; from 0 to 48% were treated with both; and from 0 to 100% reported receiving no treatment. The proportion controlled, defined as 6.5 mmol/L, ranged from 0 to 100%, the median being about 55%. These observations present stark evidence of the gap between widely recommended goals for control of hypercholesterolemia and actual practice in the first comparative analysis of this kind. In childhood and adolescence, too, international comparisons have been made by various researchers, as illustrated in Figure 11-6.42 Again, absolute values are not comparable between these independent surveys, but overall patterns by age are generally similar. Very sharp increases occur from birth to age 1 year, as also found in other data, and mean values by age appear rather stable within each population across the school-age years. However, closer evaluation reveals a systematic pattern of a decrease and subsequent increase in total cholesterol concentration in adolescence in each population, although with some population differences in ages at the inflection points in the curve. This pattern of variation in total cholesterol concentration with age, and its difference by about 1 year in timing between sexes, is sufficient to raise concern about a fixed value as the screening criterion in youth, as noted previously.43 Knuiman, who surveyed adults in multiple countries, similarly surveyed populations of boys aged 13 years in 16 countries in Africa, Europe, and South Asia.44 Again, a central laboratory was used. Standardized determinations of total and HDL-cholesterol levels were carried out for all populations. Marked population differences were found
46894_CH11_269_310.qxd
282
2/25/10
3:39 PM
Page 282
CHAPTER 11 ADVERSE BLOOD LIPID PROFILE
Men
Women
SWI-TIC SWI-VAF GER-BRE SWE-NSW YUG-NOS CZE-CZE FRA-STR ICE-ICE UNK-GLA DEN-GLO BEL-GHE BEL-CHA UNK-BEL FRA-LIL FRA-TOU ITA-BRI LTU-KAU GER-EGE ITA-FRI AUS-NAW CAN-HAL POL-WAR SWE-GOT SPA-CAT AUS-PER POL-TAR RUS-NOI RUS-MOI USA-STA RUS-MOC RUS-NOC CHN-BEI
GER-BRE SWI-TIC SWI-VAF BEL-CHA CZE-CZE YUG-NOS SWE-NSW UNK-GLA LTU-KAU ICE-ICE BEL-GHE FRA-STR ITA-BRI FRA-LIL CAN-HAL DEN-GLO UNK-BEL FRA-TOU GER-EGE POL-WAR AUS-NEW ITA-FRI POL-TAR SPA-CAT SWE-GOT AUS-PER RUS-MOC RUS-MOI RUS-NOI RUS-NOC USA-STA CHN-BEI
0
20 80 40 60 Prevalence (%) of hypercholesterolemia 6.5 mmol/l
100
0
6.5 mmol/l or treatment
20 40 60 80 Prevalence (%) of hypercholesterolemia
6.2 mmol/l
100
5.0 mmol/l
Figure 11-5 The Age-Standardized Prevalence of Hypercholesterolemia for Men and Women in Age Group 35–64, WHO MONICA Project. Source: Reprinted with permission from International Journal of Epidemiology, Vol 34. H Tolonen, U Keil, M Ferrario, A Evans, for the WHO MONICA Project, Prevalence, Awareness and Treatment of Hypercholesterolemia in 32 Populations: Results from the WHO MONICA Project. pp 181–192 © International Epidemiological Association 2004.
at this age, not attributable to laboratory variation but perhaps due in part to differences in growth tempo across populations, in relation to the preadolescent peak described above. Some insight into population differences in total cholesterol levels is provided by Project HeartBeat!, a mixed longitudinal follow-up study in which children and adolescents were examined in Shibata, Japan, and The Woodlands and Conroe, Texas (Figure 11-7)43 (Darwin R. Labarthe, unpublished data, 2009). All determinations were conducted in CDCstandardized laboratories. Unexpectedly, given prior knowledge of exceptionally low cholesterol levels among adults in Japan, total cholesterol levels were higher for Japanese than American children and adolescents across the age range from 8 to 18 years. For both girls and boys, however, the difference was entirely attributable to greater HDL-cholesterol levels in Japan than the United States—and, in both popu-
lations, girls had higher HDL-cholesterol concentrations than did boys, from the early teens. Disparities Differences in cholesterol levels by age, sex, and race/ethnicity were described above. Whatever the underlying causes of these differences, once hypercholesterolemia is recognized, appropriate action to control it is expected under widely recognized guidelines. However, the report by Ford and others of an overall rate of 17% control at the level of 5.2 mmol/L in 2005–2006 also indicated marked variation in this indicator within the adult population.34 Control was especially poor among Mexican Americans— being 4.6% among men and 9.3% among women. Older persons, Whites or African Americans, and men had better rates of control than younger persons, Mexican Americans, and women. The benefits of es-
46894_CH11_269_310.qxd
2/25/10
3:39 PM
Page 283
RELATION TO RATES AND RISKS 283
Cholesterol (mg/dl) 200
150
100
North America (Black)
Finland (White)
Japan (Japanese)
Native American (Pima Indians)
50 0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Age, Years Figure 11-6 Mean Values of Serum Total Cholesterol Concentration Among US Black, White, Japanese, and Native American Male Children, and Adolescents, Aged 0–19 Years. Source: Reprinted with permission from D Labarthe, B O’Brien, and K Dunn, International Comparisons of Plasma Cholesterol and Lipoproteins, in Hyperlipidemia in Childhood and the Development of Atherosclerosis, CL Williams and EL Wynder, eds, Vol 623, p 117, © 1991, Annals of the New York Academy of Sciences.
tablished guidelines are clearly not reaching all groups equally, nor any group to the intended degree.
RELATION TO RATES AND RISKS Population Differences The question of how the blood lipid profile relates to population differences in risk of coronary heart disease or other complications of atherosclerosis must be addressed mainly in relation to total cholesterol concentration. This is because at the inception of the long-term follow-up studies contributing such information only total cholesterol determination was practical and considered important. This is the case, for example, with the Seven Countries Study, which is uniquely valuable for such population comparisons.45 The 10-year mortality experience in that study shows the relation between coronary heart disease death rates and median serum total cholesterol
concentration for each population across the 16 cohorts of men age 40–59 years at entry, in the late 1950s to mid-1960s (Figure 11-8). Median cholesterol values ranged from about 160 mg/dl to 265 mg/dl, and 10-year coronary mortality varied from 1/1000 or less in Crete (K) to about 70/1000 in East Finland (E). The resulting regression equation indicates the positive coefficient for the relation between cholesterol concentration and coronary mortality. A correlation coefficient calculated with adjustment for age, systolic blood pressure, and smoking history was 0.82; the square of this value indicates the proportion of variation in rates among populations that is accounted for by median cholesterol concentration, approximately 67%. As is evident in this figure and as noted by Keys, there was little relation with population differences in mortality when cholesterol concentrations were below 200 or 210 mg/dl. In general, however, within-population relationships were strong at levels above 200 mg/dl, except in the lowestrate populations.
46894_CH11_269_310.qxd
284
2/25/10
3:39 PM
Page 284
CHAPTER 11 ADVERSE BLOOD LIPID PROFILE
4.9
Total Cholesterol (mmol/l)
4.7 Japanese girls
4.5
Japanese boys
4.3 4.1
US white girls
3.9 US white boys
3.7 3.5 3.3 7
8
9
10
11
12 13 Age (year)
14
15
16
17
18
HDL-Cholesterol (mmol/l)
1.7 1.6
Japanese girls
1.5
Japanese boys
White girls
1.4 White boys
1.3 1.2 1.1 1.0 7
8
9
10
11
12 13 Age (year)
14
15
16
17
18
Figure 11-7 Concentrations of Total Cholesterol and HDL-Cholesterol Among Japanese and US White Children and Adolescents, by Age, from 8 to 18 Years, Project HeartBeat!. Source: Unpublished data from Project Heartbeat!.
Further follow-up of the Seven Countries Study cohorts to 25 years permitted more detailed analysis, given greater numbers of events, especially when cohorts were grouped on the basis of geography, culture, and patterns of interim change in cholesterol concentration.46 Figure 11-9 illustrates the results of analysis by quartiles of baseline serum total cholesterol concentration. Coronary mortality increased with baseline cholesterol concentration in every group of populations except Japan, especially steeply for the higher observed cholesterol values. Further analysis took advantage of the repeated cholesterol determinations in most of the cohorts, at a 5-year follow-up
examination, to estimate and adjust for misclassification due to measurement variability (or regression dilution bias, discussed previously). This latter analysis showed, for all cohorts taken together, an overall average increase of 17% in coronary mortality for an increment of 20 mg/dl in median cholesterol concentration (relative risk 1.12, 95% confidence interval 1.09–1.16; risk estimate after adjustment, 1.17). The figure also shows wide population differences in mortality at fixed levels of cholesterol concentration, which overlap across regions at levels from between 175 mg/dl and 200 mg/dl to between 250 mg/dl and 275 mg/dl. The range in mortality for a fixed choles-
46894_CH11_269_310.qxd
2/25/10
3:39 PM
Page 285
RELATION TO RATES AND RISKS 285
E
Y = 10-Year Coronary Deaths per 1,000
70
60
50 R 40 N 30
20 S Z 10 U
V
T
W
I B
C
Y = –66 + 0.43X G
M
r = 0.80
D K
150
200
250
X = Median Cholesterol, mg/dl of Serum
Note: B, Belgrade; C, Crevalcore; D, Dalmatia; E, East Finland; G, Corfu; I, Italian Railroad; K, Crete; M, Montegiorgio; N, Zutphen; R, American Railroad; S, Slavonia; T, Tanushimaru; U, Ushibuka; V, Velika Krsna; W, West Finland; Z, Zrenjanin.
Figure 11-8 Ten-Year Coronary Death Rates and Median Serum Cholesterol Concentration. Source: Reprinted with the permission of the publisher from Seven Countries by Ancel Keys, Cambridge, Mass: Harvard University Press, © 1980 by the President and Fellows of Harvard College.
terol concentration was threefold or greater and indicates the collective influence of factors other than age, smoking, and systolic blood pressure, for which adjustment was made in the analysis. Individual Differences Total cholesterol concentration was also measured in each of the early cohort studies included in the US Pooling Project, with results as shown in Table 11-5.47 Among 8274 men age 40–59 years at entry, 647 events were recorded over 8.6 years of follow-up. The data are presented first for the standard pool of the five most comparable studies, together and separately (Albany, Chicago Gas, Chicago Western Electric, Framingham, and Tecumseh), and then for the three remaining studies (Los Angeles, Minnesota executives, and Minnesota Railroad workers). For each quintile category of baseline serum total cholesterol concentration, the incidence of first major coronary events is expressed as the ratio (times 100) of that rate to the overall rate in the total group.
Incidence below the population average results in a standardized incidence ratio less than 100; greater than average incidence results in a standardized incidence ratio greater than 100. For example, the standardized incidence ratio for the lowest quintile of total cholesterol concentration (below 195 mg/dl) was 72, whereas that for the highest quintile (greater than 268 mg/dl) was 158. It was noted that for several of the cohorts, including all of those in Pool 5, the lowest incidence ratio was in the second, not the first, quintile of cholesterol concentration. Therefore, the overall risk ratios were defined arbitrarily by relating incidence for quintile group V to that for quintile groups I and II combined. This risk ratio was 2.4 for Pool 5 and varied from 1.5 to 4.9 among studies. The 95% confidence intervals, having lower limits greater than 1 in all but one study, supported the presence of an association between serum cholesterol and coronary event rates. On the basis of the follow-up of men screened for the Multiple Risk Factor Intervention Trial, much
46894_CH11_269_310.qxd
286
2/25/10
3:39 PM
Page 286
CHAPTER 11 ADVERSE BLOOD LIPID PROFILE
35
Northern Europe Southern Europe, Mediterranean United States
30
Serbia Southern Europe, Inland Japan
CHD Mortality Rates, %
25
20
15
10
5
0 2.60 (100)
3.25 (125)
3.90 (150)
4.50 (175)
5.15 (200)
5.80 (225)
6.45 (250)
7.10 (275)
7.75 (300)
8.40 (325)
9.05 (350)
Serum Total Cholesterol, mmol/l (mg/dl)
Figure 11-9 Twenty-Five Year Coronary Death and Quartiles of Serum Cholesterol Concentration, Adjusted for Age, Cigarette Smoking, and Systolic Blood Pressure. Source: Reprinted from WMM Verschuren et al., Journal of the American Medical Association, Vol 274, p 131, © 1995.
more detailed analysis of this relation was possible.48 This very large study population of 361,662 men with, on average, 6 years of follow-up could be grouped in 20 units of 5 percentile levels, in contrast to the 5 units of 20 percentiles in the Pooling Project example. This offers greater resolution in examination of risk gradients, as shown in Figure 11-10. From the extreme values of 150 to nearly 300 mg/dl in cholesterol concentration, an exponential pattern of increasing risk was observed. Minimum risk was at the lowest levels, in contrast to the Pooling Project findings. The risk ratio between the highest and lowest of the 20 strata would be greater than 4 on the basis of this analysis. A study of the longer-term prediction of cardiovascular disease occurrence beginning at an average age of 22 years at baseline was reported from the 27- to 42-year follow-up of medical students in Johns
Hopkins University.49 The predictive value of cholesterol concentration was again demonstrated, although event rates increased sharply only after 20 years of follow-up. As shown in Figure 11-11, the lowest quartile group of baseline cholesterol concentration (118–172 mg/dl) was found to have the lowest cardiovascular disease incidence rates, still reaching 10%, whereas rates for the highest quartile group (209–315 mg/dl) reached nearly 40%. The studies cited above pertain mainly to total cholesterol concentration in men, chiefly White men of middle age. From studies in women, such as the Framingham Heart Study, individual risks in relation to blood lipids are in some ways similar to those in men but also importantly different.50,51 The age course of the blood lipid profile differs in women, with total and LDL-cholesterol increasing especially from age 40 to 60 years. VLDL-cholesterol and
MI-EX 100 70 (64) (78) (117) (117) (189) ( ) ( ) ( ) 283 4,008 28
LA 100 (42) (37) (46) 116 73 143 ( ) ( ) ( ) 1,104 10,137 72
100 49 (47) 50 77 96 194 4.0 3.4 7.6 2,551 12,484 112
MI-RR
Source: Reprinted with permission from Journal of Chronic Diseases, The Pooling Project Research Group, Vol 31, p 230, © 1978, Elsevier Science, Inc.
3:39 PM
Note: ( ), Based on fewer than 10 first events; ALB, Albany Civil Servants; CH-GAS, Chicago Gas Company; CH-WE, Chicago Western Electric Company; FRAM, Framingham; TECUM, Tecumseh, Michigan; LA, Los Angeles Civil Servants; MI-EX, Minnesota Businessmen; MI-RR, Minnesota Railroad Workers.
Serum Cholesterol: Parameters of the Bivariate Model for First Major Coronary Events, Pooling Project Study Group Pool 5 ALB CH-GAS CH-WE FRAM TECUM Quintile and Level (mg/dl) Standardized Incidence Ratio All All 100 100 100 100 100 100 I II 218 66 70 79 60 62 49 I 194 72 72 100 62 74 (10) II 194–218 61 67 61 57 50 (83) III 218–240 78 72 89 70 88 (56) IV 240–268 129 129 124 99 160 145 V 268 158 177 118 159 167 242 Risk ratio: V/(I II) 2.4 2.5 1.5 2.7 2.7 4.9 95% confidence interval: Low 1.9 1.7 0.9 1.7 1.7 2.0 High 2.9 3.8 2.4 4.6 4.0 13.1 Number of men at risk 8,274 1,765 1,264 1,980 2,130 1,135 Person-years of experience 70,781 16,878 11,064 16,505 19,480 6,854 Number of first events 647 156 123 142 177 49
2/25/10
Table 11-5
46894_CH11_269_310.qxd Page 287
RELATION TO RATES AND RISKS 287
46894_CH11_269_310.qxd
288
2/25/10
3:39 PM
Page 288
CHAPTER 11 ADVERSE BLOOD LIPID PROFILE
Figure 11-10 Serum Cholesterol Concentration and 6-Year Mortality from Coronary Heart Disease (CHD). Multiple Risk Factor Intervention Trial Screenees Aged 35–57 Years. Each point represents the median value for 5% of the population. Source: Reprinted with permission from MJ Martin et al., Serum cholesterol, blood pressure, and mortality: implications from a cohort of 361,662 men. Lancet, Vol 2, pp 933–936, © 1986, The Lancet, Ltd.
triglycerides also increase during this period, and HDL-cholesterol declines. The resulting adverse change in the HDL-/LDL-cholesterol ratio, with individual values ranging from greater than 7 to less than 5, was associated with a sharp gradient of 8year coronary risk in the Framingham Heart Study. Increased triglyceride levels also confer increased risk, as shown in a 19-year follow-up study of women aged 39–64 years, in Sweden.52 Over the period of
that study, total cholesterol and smoking prevalence both decreased among women, but triglyceride concentrations increased and offset the benefits of the other risk factor changes. It has been suggested that, whereas in general both LDL- and HDL-cholesterol relate to coronary risk in women as in men, the inverse risk gradient for HDL-cholesterol is stronger for women and that a given level of LDL-cholesterol may be less atherogenic for women.53
46894_CH11_269_310.qxd
2/25/10
3:39 PM
Page 289
RELATION TO RATES AND RISKS 289
0.40 0.35
209–315 mg/dl 190–208 mg/dl 173–189 mg/dl 118–172 mg/dl
Cumulative Incidence of Cardiovascular Disease
0.30 0.25
P<0.001 0.20 0.15 0.10 0.05 0.00 0
5
10
15
20
25
30
35
40
128 131 155 140 554
61 62 75 78 276
7 15 12 13 47
Years of Follow-Up
Quartile (mg/dl) 118–172 173–189 190–208 209–315 Total
250 258 254 255 1017
248 256 251 251 1006
245 254 248 243 990
240 250 240 235 965
234 243 228 222 927
217 216 208 196 837
Note: To convert values for cholesterol to mmol/l, multiply by 0.02586. The numbers below the figure are the numbers of men included in the analysis at each point.
Figure 11-11 Forty-Year Cumulative Incidence of Cardiovascular Disease in 1017 White Men Following Baseline Serum Cholesterol Determination at Median Age 22. Source: Reprinted with permission from MJ Klag et al., Serum cholesterol in young men and subsequent cardiovascular disease, The New England Journal of Medicine, Vol 328, No 5, pp 313–318, © 1993, Massachusetts Medical Society.
The relation of blood lipids to risk of cardiovascular diseases is not limited to cholesterol or triglyceride concentrations or to coronary heart disease. Total cholesterol concentration was related to risk of thromboembolic stroke, for example, in those screened in the Multiple Risk Factor Intervention Trial as presented in Chapter 5.54 Lp(a) concentration has been found to predict coronary heart disease in men and women, as illustrated in a cohort study from Rochester, Minnesota.55 When nearly 10,000 men and women were followed for 15 years, women exhibited strong (nearly threefold) gradients of risk of both coronary heart disease and stroke with in-
creasing baseline levels of Lp(a). Corresponding gradients were weaker (1.5- to 1.7-fold), though statistically significant, for men. Other cardiovascular conditions have also been found to be associated with Lp(a) concentration: carotid atherosclerosis, peripheral arterial disease, coronary artery bypass graft occlusion, and others.7 ApoE genotype predicts coronary artery disease in men and women, according to a meta-analysis of 14 clinical and angiographic studies.8 The three alleles (ε2, ε3, and ε4) imply six possible genotypes. Combinations that include the ε4 allele, compared with the ε3/ε3 genotype, resulted in an overall relative risk of coronary heart disease of
46894_CH11_269_310.qxd
290
2/25/10
3:39 PM
Page 290
CHAPTER 11 ADVERSE BLOOD LIPID PROFILE
1.38 (confidence interval [CI] 1.22–1.57) for men and 1.82 (CI 1.30–2.54) for women. Global Health Impact The INTERHEART study, cited in previous chapters, was a case-control study of myocardial infarction in 52 countries, “representing every inhabited continent.”56 Blood lipids were assessed by ratio of apolipoproteins B and A1 (ApoB/ApoA1) as a nonfasting approach to represent non-HDL- and HDLcholesterol, as well as total and HDL-cholesterol. In the overall study population, for the highest versus lowest quintile of the ApoB/ApoA1 distribution, the odds ratio for cases versus controls when adjusted for all other factors was 3.25, stronger than any other risk factor. The overall population attributable fraction was also greatest among all risk factors, 49.2%. The population attributable fractions for lipids among regional groups of the study population ranged from 41% in South America and China to 74% in Africa for men and from 43% in China and Australia/New Zealand to 74% in Africa for women. The authors noted:56, pp 945–946 . . . most populations in the world (at least urban) have lipid abnormalities, which increase the risk of myocardial infarction. Since ApoB/ ApoA1 ratio was the most important risk factor in all geographic regions in our study, a substantial modification of its population distribution is important for worldwide reduction of myocardial infarction. The global dimensions of adverse blood lipid profiles are further indicated by data from the Global Burden of Disease and Risk Factors Study on the basis of estimated distributions of total cholesterol concentration throughout the world.57 “High cholesterol” in this analysis is the “usual level” in a population, and the reference value is the “theoretical-minimum-risk exposure distribution.” For total cholesterol concentration, this value is taken to be 3.8 mmol/L (147 mg/dl) on the basis of “the lowest levels at which metaanalyses of cohort studies have characterized doseresponse relationships.” Table 11-6 presents, for both ischemic heart disease and cerebrovascular disease (stroke), the fraction of deaths attributable to high cholesterol in the population (population attributable fraction, PAF) for each of six geographic regions, two economic strata, and the world. Additional data from the study include PAFs for years of life lost and disability-adjusted life years lost (DALYs) and the numbers of persons at each population level estimated to be affected by death or disability. Total deaths at-
tributable to cholesterol were 3.2 million for ischemic heart disease and 0.7 million for stroke. PAFs are shown for both males and females, from age 30–44 years to 80 years. For the world total, 45% of ischemic heart disease and 13% of stroke were attributable to this factor. The range for ischemic heart disease was from 15% to 55% and for stroke, from 5% to 20%, in both cases lowest in sub-Saharan Africa and highest in Europe and Central Asia. PAFs were higher for high-income than for low- and middle-income countries overall, for both ischemic heart disease (52 versus 43%) and stroke (17 versus 12%). In general, the attributable fractions were greater at younger ages for both conditions; other factors such as blood pressure increase in prominence at later ages.
RELATION TO OTHER FACTORS Nature of the Relations The adverse blood lipid profile occurs largely in consequence of dietary imbalance, jointly with physical inactivity and obesity. It is also commonly associated with the other major diet-activity-obesity-related conditions, especially high blood pressure and diabetes. The concurrence of these and other risk factors reflects a constellation of circumstances of life that affect a growing proportion of the world population. They coexist within populations and cluster within individuals. This aspect of the major determinants of the cardiovascular diseases will be a recurring note in subsequent chapters. Implications for Prevention and Control It is a reasonable argument that ultimately these factors should be considered together, and not separately, in part because of shared underlying determinants and also because of multiple benefits of their amelioration on a population level. This logic supports the concept of “absolute risk,” a measure of estimated probability that an individual will experience a coronary––or, more broadly, cardiovascular––event within a defined period. The blood lipid profile may contribute one or more components to such a calculation, depending on the model used. The product is a risk estimate that guides policy for resource allocation, or treatment decisions for individuals, on the basis of the aggregate risk from multiple factors. A transition toward this approach is in progress, but conventional single-factor guidelines and practices remain and are the focus of the discussion that follows.
9 27 23 19 18 4 15 28 16
12 28 22 22 17 10 17 27 18
6 18 16 14 12 1 10 19 11
11 24 19 18 14 7 15 23 16
3 10 8 7 6 0 5 10 6
6 14 10 10 9 4 9 13 10
6 19 16 13 13 2 11 16 11
Source: Data from Global Burden of Disease and Risk Factors, edited by AD Lopez et al., © 2006. The International Bank for Reconstruction and Development/The World Bank.
5 23 18 16 18 3 12 24 13
8 23 20 17 16 3 13 24 14
9 21 16 16 15 8 14 17 14
8 20 16 15 14 5 12 17 13
6 25 22 18 16 2 12 27 14
11 29 23 22 23 10 17 28 18
Risk Factor: Disease: PAF of Mortality (%) East Asia and Pacific Europe and Central Asia Latin America and the Caribbean Middle East and North Africa South Asia Sub-Saharan Africa Low- and middle-income countries High-income countries WORLD
12:26 PM
High Cholesterol Cerebrovascular Disease
32 55 49 47 43 15 43 52 45
All
2/26/10
Risk Factor: Disease:
Contributions of High Cholesterol to Mortality Due to Ischemic Heart Disease and Cerebrovascular Disease, by World Region High Cholesterol Ischemic Heart Disease 30–44 years 45–59 years 60–69 years 70–79 years 80+ years Total Region Male Female Male Female Male Female Male Female Male Female Male Female PAF of Mortality (%) East Asia and Pacific 49 41 44 51 29 41 21 35 15 32 27 37 Europe and Central Asia 86 83 74 77 54 64 34 55 40 54 54 57 Latin America and the Caribbean 84 76 70 69 51 56 41 48 34 42 49 50 Middle East and North Africa 75 70 62 67 45 54 36 46 32 43 45 51 South Asia 71 76 59 70 41 45 32 38 26 38 42 46 Sub-Saharan Africa 10 18 14 36 8 28 2 20 0 18 7 24 Low- and middle-income countries 69 65 58 64 41 48 32 43 27 43 41 47 High-income countries 89 84 76 76 56 63 45 53 42 52 51 54 WORLD 71 66 61 65 43 49 34 45 32 46 43 48
Table 11-6
46894_CH11_269_310.qxd Page 291
RELATION TO OTHER FACTORS 291
46894_CH11_269_310.qxd
292
2/25/10
3:39 PM
Page 292
CHAPTER 11 ADVERSE BLOOD LIPID PROFILE
PREVENTION AND CONTROL Prevention of adverse blood lipid profiles—or promotion of the optimum profile—depends on establishment and maintenance of favorable patterns of diet and physical activity, body weight, and body composition throughout the life span, beginning in childhood, or earlier. This requirement implies conditions of living in which healthy foods are available, are affordable, and are consumed in households and other community settings—such as schools, workplaces, and restaurants. Promotion of healthy activity entails assuring social and environmental conditions that are similarly supportive of cardiovascular health. Such conditions supporting favorable diet and activity patterns tend to assure community-level, population-wide maintenance of cardiovascular health from healthy beginnings early in life. They also enable prevention at the individual level by “making healthy choices the easy choices,” as commonly phrased. Control of adverse blood lipid profiles implies taking corrective or remedial action in response to increased risk. As for prevention, control has both community- or population-level as well as individual-level aspects. A sample survey may indicate a population distribution of blood lipid components that warrants public health concern and stimulates targeted policy and environmental changes at the community or higher levels. Or, an individual clinical encounter or organized screening program may identify one or many persons for whom evaluation and a management plan are indicated. These ideas are based on evidence establishing the role of lipids in causation of cardiovascular diseases; the main determinants of blood lipid profiles; the magnitude and prevalence of increased risk attributable to them; and their amenability to interventions on lifestyle, medication, or both. How has this evidence been translated into policy and practice for prevention and control? This question can be addressed separately at individual, community, and global levels. Individual Measures Individual-level interventions have been evaluated in a great many clinical trials, conducted over several decades, which vary in specific purposes and outcomes. Diet, drugs, and combinations of these and other interventions have been tested in studies of different size and complexity in many countries. Trials have variously included adults with a prior cardiovascular event (secondary prevention trials), those without known cardiovascular disease (primary prevention trials), or less often persons with angio-
graphically identified coronary artery disease (regression/progression of atherosclerosis trials). Still less common are trials in children or adolescents with high total or LDL-cholesterol (cholesterol-reduction trials). Trials addressing multiple risk factors in individuals or in communities are reviewed in Chapter 21, “The Case for Prevention.” Several topics specific to trials for reducing cholesterol, or improving the blood lipid profile, are addressed here. Early Trials Among the earliest examples were drug trials conducted among groups at very high risk: men with previous myocardial infarction. Recurrent infarction and death were the outcomes against which treatment was evaluated. The Coronary Drug Project (CDP) was a prototype of a multicenter cardiovascular disease prevention trial and the most extensive such study undertaken up to its initiation in 1965.58 It involved 53 clinical centers and recruitment of 8341 participants for random allocation either to one of its five active treatment arms or to the placebo control group. High- and low-dose conjugated estrogens, clofibrate, dextrothyroxine, and niacin were known to have cholesterol-lowering effects, but their efficacy and safety for coronary heart disease prevention were still in very early stages of investigation. The high-dose and low-dose estrogen treatments were both discontinued in the course of the trial because of excess adverse events or lack of benefit. At the scheduled completion of the trial, neither clofibrate nor niacin was found to confer benefit in reducing mortality. However, after extended posttrial follow-up, there was a significantly lower mortality among niacin-treated patients than placebo controls. Interestingly, niacin has recently been recommended as an adjunct to statins because it raises HDL-cholesterol and reduces triglycerides.59 Several of the early trials in primary prevention of coronary heart disease are described as to their design (Table 11-7) and results (Table 11-8) on the basis of a 1989 review in Diet and Health, by the National Academy of Sciences, discussed earlier.4 These studies were conducted in Europe and the United States in either worksite or clinical populations. Several hundred to many thousands of individuals participated. Drug trials were double blinded, lending rigor to outcome evaluation, although this was not feasible for dietary or other behavioral interventions. Starting levels of cholesterol concentration were high relative to current trials. In Table 11-8, the Finnish Mental Hospital Study results are presented separately for women and men, and all studies are arrayed in increasing order of percentage net
46894_CH11_269_310.qxd
2/25/10
3:39 PM
Page 293
PREVENTION AND CONTROL 293
Table 11-7
Randomized Trials on Primary Prevention of Coronary Heart Disease (CHD), Including Cholesterol Lowering by Diet or Drugs: Design Double Interventions and Targets Study Study Population Blind of Interventions Göteborg Multifactor Trial 20,015 men ages 47 to 55 No Diet, cigarette smoking, high blood pressure WHO Multifactor Trial 66 employed groups totaling 49,781 No Diet, cigarette smoking, high men ages 40 to 59 blood pressure Multiple Risk Factor Intervention 12,886 high-risk men ages 35 to 57 No Diet, cigarette smoking, high Trial blood pressure Lipid Research Clinics 3806 hypercholesterolemic men Yes Cholestyramine Coronary Primary ages 35 to 59 Prevention Trial WHO Clofibrate Trial 11,627 hypercholesterolemic men Yes Clofibrate ages 30 to 59 Helsinki Heart Study 4081 hypercholesterolemic men Yes Gemfibrozil ages 40 to 55 Los Angeles Veterans 846 men ages 55 to 89 Yes Diet only Administration Domiciliary Study Oslo Study 1232 hypercholesterolemic norNo Diet, cigarette smoking motensive men ages 40 to 49 Finnish Mental Hospital Study 2 mental hospitals totaling No Diet only 4178 male patients and 6434 female patients ages 15 and older Source: Reprinted with permission from Diet and Health: Implications for Reducing Chronic Disease. © 1989, by the National Academy of Sciences. Courtesy of the National Academy Press, Washington, DC.
cholesterol reduction with treatment. A close relationship is apparent between reductions in cholesterol and in CHD events––either CHD death or fatal and nonfatal events combined. The statistically sig-
nificant decrease in CHD outcomes in the Lipid Research Clinics Coronary Primary Prevention Trial was a landmark finding. It provided experimental evidence supporting the large body of observational
Table 11-8
Randomized Trials on Primary Prevention of Coronary Heart Disease (CHD), Including Cholesterol Lowering by Diet or Drugs: Results % Differences Between Treated and Control Groupsa Serum Cholesterol Study at Entry (mg/dl) Serum Cholesterol CHDb Göteborg Multifactor Trial 250 0 0 WHO Multifactor Trial 216
1
7 Multiple Risk Factor Intervention Trial 254
2
7 Lipid Research Clinics Coronary Primary Prevention Trial 292
8
19* WHO Clofibrate Trial 242
9
20* Helsinki Heart Study 270
9
34* Finnish Mental Hospital Study (women) 275
12
34 Los Angeles Veterans Administration Domiciliary Study 233
13
24 Oslo Study 329
13
47* Finnish Mental Hospital Study (men) 267
15
53* a
At the end of the trial. CHD death was the endpoint in the WHO Multifactor Trial, the Multiple Risk Factor Intervention Trial, and the Finnish Mental Hospital Study. CHD death and nonfatal myocardial infarction were endpoints in the other studies. The studies varied in their technical definitions of these events. *p 0.05.
b
Source: Reprinted with permission from Diet and Health: Implications for Reducing Chronic Disease. © 1989, by the National Academy of Sciences. Courtesy of the National Academy Press, Washington, DC.
46894_CH11_269_310.qxd
294
2/25/10
3:39 PM
Page 294
CHAPTER 11 ADVERSE BLOOD LIPID PROFILE
epidemiologic studies supporting the “lipid hypothesis” and led directly to establishment of the NCEP. Other trials had equal or greater impact on cholesterol and on CHD event rates, including three dietary trials, of which the Finnish Mental Hospital Study in men had the greatest intervention impact. Dietary Trials During the 1980s and 1990s, 37 dietary intervention studies in free-living populations used the NCEP Step I diet ( 30% of total energy as fat, 10% as saturated fatty acids, and 300 mg dietary cholesterol/ day) or Step II diet ( 7% of energy as saturated fatty acids and 200 mg cholesterol/ day).60 Both diets were found to have beneficial effects on all lipid components, except that the Step II diet decreased HDLcholesterol slightly; this decrease was offset by weight loss, which increased HDL-cholesterol and decreased triglycerides. The American Heart Association reviewed the evidence for its Step I and Step II diets to reduce cholesterol levels.61 It was concluded that a 15% reduction in total or LDL-cholesterol might require reducing saturated fat intake to less than 7% of calories. Greater cholesterol reductions for high-risk patients were judged best attained by reducing calories from saturated fats to 5%, rather than further decreasing total fat intake. Optimal dietary advice was described as follows:61, p 3391 If saturated fat and cholesterol intakes are low, intake of fiber-rich foods is increased, and body weight is controlled, there will be an improvement in the cholesterol levels of most people, and risk of CHD will be reduced. Refinements in advice, such as the need for more physical activity (although exercise in terms of the LDLcholesterol profile is effective only with a reduction in adiposity), making appropriate food choices such as eating more vegetables and fruits and becoming familiar with food labels for information on food composition, including caloric content of food choices, are desirable. More Recent Drug Trials With more recent pharmaceutical developments, especially introduction of the statins, Brown and Goldstein, Nobel laureates for their work on lipid transport, forecast the disappearance of coronary heart disease as a public health problem early in the next (i.e., 21st) century.62 Although this view was overly optimistic in the face of the large forces un-
derlying epidemic coronary heart disease, it does convey the enthusiasm with which the statins have been greeted. They constitute a class of drugs whose action is to inhibit a liver enzyme, HMG CoA reductase, with a dual effect on LDL-cholesterol: Production of cholesterol in the liver is reduced, and the ability of the liver to remove LDL-cholesterol from the blood is increased.63 The primary and secondary prevention trials with statins that were the basis of ATP III guidelines are shown in Table 11-9.1 Depending on drug and dose, LDL-cholesterol was reduced by 25 to 35%, with generally significant reductions ranging from approximately 20 to 40%, in major coronary events, revascularization, and coronary and total mortality. Stroke was significantly reduced, but was evaluated only in secondary prevention trials. Statins were compared with five other classes of cholesterol-lowering drugs and with dietary interventions on the basis of a meta-analysis of 59 trials.64 Statins were found to have the greatest effect on cardiovascular and all-cause mortality, attributed to their greater efficacy in lowering LDL-cholesterol. Several dietary trials were included and had from no reduction (Gothenburg 1986) to 26% (Ornish 1990) reduction in LDL-cholesterol, but were often small with too few events for reliable estimation of effects. A meta-analysis of 14 trials of statins added reduction in fatal and nonfatal stroke to the benefits; effectiveness appeared to be greater in secondary than in primary prevention and beyond the first year of use.65 A Health Technology Assessment of statins, including economic evaluation, was reported in 2007 for the National Health Service in the United Kingdom.63 Now 31 trials could be evaluated, with similar conclusions regarding efficacy of statins in both primary and secondary prevention. Caution was noted in projecting benefits to the population at large, given selectivity of patients in the trials as to drug sensitivity, compliance, and other factors. Cost-effectiveness was considered less clear for primary than for secondary prevention. It was further concluded that:63, p iv The potential of targeting of statins at low-risk populations is however associated with major uncertainties, particularly the likely uptake and long-term compliance to lifelong medication by asymptomatic younger patients. The targeting, assessment and monitoring of low-risk patients in primary care would be a major resource implication for the NHS. It is important to note that, despite the importance and consistency of these findings regarding the efficacy of statins, pooling all placebo-controlled tri-
4.9 yrs 5 yrs 5.4 yrs 5 yrs 5 yrs
6595 6605 4444 4159 9014
Simvastatin 10/40 Pravastatin 40 mg Pravastatin 40 mg
Pravastatin 40 mg Lovastatin 20/40 188 139 150
192 150
31%*
37%*
35%*
25%*
29%*
26%*
25%*
35%*
27%*
25%*
Source: Reprinted from WMM Verschuren et al., Journal of the American Medical Association, Vol 274, p 131, © 1995.
*Statistically significant at p 0.05.
Study Primary Prevention WOSCOPS AFCAPS/ TexCAPS Secondary Prevention 4S CARE LIPID
37%*
27%*
24%*
37%*
33%*
42%*
24%*
24%*
33%* NS
Coronary Mortality
30%*
9%
23%*
22%* NS
Total Mortality
27%*
31%*
19%*
— —
Stroke
3:39 PM
More Recent Primary and Secondary Prevention Trials of Cholesterol Lowering with Statins Statin Baseline Major Drug LDL-C LDL-C Coronary RevascuPersons Duration (dose/d) (mg/dl) Change Events larization
2/25/10
Table 11-9
46894_CH11_269_310.qxd Page 295
PREVENTION AND CONTROL 295
46894_CH11_269_310.qxd
296
2/25/10
3:39 PM
Page 296
CHAPTER 11 ADVERSE BLOOD LIPID PROFILE
als showed no endpoint reduced by as much as 50%. The majority of cardiovascular events were not prevented despite this major advance in treatment, even under the idealized conditions of clinical trials. A further note is warranted on blood lipid control in women. ATP III points to a lack of data on drug therapy in primary prevention for women but describes nearly equivalent reduction in coronary heart disease event rates in women as in men in secondary prevention trials.1 American Heart Association guidelines for cardiovascular disease prevention in women were updated in 2007.66 Because of high lifetime risk for women, even a risk of cardiovascular disease that is less than 10% in 10 years was judged to justify drug therapy to lower LDL-cholesterol from levels 160 mg/dl in the presence of lifestyle therapy and other risk factors. Trials in Children and Adolescents A prominent example of a dietary intervention trial is the Dietary Intervention Study in Children (DISC).67 LDL-cholesterol concentration from the 80th to the 98th percentile was the criterion for entry, for boys and girls aged 8 to 10 years at entry. Mean values in intervention and control groups were 130.6 and 130.5 mg/dl at randomization. Results showed significant differential reduction in LDL-cholesterol concentration at years 1 and 3. Thereafter reductions were not statistically different and resembled changes expected during adolescence. Concern about possible adverse effects of a lipid-lowering diet in this period of life led to extended follow-up to gauge any difference in
growth or measures of nutritional well-being.68 There was no indication of adverse effects among participants followed up to 18 years of age. An AHA Scientific Statement reported a review of trials of cholesterol-lowering drugs in children and adolescents, all considered to have “high-risk lipid abnormalities” such as familial hypercholesterolemia. The review included 14 studies, seven of which used statins at one or multiple doses.69 The studies were small and of short duration but were taken to be consistent with studies in adults regarding safety and efficacy. Reductions in total and LDL-cholesterol were observed in all studies and nearly all showed increases in HDL-cholesterol, but results were mixed for triglycerides. The Statement concluded that “drug therapy should be targeted only toward individuals with highrisk lipid abnormalities or high-risk conditions who have not yet reached target lipid levels with lifestyle modification and should not be used as a first-line therapy for those whose lipid abnormalities are primarily lifestyle related.”69, p 1963 Guidelines for Intervention Individual approaches specific to blood lipid control in adults are addressed in ATP III.1 The scheme for adult screening, confirmation, and treatment is summarized in Figures 11-12A–E, which illustrate (A) the first patient encounter, and the subsequent steps depending on the risk status that is determined (in descending levels of risk); (B) presence of coronary heart disease (CHD) or “CHD risk equivalents” (existing CHD, other clinical atherosclerotic diseases,
Physician Responsibilities Visit 1 Control Risk Factors Public Health Message Reevaluate 1–5 Yrs
Patient Encounter
Lipoprotein Evaluation Risk Factor Evaluation
Assign Risk Status
(OR)
Initiate TLC* * If CHD or CHD risk equivalent is present, drug therapy can be started simultaneously with TLC when LDL C is $130 mg/dL.
Figure 11-12A Physician Responsibility for Visit 1. Source: Reprinted from Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). Final Report. NIH Publication No. 02-5215, September 2002, p IV-6.
46894_CH11_269_310.qxd
2/25/10
3:39 PM
Page 297
PREVENTION AND CONTROL 297
Continue TLC and Current Drug(s) LDL ⬍100 LDL ⱖ130
TLC ⫹ LDL-Lowering Drug(s) LDL 100⫺129
CHD and CHD Risk Equivalents
LDL 100⫺129
TLC ⫹ Therapeutic Options*
LDL ⬍100
TLC ⫹ Control Other Risk Factors
Consider Other Therapeutic Options*
* Therapeutic options include intensifying LDL lowering dietary or drug therapies, emphasizing weight reduction and increased physical activity, adding drugs to lower triglycerides or raise HDL cholesterol (nicotinic acid or fibrates), and intensifying control of other risk factors.
Figure 11-12B Therapeutic Approaches to Persons with 0–1 Risk Factor. The LDL-cholesterol goal is 100 mg/dl. Source: Reprinted from Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). Final Report. NIH Publication No. 02-5215, September 2002, p IV-6.
LDL ,130
Control Other Risk Factors Public Health Message on Healthy Life Habits Reevaluation in 1 Year
Multiple (21) Risk Factors 10-yr Risk 10220%
LDL $130
TLC
LDL ,130
Continue TLC
LDL $130
Continue TLC & Consider Adding LDL-Lowering Drugs
3 mos
Figure 11-12C Therapeutic Approaches to Persons with Mutiple Risk Factors, 10-Year Risk 10–12 Percent. The LDL-cholesterol goal is 130 mg/dl. Drugs can be considered if necessary to attain the LDL-cholesterol goal if the LDL-cholesterol level is 130 mg/dl after a trial of TLC. Source: Reprinted from Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). Final Report. NIH Publication No. 02-5215, September 2002, p IV-6.
46894_CH11_269_310.qxd
298
2/25/10
3:39 PM
Page 298
CHAPTER 11 ADVERSE BLOOD LIPID PROFILE
LDL ,130
Control Other Risk Factors Public Health Message on Healthy Life Habits Reevaluation in 1 Year
Multiple (21) Risk Factors 10-yr Risk ,10%
LDL $130
TLC
LDL ,160
Continue TLC
LDL $160
Continue TLC & Consider Adding LDL-Lowering Drugs
3 mos
Figure 11-12D Therapeutic Approaches to Persons with Multiple (2+) Risk Factors, 10-Year Risk 10 Percent. The LDL-cholesterol goal is 130 mg/dl. Drug therapy can be considered if LDL-cholesterol is 160 mg/dl after a trial of TLC. Source: Reprinted from Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). Final Report. NIH Publication No. 02-5215, September 2002, p IV-6.
LDL ,130
Public Health Messages on Healthy Life Habits Reevaluation: 5 Years
LDL 1302159
Public Health Messages on Healthy Life Habits Reevaluation: 1 Year
021 Risk Factor (10-year risk usually ,10%)
LDL ,160
LDL $160
TLC
Continue TLC
3 mos
LDL 1602189
Continue TLC & LDL - Lowering Drugs Optional*
LDL $190
Continue TLC & Consider Adding LDL - Lowering Drugs
* Factors favoring drug use are a severe single risk factor, a family history of premature CHD, and/or underlying or emerging risk factors in addition to a single major risk factor.
Figure 11-12E Therapeutic Approaches to Persons with 0–1 Risk Factor. The LDL-cholesterol goal is 160 mg/dl. Drug therapy can be considered if the LDL-cholesterol level is 190 mg/dl after a trial of TLC. If LDL-cholesterol is 160–189 mg/dl, drug therapy is optional depending on clinical judgement. Source: Reprinted from Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). Final Report. NIH Publication No. 02-5215, September 2002, p IV-6.
46894_CH11_269_310.qxd
2/25/10
3:39 PM
Page 299
PREVENTION AND CONTROL 299
diabetes mellitus, or multiple risk factors and 10year risk of CHD 20%); (C) multiple (2) risk factors with 10-year risk 10–20%; (D) multiple (2) risk factors with 10-year risk 10%; or (E) 0–1 risk factor (10-year risk usually 10%). Because “TLC” (therapeutic lifestyle changes) are called for in nearly every scenario, the plan for implementing and monitoring this aspect of management is also presented, in Figure 11-13. This illustrates the sequence of visits at intervals of 6 weeks and then up to every 6 months to evaluate the LDL-cholesterol response to lifestyle change. In every scenario, regardless of risk level, LDL-cholesterol is the focus of management, with target levels specific to initial risk. Level and intensity of therapy are guided by risk level and response to previous therapy. For children and adolescents, an analogous assessment is to be performed under the NCEP guidelines, selectively on the basis of family history, as discussed previously.20 Those with borderline or high LDL-cholesterol are to be treated with diet. Step One Diet is the initial approach, then Step Two is taken if the response is inadequate and further reduction of saturated fat and cholesterol intake is advised. Later, only as a last resort, drugs are used for those who are unresponsive to diet and are at least 10 years of age. The Expert Panel emphasized that the primary strategy for blood lipid control in childhood and adolescence is the population strategy and that the high-risk approach is intended only for those with the strongest presumption of future risk, based on family history.
Community and Population-Wide Measures Evidence for community or population-wide interventions to reduce cholesterol levels or improve blood lipid profiles is largely embedded within the experience of multifactor community trials. This is because these undertakings are most often designed to impact the major risk factors all at once, rather than to address a single factor, so as to achieve maximum reduction of cardiovascular event rates. Being broader in scope than any one risk factor, they are reviewed together in Chapter 21, “The Case for Prevention.” Documented there, for example, are declines in total cholesterol concentration in the North Karelia Project, in Finland, of about 1 mmol/L (39 mg/dl) as a population mean over the period 1972–1992.70 These population changes resulted from broad-based interventions affecting agriculture, public knowledge, and eating habits as well as other aspects of lifestyle and health care. The triad of community intervention trials in the United States, at Stanford, Minnesota, and Pawtucket, Rhode Island, had mixed success in reducing cholesterol levels; improvements in intervention communities were most often equaled or exceeded in the control communities, with greater reduction overall in control than in intervention communities when the results were pooled.71 It was concluded from this era of community intervention studies that achieving change beyond the favorable secular trends in cholesterol (as well as blood pressure and smoking) requires multilevel, as well as multifactor, strategies to include
Visit 2 Visit 1 Begin Lifestyle Therapies Emphasize reduction in saturated fat & cholesterol Encourage moderate physical activity Consider referral to a dietitian
6 wks
Evaluate LDL response If LDL goal not achieved, intensify LDL-lowering Tx Reinforce reduction in saturated fat and cholesterol Consider adding plant stanols/sterols Increase fiber intake Consider referral to a dietitian
Visit 3 6 wks
Evaluate LDL response If LDL goal not achieved, consider adding drug Tx
Visit N Q 4⫺6 mo Monitor Adherence to TLC
Initiate Tx for Metabolic Syndrome Intensify weight management & physical activity Consider referral to a dietitian
Figure 11-13 A Model of Steps in Therapeutic Lifestyle Changes (TLC). Source: Reprinted from Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). Final Report. NIH Publication No. 02-5215, September 2002, p IV-X.
46894_CH11_269_310.qxd
300
2/25/10
3:39 PM
Page 300
CHAPTER 11 ADVERSE BLOOD LIPID PROFILE
personal, family, community, and policy/environmental changes that would be mutually reinforcing. Interventions approaching this concept and assessing impact on cholesterol as well as other risk factors have been undertaken in communities and workplaces. The Asheville (NC) Project utilized community and hospital pharmacy clinics to support control of high cholesterol and blood pressure for two cohorts of patients with these conditions.72 Over a 6-year period, mean LDL-cholesterol levels decreased from 127 to 108 mg/dl and control improved from 40 to 67%. Total cholesterol and triglycerides also decreased; HDL-cholesterol was unaffected. Despite increased medication costs, overall healthcare costs decreased because of reduced frequency of emergency department visits and hospital admissions. Workplace studies can be illustrated by example from the New York City Department of Health and Mental Hygiene’s Wellness at Work program.73 In a public–private collaboration involving the department and several partners, health risks were assessed for employees in 10 New York City organizations. Two levels of intervention provided either a moderate-intensity program to support positive behavior changes or a high-intensity program with additional individualized intervention. The proportions of employees with highrisk cholesterol 2 years after baseline decreased from 8.4 to 4.9%. Moderate-intensity intervention was more effective than the high-intensity program for cholesterol reduction and was not statistically less successful across several risk factors evaluated. Interventions in childhood can be illustrated by studies in both the United States and Finland. The Child and Adolescent Trial for Cardiovascular Health (CATCH) intervened in schools to improve meals and physical activity, producing significant relative declines in consumption of total and saturated fat and cholesterol in intervention versus control schools.74 The North Karelia Youth Programs comprised two small family-based and two larger community- and school-based intervention programs aimed to determine whether serum cholesterol and blood pressure could be affected by dietary modifications.75 The family-based, but not the school-based, interventions resulted in significant decrease in total cholesterol concentration, by 15% from baseline values, after major decreases in intakes of total and saturated fats, cholesterol, and energy. In the school interventions, dietary change was also substantial, but the gradient of reduced cholesterol was significant only for girls. The starting age at 13 years and expected subsequent “spontaneous” change in total cholesterol over 2 years’ follow-up, discussed earlier, may have clouded the evaluation.
Population-wide measures for prevention and control of blood lipids have been included in recommendations from national and regional organizations, the World Health Organization, and in the context of global efforts to prevent chronic diseases. The evolution of these recommendations is reviewed in Chapter 20, “Recommendations, Guidelines, and Policies.” In the United States, for example, ATP III emphasizes the public health approach of the NCEP:1, p I-3 Lowering LDL-cholesterol levels in the whole population and keeping them low requires adoption of a low saturated fat and low cholesterol diet, maintenance of a healthy weight, and regular physical activity. . . . The population approach for controlling CHD risk factors will, in the long term, have the greatest impact on reducing the magnitude of cardiovascular disease in the United States. . . . The clinical approach alone cannot overcome the burden of atherosclerotic disease in the general population. A parallel and simultaneous effort must be made to promote changes in population life habits to retard atherogenesis. Reference is made by ATP III to NCEP’s Report of the Expert Panel on Population Strategies for Blood Cholesterol Reduction, published in 1990. The Report presented recommendations in 11 areas:76 1. nutrient intake—quantitative goals for fats and cholesterol 2. eating patterns—food choices to promote the desired nutrient intake, along with attention to other risk factors 3. healthy children and adolescents—adoption of similar eating patterns from age two 4. special groups—general application of the above recommendations to all (women, elderly, diverse ethnic groups, low-income groups) with special considerations when appropriate 5. health professionals—their role in communicating these recommendations 6. the food industry—support to make foods lower in saturated and total fats and cholesterol more readily available and better labeled and advertised 7. mass media—publicity for the desired eating pattern 8. government—adoption of appropriate policies 9. educational systems—enhanced dissemination of information supporting the desired eating
46894_CH11_269_310.qxd
2/25/10
3:39 PM
Page 301
PREVENTION AND CONTROL 301
patterns in a consistent way between official education agencies and health organizations 10. measurement of blood cholesterol—assurance of high-quality standards for cholesterol screening 11. research and surveillance—continuing investigation of relations between food and health and revision of guidelines accordingly Effectively implemented, these recommendations would be expected to reduce the prevalence of adverse blood lipid profiles among older and younger adults and prevent their initial development in youth and young adults. Regarding lifestyle changes to reduce alreadyincreased risk, and the implied linkage between public health and clinical approaches to primary prevention, ATP III states:1, p II-31 A strong case exists for the efficacy and safety of primary prevention through lifestyle changes. Primary prevention efforts extend to both public health and clinical arenas. . . . ATP III affirms the validity of lifestyle changes as first-line therapy for primary prevention. It places priority on LDL-lowering modifications because of the identification of LDL-cholesterol as the primary target of therapy; however, ATP II also urges the use of a broad approach to lifestyle changes for CHD risk reduction in primary prevention. Where “public health messages” are cited in the ATP III algorithms shown in Figure 11-12, the content is to address “avoidance or cessation of cigarette smoking, reduction of intakes of saturated fats and cholesterol, achieving and maintaining a healthy body weight, regular physical activity, and routine medical check-ups for blood pressure and cholesterol.”1, p V-1 Similarly, the NCEP report on cholesterol in children and adolescents identified the population approach as “the principal means for preventing CHD” and advocated a strategy of population-wide changes in nutrient intake and eating patterns to reduce blood cholesterol in all American children.20 Global Strategies Recommendations in Technical Reports from WHO are directed to the attention of all Member States as a call to action. It is anticipated that corresponding national policies will be developed in consequence of the reports. They are in this sense potentially global in reach. The World Health Assembly in 1976 adopted a resolution calling for development of a
comprehensive research program and for coordination of international activities in prevention of coronary heart disease. In keeping with that charge, a WHO Expert Committee was convened in late 1981. Their report reviewed concepts of prevention at the population and individual levels, outlined elements of the population-wide strategy in particular— including reference to the problem in developing countries—and offered recommendations for program implementation.77 Directly relevant here are the conclusions and recommendations concerning blood lipids: 1. attainment of a population mean value of total cholesterol concentration less than 200 mg/dl by reduction of saturated fat intake to less than 10% of energy intake and of dietary cholesterol intake to less than 300 mg/day and by avoidance of obesity 2. increase in habitual physical activity for several purposes, including reduction of cholesterol concentration 3. beginning intervention in youth to avert the development of elevated blood cholesterol concentrations in the first place A subsequent Expert Committee Report, Prevention in Childhood and Youth of Adult Cardiovascular Diseases—Time for Action, addressed atherosclerotic and hypertensive diseases and other cardiovascular conditions of major public health concern.78 Foremost among recommendations in this 1990 report were: modifying dietary patterns to reduce intake of saturated fat and cholesterol, increasing the intake of complex carbohydrates, and avoiding or correcting overweight, all for the purpose of reducing undesirable cholesterol concentrations or averting them in the first place. The Disease Control Priorities in Developing Countries Project is also a global health initiative, with emphasis on low- and middle-income countries (see Chapter 1, “Cardiovascular Diseases: A Global Public Health Challenge”). In the second edition of the Project’s monumental report, DCP2, Rodgers and others address high cholesterol, together with high blood pressure and body weight—their contribution to the global burden of chronic diseases, potential interventions, their effectiveness and cost-effectiveness, and broader economic considerations in policy development.79 Both personal and population-based interventions are considered, separately in developed and developing countries. A strong case is made for cost-effectiveness of an intervention strategy based on absolute risk—
46894_CH11_269_310.qxd
302
2/25/10
3:39 PM
Page 302
CHAPTER 11 ADVERSE BLOOD LIPID PROFILE
the probability of a coronary or cardiovascular event that is estimated by taking multiple risk factors into account—rather than targeting any single risk factor, such as high cholesterol. This argument is well developed for personal interventions that depend on often costly medications and the clinical services surrounding their use—as illustrated in Figures 11-12A–E, shown previously. This is because within any risk category, such as high or borderline levels defined for one factor, a range of estimated risk applies—as much as 10-fold—when other factors are taken into account. Two potential consequences of failure to take absolute risk into account are treating individuals at low risk, thus having little impact, and neglect of treatment for individuals at high risk, where impact could be greatest. By contrast, it is inherent in dietary and other lifestyle interventions that multiple factors—high cholesterol, blood pressure, and body weight, as well as diabetes and other conditions—are all likely to be improved when these interventions are implemented effectively and sustained. Accordingly, recommendations at the personal level in DCP2 begin with lifestyle modification, such as education programs. Costeffectiveness in this case depends on both program costs and the percentage reduction achieved in the population level of a given risk factor: The greater the reduction, the more favorable the cost-effectiveness ratio for a given intervention. (Logical consistency would require measuring impact as reduction in the population distribution of absolute risk, but this is not addressed.) Developed and developing countries present different cost-effectiveness ratios for personal interventions with drugs because of differences in costs of medicines and systems for their delivery. Even among developed countries, estimates vary widely both as a result of these factors and of differences in population risks. However, in several of the WHO Regions, if low-cost medicines are assumed, there is a consistent relationship among cost-effectiveness ratios for alternative strategies (Table 11-10). “Prevention” in the case of high blood pressure and cholesterol equates to population-wide measures, salt-lowering legislation (SL), and diet-related health education (HE). “Targeted risk factors” refers to blood pressure- and cholesterol-specific individual pharmacological interventions. “Absolute risk (TRI SL HE)” means triple therapy with a beta-blocker, statin, and aspirin plus the population-wide measures. This is separately evaluated depending on whether the risk threshold for treatment is set at 35, 25, 15, or 5% 10-year risk of CVD. In terms of cost per DALY saved, prevention generally has the lowest (most favorable) cost-
effectiveness ratio whether in local purchasing power (“International$”) or in US$. Targeted risk strategies were generally rated as “dominated”—the absolute risk strategy at 35% risk cost less and saved more DALYs. While addressing personal and populationwide interventions by lifestyle or pharmacological means, in both developed and developing countries, what is presented here as a global strategy is adoption of cost-effectiveness analysis as the appropriate tool, with intervention choices based on absolute risk as the outcome. Other dimensions of global strategy that bear on adverse blood lipid profiles are those affecting dietary imbalance and physical inactivity, discussed in earlier chapters.
CURRENT ISSUES Present Guidelines Guidelines for detecting, evaluating, and treating high cholesterol and other components of the blood lipid profile in adults, such as ATP III in the United States and the recent National Institute for Health and Clinical Excellence (NICE) guidance in the United Kingdom, are intended to determine practice.1,80 Taking the United States as an example, the extent to which they actually do so is quite limited on the evidence of low awareness, treatment, and control of high cholesterol in national surveys. The WHO MONICA Project data similarly reflect limited impact of such policies in the United Kingdom and Europe. This issue is further emphasized in ATP III:1, p IX-1 Only about half of the persons who are prescribed a lipid-lowering drug are still taking it six months later; after 12 months this falls to 30–40 percent of persons. This is especially disconcerting, since it takes 6 months to 1 year before a benefit from treatment becomes apparent. . . . For this benefit to be realized, treatment will have to be continued for years and probably for the duration of the patient’s life. Thus, paying attention to ways of improving adherence with treatment is just as important to the ultimate success of these guidelines as are the rudiments of the guidelines themselves. ATP III cited a large systematic review that found little evidence for consistently effective strategies to improve medication adherence.1 An inventory of approaches to improve adherence by both patient and provider is presented, but it remained to be determined which if any of these will be successful in one or another type of treatment setting. Whether over-
46894_CH11_269_310.qxd
2/25/10
3:39 PM
Page 303
CURRENT ISSUES 303
Table 11-10
Region Africa E
Comparison of the Cost-Effectiveness of Absolute Risk with Treatment According to Either Blood Pressure or Lipid Targets Alone in Addition to Population-Based Strategies, Selected WHO Regions Incremental Cost-Effectiveness Ratio (Cost/DALY Saved)a Strategy Risk (Percent) International$ US$ Prevention (SL and/or HE) Dominatedb Targeted risk factorsc Dominatedb Absolute riskc (TRI) 35 138 42 25 778 295 15 1445 639
Latin America and the Caribbean B
Southeast Asia B
Western Pacific B
Prevention (SL) Prevention (SL HE) Targeted risk factorsc Absolute riskd (TRI SL HE)
Prevention (SL) Prevention (SL HE) Targeted risk factorsc Absolute riskd (TRI SL HE)
Prevention (SL) Targeted risk factorsc Absolute riskd (TRI SL HE)
35 25 15 5
127 145 Dominatedb 286 1598 2391 4319
35 25 15 5
70 127 Dominatedb 301 1197 2094 3952
35 25 15 5
97 Dominatedb 1124 1278 2092 4028
65 74 178 1058 1664 3075 18 32 133 578 1120 2233 18 423 564 1042 2135
Source: Murray and others, 2003. B low child mortality and low adult mortality; E high child mortality and very high adult mortality; HE health education through the mass media to reduce cholesterol; SL legislation to decrease the salt content of processed foods, including appropriate labeling and enforcement; TRI treatment with aspirin, beta-blockers, and a statin. a Costs of prevention and nondrug costs for treatment according to absolute risk are converted at an estimated regional average ratio of exchange rate to purchasing-power parity rate; drug costs are not converted, assuming drugs to be imported at world prices. The share of drug costs in total treatment cost, as a function of risk, is taken from the estimates for India in table 45.6 and assumed to be the same for all regions. b Dominated strategies were both less effective and more costly than comparator strategies. c Treating SBP greater than 140 mm Hg or 160 mm Hg or total cholesterol greater than 5.7 mmol/L or 6.2 mmol/L (220 or 240 mg/dl). d Risk refers to 10-year risk of CVD greater than or equal to the number listed. Source: Reprinted with permission from Disease Control Priorities in Developing Countries, Second Edition, A Rodgers et al., p 861, © 2006. The International Bank for Reconstruction and Development/The World Bank.
the-counter (OTC) availability of statins would contribute to a solution, notwithstanding any concerns about this policy, is uncertain.81 Other issues also contribute to the gap between the vision reflected in guidelines and the reality of their public health impact. These include: • the large and growing size of the target population, including ever-lower treatment thresholds and targets, with consequently rising costs of full implementation82,83 • outdating of guidelines due to new data and changing views84
• uncertain applicability or population impact of guidelines across national boundaries or among different target populations85–87 Within the United States, there is the further issue of conflicting recommendations, for example, between the United States Preventive Services Task Force (USPSTF) and NCEP guidelines for both adults and children.88 In contrast to ATP III, the USPSTF does not recommend screening of men aged 20–35 years or women aged 20–45 years in the absence of other coronary heart disease risk factors. In contrast to both the NCEP guidelines and recent guidelines
46894_CH11_269_310.qxd
304
2/25/10
3:39 PM
Page 304
CHAPTER 11 ADVERSE BLOOD LIPID PROFILE
from the American Academy of Pediatrics (AAP), USPSTF makes no recommendation regarding blood lipid screening in children or adolescents.20,22 Within the pediatric age range, there are also differences between the NCEP (1992) and AAP (2008) guidelines, including provision in the latter for age-specific criteria, although in 5-year rather than single year of age categories as proposed elsewhere.20,22 The NCEP guideline appears intended for children age 2 years or older; the AAP guideline calls explicitly for initial screening no earlier than age 2 years and no later than age 10 years. Changing Guidelines For reasons indicated above, guidelines addressing blood lipids, as in other areas, can be expected to change as underlying evidence and larger policy or system changes suggest. Where absolute risk estimation is the approach, questions are debated (see Chapter 20, “Recommendations, Guidelines, and Policies”): How valid are the risk estimates, and are they improved by incorporating additional predictors? How applicable is one model across populations with different risk distributions and disease rates? What is the appropriate scope of predicted outcomes whose risk is to be reduced—coronary events alone, broader cardiovascular outcomes, or still others such as vascular dementia? Is the conventional 10-year horizon for risk prediction, limiting attention to short-term risk, the best policy? In the process of reviewing existing guidelines, NHLBI is undertaking a dual process. It will lead to Adult Treatment Panel IV and other risk-factorspecific updates and, concurrently, produce two new, integrated, multifactor guidelines: for adults, the Cardiovascular Disease Risk Reduction guideline; and for children and adolescents, an analogous integrated guideline to address overall cardiovascular risk-factor identification and risk reduction.89,90 Beyond the Guidelines The public health dimensions of prevention and control of adverse blood lipid profiles are expressed clearly in the guideline documents reviewed here. But, as with the clinical components, public health recommendations remain to be implemented sufficiently to have sustained effect at the population level— whether measured by maintenance of favorable distributions of blood lipid levels from childhood; improvement of unfavorable distributions typically found in adulthood; or specifically reducing individual risks as measured by high prevalence of awareness, treatment, and control.
The difficulties encountered in development and implementation of policies and guidelines for riskfactor control make a self-evident case for prevention of adverse blood lipids in the first place. Only by effectively establishing and maintaining patterns of diet and activity that favor optimum cardiovascular health can the perpetual progression of adverse blood lipids from each generation of children into the next generation of adults at risk be averted. As expressed nearly two decades ago, it is “Time for Action.”78 REFERENCES 1. National Cholesterol Education Program Expert Panel on Detection, Evaluation and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). NIH Publication No. 02-5215. Bethesda, MD: National Cholesterol Education Program, National Heart, Lung and Blood Institute, National Institutes of Health; September 2002. 2. Stamler J. Established major coronary risk factors. In: Marmot M, Elliott P, eds. Coronary Heart Disease Epidemiology: From Aetiology to Public Health. Oxford (England): Oxford University Press; 1992:35–66. 3. Keys A. From Naples to Seven Countries—a sentimental journey. Prog Biochem Pharmacol. 1983;19:1–30. 4. National Research Council (US) Food and Nutrition Board. Diet and Health. Implications for Reducing Chronic Disease Risk. Washington, DC: Committee on Diet and Health, Food and Nutrition Board, Commission on Life Sciences, National Research Council, National Academy Press; 1989. 5. Gotto AM, Jr. Lipid and lipoprotein disorders. In: Pearson TA, Criqui MH, Luepker RV, Oberman A, Winston M, eds. Primer in Preventive Cardiology. Dallas, TX: American Heart Association; 1994:107–129. 6. Ginsberg HN. Lipoprotein metabolism and its relationship to atherosclerosis. Med Clin North Am. 1994;78:1–20.
46894_CH11_269_310.qxd
2/25/10
3:39 PM
Page 305
REFERENCES 305
7. Sullivan DR. Lipoprotein (a). J Cardiovasc Risk. 1994;1:212–216. 8. Wilson PWF, Schaefer EJ, Larson MG, Ordovas JM. Apolipoprotein E alleles and risk of coronary disease. A meta-analysis. Arterioscler Thromb Vasc Biol. 1996;16:1250–1255. 9. NIH Consensus Development Panel on Triglyceride, High-Density Lipoprotein and Coronary Heart Disease. Triglyceride, highdensity lipoprotein, and coronary heart disease. JAMA. 1993;269:505–510. 10. Viles-Gonzalez JF, Fuster V, Corti R, Badimon JJ. Emerging importance of HDL cholesterol in developing high-risk coronary plaques in acute coronary syndromes. Curr Opinion Cardiol. 2003;18:286–294. 11. Choi BG, Vilahur D, Viles-Gonzalez JF, Badimon JJ. The role of high-density lipoprotein cholesterol in the prevention and possible treatment of cardiovascular diseases. Curr Molec Med. 2006;6:571–587. 12. Toth PP. High-density lipoprotein as a therapeutic target: clinical evidence and treatment strategies. Am J Cardiol. 2005;96(suppl): 50K–58K. 13. Thompson WG, Gau GT. Invited commentary. Arch Intern Med. 2009;169:578–579. 14. Maki KC, Galant R, Davidson MH. Non-highdensity lipoprotein cholesterol: the forgotten therapeutic target. Am J Cardiol. 2005; 96(suppl):59K–64K. 15. Natarjan S, Glick H, Criqui M, Horowitz D, Lipsitz SR, Kinosian B. Cholesterol measures to identify and treat individuals at risk for coronary heart disease. Am J Prev Med. 2003;25(1):50–57. 16. Ballantyne CM, Hoogeveen RC. Role of lipid and lipoprotein profiles in risk assessment and therapy. Am Heart J. 2003;146:227–233. 17. ATP III At-A-Glance: Quick desk reference. http://www.nhlbi.nih.gov/guidelines/cholesterol/ atglance/.htm. Accessed June 27, 2007.
18. Graham IM. Guidelines on cardiovascular disease prevention in clinical practice: the European perspective. Curr Opin Cardiol. 2005;20:430–439. 19. World Health Organization. Prevention of Cardiovascular Disease. Guidelines for Assessment and Management of Cardiovascular Risk. Geneva: World Health Organization; 2007. 20. National Cholesterol Education Program. Report of the Expert Panel on Blood Cholesterol Levels in Children and Adolescents. NIH publication 91-2732. Bethesda, MD: National Cholesterol Program, National Heart, Lung and Blood Institute, National Institutes of Health, Public Health Service, US Department of Health and Human Services; 1991. 21. Labarthe DR, Dai S, Fulton JE. Cholesterol screening in children: insights from Project HeartBeat! and NHANES III. Prog Pediatr Cardiol. 2003;17:169–178. 22. Daniels SR, Greer FR, and the Committee on Nutrition. Lipid screening and cardiovascular health in childhood. Pediatrics. 2008;122: 198–208. 23. Joliffe CJ, Janssen I. Distribution of lipoproteins by age and gender in adolescents. Circulation. 2006;114:1056–1062. 24. United States General Accounting Office. Report to the Chairman, Subcommittee on Investigations and Oversight, Committee on Science, Space, and Technology, House of Representatives. Cholesterol Measurement, Test Accuracy and Factors That Influence Cholesterol Levels. GAO/PEMD-95-8. Washington, DC: General Accounting Office; 1994. 25. Guidelines for Cholesterol Screening Programs. Dallas, TX: American Heart Association; 1989. 26. Labarthe DR. Emergence of risk factors in children. In: Marmot M, Elliott P, eds. Coronary Heart Disease Epidemiology: From Aetiology to Public Health. 2nd ed. Oxford (England): Oxford University Press; 2005:591–605.
46894_CH11_269_310.qxd
306
2/25/10
3:39 PM
Page 306
CHAPTER 11 ADVERSE BLOOD LIPID PROFILE
27. National Center for Health Statistics. Health, United States, 2008. With Chartbook. Hyattsville, MD; 2009. 28. Williams RR, Schumacher C, Hopkins PN, et al. Practical approaches for finding and helping coronary-prone families with special reference to familial hypercholesterolemia. In: Goldbourt U, de Faire U, Berg K, eds. Genetic Factors in Coronary Heart Disease. Dordrecht, the Netherlands: Kluwer Academic Publishers; 1994:425–445. 29. III. The individualized approach: detection/ diagnosis/evaluation. Pediatrics. 1992;89: 545–554. 30. Hegsted DM, Ausman LM, Johnson JA, Dallal GE. Dietary fat and serum lipids: an evaluation of the experimental data. Am J Clin Nutr. 1993;57:875–883. 31. St. Clair RW. Biology of atherosclerosis. In: Pearson TA, Criqui MH, Luepker RV, Oberman A, Winston M, eds. Primer in Preventive Cardiology. Dallas, TX: American Heart Association; 1994:11–24. 32. US Department of Health and Human Services. Healthy People 2010 Midcourse Review: Washington, DC: US Government Printing Office; December 2006. 33. Goff Jr DC, Labarthe DR, Howard G, Russell GB. Primary prevention of high blood cholesterol concentrations in the United States. Arch Intern Med. 2002;162:913–919. 34. Ford ES, Mokdad AH, Giles WH, Mensah GA. Serum total cholesterol concentrations and awareness, treatment, and control of hypercholesterolemia among US adults. Findings from the National Health and Nutrition Examination Survey, 1999 to 2000. Circulation. 2003; 107:2185–2189. 35. Ford ES, Li C, Pearson WS, Zhao G, Mokdad AH. Trends in hypercholesterolemia, treatment and control among United States adults. Int J Cardiol. 2008; doi:10.1016/j.ijcard.2008. 11.033. 36. Klingman D, Williams SA, Benner JS, Smith TW, Ahn J, O’Donnell JC. Gauging the treat-
ment gap in dyslipidemia: findings from the 1999–2000 National Health and Nutrition Examination Survey. Am Heart J. 2005;150: 595–601. 37. Ford ES, Li C, Zhao G, Pearson WS, Mokdad AH. Hypertriglyceridemia and its pharmacologic treatment among US adults. Arch Int Med. 2009;169:572–578. 38. Lloyd Jones D, Adams R, Carnethon M, et al. Heart disease and stroke statistics 2009 update. A report from the American Heart Association Statistics Committee and Stroke Statistics Committee. doi:10.1161/ CIRCULATIONAHA.108.191261. http:// circ.ahajournals.org. Accessed December 18, 2008. 39. Keys A. Serum cholesterol and the question of “normal.” In: Benson ES, Strandjord PE, eds. Multiple Laboratory Screening. New York: Academic Press; 1969:147–170. 40. Knuiman JT, West CE, Burema J. Serum total and high density lipoprotein cholesterol concentrations and body mass index in adult men from 13 countries. Am J Epidemiol. 1982;116: 631–642. 41. Tolonen H, Keil U, Ferrario M, Evans A for the WHO MONICA Project. Prevalence, awareness and treatment of hypercholesterolemia in 32 populations: results from the WHO MONICA Project. Int J Epidemiol. 2005;34: 181–192. 42. Labarthe DR, O’Brien B, Dunn K. International comparisons of plasma cholesterol and lipoproteins. In: Williams CL, Wynder EL, eds. Hyperlipidemia in Childhood and the Development of Atherosclerosis: Annals of the New York Academy of Sciences. New York: New York Academy of Sciences; 1991; 623: 108–119. 43. Labarthe DR, Nichaman MZ, Harrist RB, Grunbaum JA, Dai S. Development of cardiovascular risk factors from ages 8 to 18 in Project HeartBeat! Study design and patterns of change in plasma total cholesterol concentration. Circulation. 1997;95:2636–2642.
46894_CH11_269_310.qxd
2/25/10
3:39 PM
Page 307
REFERENCES 307
44. Knuiman JT, Hermus RJ, Hautvast JG. Serum total and high density lipoprotein (HDL) cholesterol concentrations in rural and urban boys from 16 countries. Atherosclerosis. 1980; 36(4):529–537. 45. Keys A. Seven Countries. A Multivariate Analysis of Death and Coronary Heart Disease. Cambridge, MA: Harvard University Press; 1980. 46. Verschuren WMM, Jocobs DR, Bloemberg BPM, et al. Serum total cholesterol and longterm coronary heart disease mortality in different cultures: twenty-five year follow-up of the Seven Countries Study. JAMA. 1995;274: 131–136. 47. The Pooling Project Research Group. Relationship of blood pressure, serum cholesterol, smoking habit, relative weight and ECG abnormalities to incidence of major coronary events: Final Report of the Pooling Project. J Chronic Dis. 1978;31:201–306. 48. Martin MJ, Hulley SB, Browner WS, et al. Serum cholesterol, blood pressure, and mortality: implications from a cohort of 361,622 men. Lancet. 1986;2:933–936. 49. Klag MJ, Ford DE, Mead LA, et al. Serum cholesterol in young men and subsequent cardiovascular disease. N Engl J Med. 1993;328: 313–318. 50. Mosca L, Manson JE, Sutherland SE, Langer RD, Manolio T, Barrett-Connor E. Cardiovascular disease in women. A statement for healthcare professionals from the American Heart Association. Circulation. 1997;96: 2468–2482. 51. Welty FK. Cardiovascular disease and dyslipidemia in women. Arch Intern Med. 2001;161: 514–522. 52. Johansson S, Wilhelmsen L, Lappas G, Rosengren A. High lipid levels and coronary disease in women in Göteborg—outcome and secular trends: a prospective 19 year follow-up in the BEDA study. Eur Heart J. 2003;24: 704–716.
53. LaRosa JC. Lipoproteins and lipid disorders. In: Douglas PS, ed. Cardiovascular Health and Disease in Women. Philadelphia, PA: WB Saunders Co; 1993:175–189. 54. Iso H, Jacobs DR, Jr, Wentworth D, Neaton JD, et al., for the MRFIT Research Group. Serum cholesterol levels and six-year mortality from stroke in 350,977 men screened for the Multiple Risk Factor Intervention Trial. N Engl J Med. 1989;320:904–910. 55. Nguyen TT, Ellefson RD, Hodge DO, et al. Predictive value of electrophoretically detected lipoprotein(a) for coronary heart disease and cerebrovascular disease in a community-based cohort of 9936 men and women. Circulation. 1997;96:1390–1397. 56. Yusuf S, Hawken S, Ôunpuu S, et al. Effect of potentially modifiable risk factors associated with myocardial infarction in 52 countries (the INTERHEART study): case-control study. Lancet. 2004;364;937–952. 57. Mazzati E, Vander Hoorn S, Lopez AD, et al. Comparative quantification of mortality and burden of disease attributable to selected risk factors. In: AD Lopez et al., eds. Global Burden of Disease and Risk Factors. The International Bank for Reconstruction and Development/The World Bank, Washington, DC; 2006:241–396. 58. The Coronary Drug Project Research Group. Clofibrate and niacin in coronary heart disease. JAMA. 1975;231:360–381. 59. Levy DR, PearsonTA. Combination of niacin and statin therapy in primary and secondary prevention of cardiovascular disease. Clin Cardiol. 2005;28:317–320. 60. Yu-Poth S, Zhao G, Etherton T, Naglak M, Jonnalagadda S, Kris-Etherton PM. Effects of the National Cholesterol Education Program’s Step I and Step II dietary intervention programs on cardiovascular disease risk factors: a meta-analysis. Am J Clin Nutr. 1999;69: 632–646. 61. Stone NJ, Nicolosi RJ, Kris-Etherton P, Ernst ND, Krauss RM, Winston M. Summary of
46894_CH11_269_310.qxd
308
2/25/10
3:39 PM
Page 308
CHAPTER 11 ADVERSE BLOOD LIPID PROFILE
the scientific conference on the efficacy of hyhpocholesterolemic dietary interventions. Circulation.1996;94:3388–3391. 62. Brown MS, Goldstein JL. Heart attacks: gone with the century? Science. 1996;272:629. 63. Ward S, Lloyd Jones M, Pandor A, et al. A systematic review and economic evaluation of statins for the prevention of coronary events. Health Technol Assess. 2007;11(14). 64. Bucher HC, Griffith LE, Guyatt GH. Systematic review on the risk and benefit of different cholesterol-lowering interventions. Arterioscler Thromb Vasc Biol. 1999;19: 187–195. 65. Cholesterol Treatment Trialists’ (CTT) Collaborators. Efficacy and safety of cholesterol-lowering treatment: prospective metaanalysis of data from 90 056 participants in 14 randomised trials of statins. Lancet. 2005; 366:1267–1278. 66. Mosca L, Banka CL, Benjamin EL for the Expert Panel/Writing Group. Evidence-based guidelines for cardiovascular disease prevention in women: 2007 update. Circulation. 2007;115:1481–1501. 67. Writing Group for the DISC Collaborative Research Group. Efficacy and safety of lowering dietary intake of fat and cholesterol in children with elevated low-density lipoprotein cholesterol: the Dietary Study in Children (DISC). JAMA. 1995;273:1429–1435. 68. Obarzanek E, Kimm SYS, Barton BA, et al. Long-term safety and efficacy of a cholesterollowering diet in children with elevated lowdensity lipoprotien cholesterol: seven-year results of the Dietary Intervention Study in Children (DISC). Pediatrics. 2001;107: 256–264. 69. McCrindle BW, Urbina EM, Dennison BA, et al. Drug therapy of high-risk lipid abnormalities in children and adolescents. A Scientific Statement from the American Heart Association Atherosclerosis, Hypertension, and Obesity in Youth Committee, Council of Cardiovascular Disease in the Young, with the Council on
Cardiovascular Nursing. Circulation. 2007; 115:1948–1967. 70. Puska P, Tuomilehto J, Nissinen A, Vartiainen E, eds. The North Karelia Project. 20 Year Results and Experiences. Helsinki, Finland: National Public Health Institute, KTL; 1995. 71. Winkelby MA, Feldman HA, Murray DM. Joint analysis of three US community intervention trials for reduction of cardiovascular disease risk. J Clin Epidemiol. 1997;50:645–658. 72. Bunting BA, Smith BH, Sutherland SE. The Asheville Project: clinical and economic outcomes of a community-based long-term medication therapy management program for hypertension and dyslipidemia. J Am Pharm Assoc. 2008;48:23–31. 73. Goetzel RZ, Roemer EC, Short ME, et al. Health improvement from a worksite health promotion private-public partnership. J Occup Environ Med. 2009;51:296–304. 74. Luepker RV, Perry CL, McKinlay SM, et al. Outcomes of a field trial to improve children’s dietary patterns and physical activity: the Child and Adolescent Trial for Cardiovascular Health (CATCH). JAMA. 1996;275:768–776. 75. Vartiainen E, Tossavainen K, Viri L, Niskanen E, Puska P. The North Karelia Youth Programs. Ann NY Acad Sci. 1991;623: 332–349. 76. National Cholesterol Education Program. Report of the Expert Panel on Population Strategies for Blood Cholesterol Reduction. NIH publication 90-3046. Bethesda, MD: National Heart, Lung and Blood Institute, National Institutes of Health, Public Health Service, US Department of Health and Human Services; 1990. 77. World Health Organization. Report of a WHO Expert Committee. Prevention of Coronary Heart Disease. Technical Report Series 678. Geneva (Switzerland): World Health Organization; 1982. 78. World Health Organization. Report of a WHO Expert Committee. Prevention in Childhood and Youth of Adult Cardiovascular Diseases:
46894_CH11_269_310.qxd
2/25/10
3:39 PM
Page 309
REFERENCES 309
Time for Action. Technical Report Series 792. Geneva (Switzerland): World Health Organization; 1990. 79. Rodgers A, Lawes CMM, Gaziano T, Vos T. The growing burden of risk from high blood pressure, cholesterol, and bodyweight. In: Jamison DT, Breman JG, Measham AR, Alleyne G, Claeson M, Evans DB, et al., eds. Disease Control Priorities in Developing Countries. 2nd ed. Washington, DC: International Bank for Reconstruction and Development/The World Bank; 2006: 851–868. 80. Cooper A, O’Flynn N on behalf of the Guideline Development Group. Risk assessment and lipid modification for primary and secondary prevention of cardiovascular disease: summary of NICE guidance. BMJ. 2008;336:1246–1248. 81. Pearson TA. The epidemiologic basis for population-wide cholesterol reduction in the primary prevention of coronary artery disease. Am J Cardiol. 2004;94(suppl):4F–8F. 82. Fedder DO, Koro CE, L’Italien GJ. New National Cholesterol Education Program III guidelines for primary prevention lipid-lowering drug therapy. Projected impact on the size, sex, and age distribution of the treatmenteligible population. Circulation. 2002;105: 152–156. 83. Watson KE. What is the evidence in favor of; low-dose statin therapy in 2008? Curr Atheroscler Rep. 2008;10:33–38. 84. Grundy SM, Cleeman JI, Bairey Merz CN, et al. for the Coordinating Committee of the National Cholesterol Education Program. Arterioscler Thromb Vasc Biol. 2004;24: e149–e161.
85. Rojas R, Aguilar-Salinas CA, Gómez-Pérez FJ, et al. Applicability of the National Cholesterol Education Program III (NCEP-III) Guidelines for treatment of dyslipidemia in a nonCaucasian population: a Mexican NationWide Survey. Revista Invest Clín. 2005;57: 28–37. 86. Ballantyne C, Arroll B, Shepherd J. Lipids and CVD management: towards a global consensus. Eur Heart J. 2005;26:2224–2231. 87. Manuel DG, Kwong K, Tanuseputro P, et al. Effectiveness and efficiency of different guidelines on statin treatment for preventing deaths from coronary heart disease: modeling study. BMJ. 2006;332:1419. doi:10.1136/bmj. 38849.487546.DE 88. US Department of Health and Human Services. Agency for Healthcare Research and Quality. US Preventive Services Task Force. The Guide to Clinical Preventive Services 2006. Recommendations of the U.S. Preventive Services Task Force. AHRQ Publication No. 06-0588. Washington, DC: Agency for Healthcare Research and Quality; 2006. http://www.ahrq.gov/clinic/uspstf/uspstbac .htm. Accessed October 14, 2007. 89. US Department of Health and Human Services. Cardiovascular Risk Reduction, Adults. Cholesterol Guidelines Update, ATP IV. Hypertension Guidelines Update, JNC 8. Obesity Guidelines Update, Adults. http:// www.nhlbi.nih.gov/guidelines/cvd_adult/ background.htm. Accessed June 5, 2009. 90. US Department of Health and Human Services. Pediatric Cardiovascular Risk Reduction Initiative. http://www.nhlbi.nih.gov/guidelines/ cvd_ped/background.htm. Accessed June 5, 2009.
46894_CH11_269_310.qxd
2/25/10
3:39 PM
Page 310
46894_CH12_311_360.qxd
2/25/10
3:40 PM
Page 311
C H A P T E R
12 High Blood Pressure as family history. The distribution of blood pressure may also vary over time within a population. The secular trend in blood pressure in the United States, for example, indicates a downward shift in the distribution, with reduced frequencies of extreme high values and lower median values, from the 1960s to the mid-1990s, with reversal of the trend in more recent years. Differences between populations in rates of coronary heart disease and stroke are explained to a significant degree by differences in their respective blood pressure distributions. Blood pressure is also a major factor in individual risks of these events within a population. Interventions to prevent and control high blood pressure in whole communities or to treat individuals to reduce high blood pressure have been shown to be effective and relatively affordable. Yet in the United States and throughout the world, prevalence of high blood pressure ranges from roughly 20 to 50 percent of the adult population, globally affecting nearly 1 billion persons. Monitoring of populations to track the proportions of those who have high blood pressure, are aware of it, are under treatment, and have levels considered to be “controlled” indicates that in few if any general populations are the majority of persons with high blood pressure adequately treated. Prevention of high blood pressure in the first place, beginning in childhood, has also been advocated for many years but remains to be addressed sufficiently to achieve known potential benefits. Interventions to prevent and control high blood pressure include reduction of salt content of manufactured and restaurant foods, community supports for healthy lifestyle patterns, and community- and clinicbased resources for prevention, detection, and longterm management of high blood pressure. Public
SUMMARY Blood pressure is a quantitative characteristic, continuously distributed in populations. High blood pressure or hypertension represents the more or less extreme upper part of the distribution in a population and is defined somewhat arbitrarily for purposes of classification and individual decision making in clinical practice. High blood pressure is a consequence of certain determinants and is at the same time a determinant of other cardiovascular conditions, such as coronary heart disease, heart failure, stroke, and others. Methods of measurement of blood pressure for population studies have been well standardized for many years. Clinical classification schemes for detection of high blood pressure and management of individual patients are long established in the United States, the United Kingdom, and Europe. Most commonly a substantial upward shift in the blood pressure distribution occurs with increasing age, resulting in a high proportion of the population—especially those at older ages—having high blood pressure. Populations are known, however, in which low average blood pressure levels persist across the age range, so high blood pressure is not simply a consequence of aging. Demographic differences—such as by sex, race/ethnicity, income, or education—are also associated with differences in blood pressure distributions. In the United States, for example, high blood pressure is significantly more common among African Americans than other groups, and it is less often effectively treated among Mexican Americans than others. Major factors in population and individual differences in blood pressure levels are dietary imbalance (especially salt consumption), excessive alcohol intake, physical inactivity, and obesity, as well
311
46894_CH12_311_360.qxd
312
2/25/10
3:40 PM
Page 312
CHAPTER 12 HIGH BLOOD PRESSURE
health has both the challenge and the opportunity of achieving population-wide blood pressure prevention and control and eliminating the disparities in health that accompany this pandemic condition.
INTRODUCTION High blood pressure, or hypertension, is regarded both as a cardiovascular condition in itself and as a major risk factor or determinant for atherosclerosis, coronary heart disease, stroke, heart failure, chronic kidney disease, and other cardiovascular conditions. It affects more than 70 million people in the United States and 1 billion people worldwide.1,2 The numbers of American adults affected, by sex and race/ ethnicity, are shown in Table 1-4a. The populationattributable fractions relating high blood pressure to ischemic heart disease and stroke globally—45% and 54%, respectively—are shown in Table 1-8. The magnitude of the public health problem of high blood pressure in the United States was already well recognized when, in the 1970s, it was considered that as many as 60 million Americans might be affected. One-third of them were undetected, only one-half of those detected were on treatment, and few were treated adequately.3 Important advances have taken place in the intervening decades, yet prevalence is even greater and treatment remains inadequate for the majority of those affected. Blood pressure is a critical measure of circulatory function. At a given moment it reflects the balance between the blood volume ejected from the left ventricle of the heart with each cardiac cycle and arterial resistance to blood flow, which is controlled especially by the distal vasculature. Many physiologic mechanisms operate continuously to maintain blood pressure at a level sufficient to perfuse the body tissues—an essential homeostatic function. Blood pressure varies from moment to moment throughout each cardiac cycle. It changes rapidly in response to acute physical and psychological influences. Blood pressure too low to sustain tissue perfusion constitutes a medical emergency and is part of the clinical syndrome of circulatory shock. On a longer timeline, higher-than-optimum blood pressure causes pathological changes in several organ systems, including progression of atherosclerosis and other vascular changes, especially in the kidneys, retina, heart, and brain. In most but not all populations, the frequency of high blood pressure increases throughout the life span. High blood pressure is a significant public health priority not only because it is common and its consequences are severe but also because it can be pre-
vented or when necessary treated, greatly reducing its attendant morbidity and mortality. These considerations led Epstein, 25 years ago, to conclude a review of the epidemiology of hypertension as follows:4, p 16 . . . screening for hypertensives must be improved further . . . the blood pressure of as many patients as possible be brought into a range where their risk of clinical complications is adequately reduced. At the community level, enough is known to justify every effort to prevent obesity, starting in youth, and to reduce the average consumption of salt in the population. These two measures alone are likely to lessen the degree to which blood pressure rises with age and, thus, to lower the prevalence of hypertension.
CONCEPTS AND DEFINITIONS Classification of blood pressure levels is necessary for clinical purposes, concerning diagnosis and treatment of individual patients and public health purposes, such as population screening or surveillance. Practice conventions have evolved continually over several decades, with changing concepts and definitions. Dichotomy Versus Continuity Until the latter 1900s, levels of blood pressure recognized as abnormally high were widely regarded as characterizing a distinct disease. Classically, the disease was termed “hypertension” or, in especially severe and rapidly progressive cases, “malignant hypertension.” A specific cause was identifiable in only a small proportion of cases. The first task of the physician concerned about a patient’s blood pressure was to determine whether the patient was “hypertensive” or “normal.” This discrimination required measuring the blood pressure and applying some fixed numerical criterion thought to distinguish correctly between these two categories. Critical values chosen for this purpose were often defended vigorously but varied among authorities in the field and were revised from time to time. In the 1960s, Pickering argued especially effectively that the absolute distinction between hypertensive and normal was a false dichotomy.5 Instead, he advanced the concept of blood pressure as a continuous or graded characteristic: In his view, the higher the usual blood pressure level, the greater was the risk of progressive pathological change and cardiovascular complications, and no clear dividing line separated diseased from normal persons. Although this concept has long been accepted, decisions are still needed in practice about which pa-
46894_CH12_311_360.qxd
2/25/10
3:40 PM
Page 313
CONCEPTS AND DEFINITIONS 313
tients’ blood pressure is high enough to receive clinical attention. Therefore, some aspects of the earlier terminology persist, and numerical blood pressure criteria, however arbitrary, continue to be used in making individual decisions about diagnosis and treatment. Blood pressure is usually measured at two points in its continuous cycle—the higher reading representing the pressure peak after ventricular contraction, or systole (systolic pressure), and the lower reading representing a lower point in the pressure cycle, closer to the maximum ventricular relaxation, or diastole (diastolic pressure). Current definitions and criteria usually refer to both measurements. Figure 12-1 illustrates the fact that the frequencies of blood pressure levels (here, diastolic) in a large screened population show no distinct separation between normal and high but indicate a single continuous distribution. For both systolic and diastolic pressure the distributions are typically skewed to the right, especially for older age groups.6 Also, the an-
swer to the question how large a proportion of the population has hypertension—an important public health question—is shown to depend simply on the level of pressure chosen to make the distinction. This is not a trivial point, because for every downward change of 5 units in the criterion, for example from 115 to 110 mm Hg (millimeters of mercury), the estimated proportion of the population with hypertension approximately doubles. “Essential” Versus “Secondary” Hypertension Beyond the multiple physiologic regulatory functions that determine the blood pressure level in an individual from moment to moment, many influences operate in ways that cause sustained elevation of blood pressure above the physiologic optimum. As suggested by the classifications above, optimum blood pressure is the low end of the distributions of both systolic and diastolic pressure. Higher values, and in the past extremely high values, were considered to represent either “essential hypertension” or “secondary” hypertension.
20
% of Screened Population
15 Prevalence of “Hypertension” by different cut points ≥ 90 = 25.3% 10
≥ 95 = 14.5% ≥ 100 = 8.4% ≥ 105 = 4.7% ≥ 110 = 2.9%
5
≥ 115 = 1.4%
0 50
60
70
80
90
100
110
120
130
Diastolic Blood Pressure (mm Hg)
Figure 12-1 Frequency Distribution of Diastolic Blood Pressure at Home Screen of 158,906 Persons, 30–69 Years of Age. Source: Reprinted with permission from Hypertension Detection and Follow-up Program: A Progress Report, Circulation Research, Supplement, 40, pp 1–107, © 1997, American Heart Association.
46894_CH12_311_360.qxd
314
2/25/10
3:40 PM
Page 314
CHAPTER 12 HIGH BLOOD PRESSURE
Evolution of these concepts is traced in Pickering’s classic 1961 monograph, The Nature of Essential Hypertension, in which essential hypertension is defined as presence of hypertension in the absence of any known causal mechanism, such as diseases of the kidney or endocrine disorders.5 These terms no longer appear in the clinical classification schemes. In accordance with The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure (JNC VII) (notably, the term is not hypertension), for example, once elevated blood pressure is identified on two occasions, within 1 or 2 months, limited clinical and laboratory evaluation is recommended.7 Only in the presence of specific indications is more extensive investigation for “identifiable causes of hypertension” recommended. The primary focus is on the level of blood pressure itself and its response to treatment when needed.
follow-up of men screened for the Multiple Risk Factor Intervention Trial. Because more than 350,000 men were studied, baseline blood pressure data could be stratified by quintiles of both systolic and diastolic pressure simultaneously and cross-tabulated to observe the joint risks.8 Within each of five categories of diastolic pressure, increasing levels of systolic pressure were associated with increases in risk of coronary heart disease death. Conversely, within each stratum of systolic blood pressure at baseline, risk tended to increase with increasing diastolic pressure; however, the increases were not so great nor were they quite so consistent as in the first comparison. Whether the apparent difference is real or important is unclear, as it may be an artifact of less-precise measurement of diastolic pressure. In any case, both systolic and diastolic pressure are incorporated in the definition and classification of blood pressure.
Systolic or Diastolic? Much attention has been given over the years to whether systolic or diastolic pressure is the “better” measure. Each has been shown to predict outcomes such as coronary events, stroke, and death; each is correlated with most of the same factors; and each responds to treatment to reduce risks. An important contribution to the debate is based on data from
Current Classifications in the United States and Europe One example of current terminology and criteria in the United States is shown in Table 12-1, which presents the classification of blood pressure for adults age 18 years and older. This is taken from JNC VII, published in 2003.7 Four categories, “normal,” “prehypertension,” “stage 1 hypertension,” and “stage 2
Table 12-1
Classification and Management of Blood Pressure for Adults*
SBP* mm Hg 120 120–139
DBPmm Hg and 80 or 80–89
Lifestyle Modification Encourage Yes
Stage 1 Hypertension
140–159
or 90–99
Stage 2 Hypertension
160
or 100
BP Classification Normal Prehypertension
Initial Drug Therapy Without With Compelling Compelling Indications Indications No antihypertensive drug indicated.
Drug(s) for compelling indications.‡
Yes
Thiazide-type diuretics for most. May consider ACEI, ARB, BB, CCB, or combination.
Drug(s) for the compelling indications.‡ Other antihypertensive drugs (diuretics, ACEI, ARB, BB, CCB) as needed.
Yes
Two-drug combination for most† (usually thiazide-type diuretic and ACEI or ARB or BB or CCB).
DBP, diastolic blood pressure; SBP, systolic blood pressure. Drug abbreviations: ACEI, angiotensin converting enzyme inhibitor; ARB, angiotensin receptor blocker; BB, beta-blocker; CCB, calcium channel blocker. *Treatment determined by highest BP category. † Initial combined therapy should be used cautiously in those at risk for orthostatic hypotension. ‡ Treat patients with chronic kidney disease or diabetes to BP goal of 130/80 mm Hg. Source: JNC Express. The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation and Treatment of High Blood Pressure, National Heart, Lung and Blood Institute, National Institutes for Health, NIH Publication No. 03-5233, May 2003, p 3.
46894_CH12_311_360.qxd
2/25/10
3:40 PM
Page 315
CONCEPTS AND DEFINITIONS 315
hypertension,” are used. Of present interest is the use of dual criteria incorporating both systolic and diastolic pressure, and introduction of the term “prehypertension.” Either systolic or diastolic pressure, whichever indicates the higher category, is decisive in classifying an individual. Prehypertension identifies a stratum of the population at increased risk of reaching stage 1 hypertension. This classification and the terms employed to describe it reflect current distinctions among blood pressure levels at which treatment is warranted, levels at which the risk of progression to hypertension is high, and levels that may be considered normal or optimal. More recently, the European Society of Hypertension and European Society of Cardiology published new guidelines that reflect differences in approaches to classification.9 In comparison with JNC VII, “optimal” blood pressure corresponds to “nor-
Table 12-2
mal”; “normal” and “high normal” divide “prehypertension”; and “Grade 1,” “Grade 2,” and “Grade 3” hypertension represent “Stage 1” and “Stage 2” hypertension. The European classification also distinguishes “isolated systolic hypertension,” in which only the systolic pressure is elevated, whereas the diastolic pressure is not. In particular, the term “prehypertension” as used in JNC VII was not included. Among children and adolescents, blood pressure levels increase but on average remain below those found in adults. Classification of blood pressure levels in this age range is based not on absolute values, but rather on relative rank of an individual within the blood pressure distribution of his or her peers. The current scheme in the United States is as presented in The Fourth Report on the Diagnosis, Evaluation, and Treatment of High Blood Pressure in Children and Adolescents (Table 12-2).10 Reference distributions of
Classification of Hypertension in Children and Adolescents, with Measurement Frequency and Therapy Recommendations SBP or DBP Frequency of Therapeutic Pharmacologic Percentile* BR Measurement Lifestyle Changes Therapy
Normal
90th
Recheck at next scheduled physical examination.
Encourage healthy diet, sleep, and physical activity.
—
Prehypertension
90th to 95th or if BP exceeds 120/80 mmHg even if below 90th percentile up to 95th percentile†
Recheck in 6 months.
Weight-management counseling if overweight, introduce physical activity and diet management.‡
None unless compelling indications such as CKD, diabetes mellitus, heart failure, or LVH exist.
Stage 1 hypertension
95th percentile to the 99th percentile plus 5 mmHg
Recheck in 1–2 weeks or sooner if the patient is symptomatic; if persistently elevated on two additional occasions, evaluate or refer to source of care within 1 month.
Weight-management counseling if overweight, introduce physical activity and diet management.‡
Initiate therapy based on indications in Table 6 or if compelling indications as above.
Stage 2 hypertension
99th percentile plus 5 mmHg
Evaluate or refer to source of care within 1 week or immediately if the patient is symptomatic.
Weight-management counseling if overweight, introduce physical activity and diet management.‡
Initiate therapy.
BP, blood pressure; CKD, chronic kidney disease; DBP, diastolic blood pressure; LVH, left ventricular hypertrophy; SB, systolic blood pressure *For sex, age, and height measured on at least three separate occasions; if systolic and diastolic categories are different categorize by the higher value. † This occurs typically at 12 years old for SBP and at 16 years old for DBP. ‡ Parents and children trying to modify the eating plan to the Dietary Approaches to Stop Hypertension (DASH) eating plan could benefit from consultation with a registered or licensed nutritionist to get them started. § More than one drug may be required. Source: The Fourth Report on the Diagnosis, Evaluation, and Treatment of High Blood Pressure in Children and Adolescents, National Heart, Lung and Blood Institute, National Institute of Health, NIH Publication No. 05-5267, May 2005, p 14.
46894_CH12_311_360.qxd
316
2/25/10
3:40 PM
Page 316
CHAPTER 12 HIGH BLOOD PRESSURE
systolic and diastolic blood pressure by sex, year of age, and height are used for determining the 90th, 95th, and 99th percentiles to distinguish among the four categories termed in the same manner as for adults. An individual’s blood pressure levels (systolic and diastolic, separately) are compared against the reference percentile values, and he or she is classified according to the higher category of the two readings.
MEASUREMENT Technique of Indirect Auscultatory Determination Blood pressure measurement became widely applicable as a diagnostic procedure after a report by the Russian Korotkov in 1905.11 His method was to surround the arm with an inflatable cuff that, with sufficient pressure, would occlude the underlying brachial artery (the major artery to the forearm). He described the changes in sounds, or auscultatory phenomena, heard by means of a stethoscope applied over the brachial artery as the pressure is released and blood flow returns to the artery. His apparatus combined the stethoscope of von Recklinghausen, the cuff of Riva-Rocci, and the mercury manometer of Poiseuille (hence the convention of millimeters of mercury, or mm Hg, as the units of measure of blood pressure). His method linked the changes in sounds with the successive pressure levels that can be read from the scale on the manometer. Unlike prior laboratory methods, it required no insertion of a needle directly into the artery to make the measurement. This noninvasive method of blood pressure measurement, the “indirect auscultatory method,” was simple, could be performed easily in routine examinations, and quickly became the standard technique. The so-called “Korotkov sounds” identify systolic pressure at the first “phase,” when sounds accompanying the beginning of blood flow are heard, and two levels of diastolic pressure, related to his fourth and fifth phases, when sounds first change character to become “muffled” in quality and then disappear as cuff pressure no longer constricts flow in the brachial artery. The American Heart Association has published guidelines for blood pressure measurement from time to time since 1939 and has variously recommended use of the fourth or the fifth phase Korotkov sound as the basis for reading the diastolic pressure. In adults, fifth-phase pressure has been favored recently.7 Special concerns about its interpretation in some readings in children led to use of fourth-phase diastolic pressure for those younger than age 14, until appearance of a 1996 update on recommendations for blood pressure measurement in
children and adolescents, calling for use of fifth-phase readings.12 There is typically a difference of several millimeters of mercury between these two values. Alternative Techniques Other techniques and devices have been introduced since the time of Korotkov. Aneroid manometers are less reliable but are in common use as an alternative to the mercury type, which remains the standard for calibration. Electronic devices of several kinds are used where they have advantages due to special measurement requirements, such as in intensive-care settings, measurements in newborns and infants, and ambulatory monitoring for clinical or research purposes, in which blood pressure measurements may be recorded very frequently throughout a period of usual activity for 24 hours or longer. For reasons of practicality and consistency with long-established methods, Korotkov’s indirect auscultatory method with a mercury manometer was generally preferred throughout several decades of epidemiologic research. A variant on this approach is a device that blinds the observer to the true zero-pressure level of mercury in the manometer column until the reading has been completed. This is the “zero-muddler” or “randomzero” device. It has been used in many studies in which observer bias may be a particular concern, such as in trials of high blood pressure prevention or treatment. A serious concern has arisen recently in the United States over threatened abandonment of the mercury sphygmomanometer under regulatory policies concerning exposure to mercury. Loss of the reference method for calibration of alternative devices risks deterioration of accuracy of this critical clinical measurement.13 The continued practice of reporting blood pressure readings, regardless of method, in units of mmHg, may become purely symbolic. Standardization of Procedures The most significant advances in blood pressure measurement for epidemiologic research and public health practice have been those in training of blood pressure observers and standardizing the circumstances of measurement.14,15 These approaches are aimed at eliminating sources of incidental variation in measurements. Such variation may compromise withinor between-person reliability and comparability of data or impair entire research projects. The goal is to attain the highest standards of data collection. Therefore, study protocols often address in some detail the provisions for training and certification of blood pressure observers (such as by use of quantitative testing by videotape methods); conditions of measurement (such as a resting period, quiet, proper
46894_CH12_311_360.qxd
2/25/10
3:40 PM
Page 317
DETERMINANTS 317
relative positions of the subject and equipment, voiding in advance, and others); and numbers of readings to be obtained (such as a fixed number to be averaged on each occasion and possibly multiple occasions of examination). Some issues remain despite these efforts at standardization, such as the awkward possibility that an individual may respond psychologically to the blood pressure measurement situation and exhibit higher than usual blood pressure. This phenomenon has often been associated with the physician as the observer and is thus designated “white coat hypertension,” the physician’s professional uniform being emblematic of the problem. Concern about this problem as well as recognized variability of blood pressure throughout the day and night has stimulated a substantial body of work in ambulatory 24-hour monitoring of blood pressure in clinical research and its use in special circumstances of patient care.16
DETERMINANTS As noted above, high blood pressure is regarded as a cardiovascular condition in much the same sense as atherosclerosis and the others discussed in Part II. While at the same time operating as a major risk factor for those conditions, it shares with them many of the same determinants. For these reasons, discussion of high blood pressure and its determinants here
overlaps with several other chapters in Part III. In general, these overlapping topics are addressed only briefly here and are cross-referenced to the corresponding chapters. The major exception is nutrition, and specifically salt, which is central to the topic of high blood pressure. Age, Sex, and Race/Ethnicity Figure 12-2a and b shows the cross-sectional patterns of blood pressure typical of Western industrial societies in recent decades, illustrated from the National Health and Nutrition Examination Survey of the United States, 1988–1991.17 These patterns by age indicate generally increasing systolic blood pressure; initially increasing, then decreasing diastolic blood pressure; mixed patterns in blood pressure levels by race/ethnicity, excepting generally higher levels for non-Hispanic Blacks than others; and somewhat lower levels by sex for younger women and for older men. Populations have long been known to differ in the degree of blood pressure increase with age in adulthood. High blood pressure is virtually absent among adults in some populations and very common in others.18 Population differences in the increase of blood pressure with age and factors associated with these differences were investigated in the INTERSALT Study of 200 men and women age 20–59 years in each of 52 populations from 32 countries around the world.19 The selected populations demonstrated the same wide
Women
Men 150
150
Systolic blood pressure
Systolic blood pressure 130
130
Non-Hispanic Black Non-Hispanic White Mexican American
110
mm Hg
mm Hg
110
80
80
70
A
Diastolic blood pressure
70
0 18–29
Non-Hispanic Black Non-Hispanic White Mexican American
30–39
40–49
50–59 60–69 Age, Year
70–79
≥80
B
Diastolic blood pressure
0 18–29
30–39
40–49
50–59 60–69 Age, Year
70–79
≥80
Figure 12-2 Mean Systolic and Diastolic Blood Pressure by Age and Race/Ethnicity, US Men (Panel A) and Women (Panel B) 18 Years and Older. Source: Reprinted with permission from VL Burt et al., Hypertension, Vol 25, p 309, © 1995, American Heart Association.
46894_CH12_311_360.qxd
318
2/25/10
3:40 PM
Page 318
CHAPTER 12 HIGH BLOOD PRESSURE
range of slopes of increasing systolic blood pressure by age previously described, and the slopes were related to several characteristics of these populations, as shown in Table 12-3.20 The range of slopes coded from 0 to 4 corresponds to increases in systolic pressure from 0 to 0.67 mm Hg per year of age as a population average; slope 4 represents an average increase of nearly 7 mm Hg over each 10-year age interval, such as from age 30 to age 40. In the INTERSALT Study, 24-hour urine collections were the basis for estimating daily excretion of sodium and potassium, as surrogates for their dietary intake; body mass index (BMI) was calculated from measured weight and height; and alcohol intake was assessed by interview. Sodium excretion was positively and potassium excretion inversely related to blood pressure slope; body mass index was associated with slopes of 2 and greater; and blood pressure slope increased with greater average alcohol intake. For children and adolescents, the situation is different. Populations around the world are more alike in their patterns of increasing blood pressure with age. One example, from a mixed longitudinal study in one United States community (Project HeartBeat!, cited earlier), illustrates age-related changes in systolic and fourth- and fifth-phase diastolic blood pressure from age 8 to 18 years (Figure 12-3).21 Age patterns are clear and distinct for each blood pressure measure. Statistically significant differences occur by sex for systolic pressure, by sex and race/ethnicity for fourth-phase diastolic pressure, and by race/ethnicity for fifth-phase diastolic pressure. Determinants of these different trajectories of change in blood pressures are not yet well understood. Regarding race and ethnicity, there are marked disparities in the occurrence of high blood pressure, differentially afflicting African Americans. A review of racial differences in blood pressure in the United States indicates that this was recognized as early as the 1930s, and it persists to the present.22 Explanations
Table 12-3 Step 0 1 2 3 4
N 393 929 2695 3300 2762
have been sought through decades of research and have included selective factors related to slavery, sociocultural factors related to subjugation and racism, differences in income and education, nutrition, and others. One area of recent interest has been the development of blood pressure in relation to birth weight and subsequent growth into adolescence or early adulthood. In the Perinatal Collaborative Project, for example, childhood growth but not birth weight was found to predict early adult blood pressure in Blacks.23 By contrast, the Bogalusa Heart Study found average birth weight to be lower among groups of Black than White infants; this difference accounted for higher blood pressure levels among Blacks at ages 15–17 years.24 A subsequent report from that study indicated continuation of the inverse relation of birth weight to blood pressure into early adulthood.25 The inverse association was similar for individual Blacks and Whites. These and other findings on birth weight and blood pressure, in general or between racial or other groups, continue to stimulate interest in the “fetal origins hypothesis” discussed further in Chapter 16, “Social and Physical Environment.” Family History A 2007 review of genetic aspects of hypertension and other cardiovascular conditions, noted in Chapter 7, “Genes and Environment,” emphasized the importance of family history.26 Table 12-4 demonstrates the frequency of a positive family history of high blood pressure, and the relative risk of having high blood pressure at a given age, under two definitions of positive family history.27 If the criterion were having one first-degree relative (parent, sibling, or offspring) with high blood pressure at any age, 55 percent of persons were positive and the relative risk of having high blood pressure was 2.3 at ages 20–49 years and 1.3 at ages 50–69 years, compared with persons with a negative family history. A second criterion, having two first-degree relatives affected be-
Mean Values of Urinary Sodium and Other Factors in Relation to Slope of Systolic Blood Pressure Increase with Age Na, mmol/24 h K, mmol/24 h Na-K Ratio BMI, kg/m2 Alcohol, ml/wk 6.6 75.3 0.1 22.3 0.0 133.7 53.6 2.8 22.2 74.0 154.1 56.4 3.0 25.1 115.1a 159.3 57.2 3.1 25.8 126.0 182.7 49.2 4.2 25.0 130.0
Note: BMI, body mass index. a Mexico was excluded from the alcohol analysis because of an unusual high consumption during data collection because of a holiday. Source: Reprinted with permission from BL Rodriguez et al., Hypertension, Vol 24, p 784, © American Heart Association.
46894_CH12_311_360.qxd
2/25/10
3:40 PM
Page 319
DETERMINANTS 319
Males Females
70 DBP4 (mmHg)
SBP (mmHg)
115 110 105 100 95
68 66 Black Males Non-Black Males Black Females Non-Black Females
64 62 60
8
10
14 12 Age (Year)
16
18
8
10
12 14 Age (Year)
(a)
18
(b) 64
DBP5 (mmHg)
16
Black Non-Black
62 60 58 56 54 8
10
14 12 Age (Year)
16
18
(c)
Figure 12-3 Trajectories of Change in Blood Pressure, Project HeartBeat!, 1991–1995. Source: Reprinted with permission from Coronary Heart Disease Epidemiology: From Aetiology to Public Health, M Marmot, P Elliott eds, D Labarthe, p 598. © 2005, Oxford University Press.
fore age 55, was met in a much smaller proportion (11 percent) of the study population, and the relative risk for high blood pressure at an early adult age was 3.8, nearly 1.5 times as great as under the first definition. These findings illustrate that predictability of high blood pressure from family history can be strong and is dependent on stringency of the criteria. Salt The relation of nutrition to cardiovascular conditions is addressed mainly in Chapter 8, “Dietary Imbal-
ance.” The importance of salt for blood pressure requires fuller discussion here. The relation between salt and blood pressure has been studied for more than a century. One of many sources of the history of salt and knowledge of its relation to hypertension is MacGregor’s Salt, Diet & Health, a 1998 publication whose flavor is suggested by the subtitle: Neptune’s Poisoned Chalice: The Origins of High Blood Pressure.28 Other noteworthy reviews of the background and epidemiologic evidence include those found in Diet and Health,29 an extensive account by
Table 12-4
Frequency of a Positive Family History of High Blood Pressure (FHx HBP) and Relative Risk of Future Hypertension for First-Degree Relatives According to Age Relative Risk for HBP Definition of ⴙ FHx HBP Frequency Ages 20–49 Ages 50–69 1 One or more first-degree 55% of relatives with HBP at any age all persons 2.3 1.3 2 Two or more first-degree relatives with high blood pressure before age 55
11% of all persons
3.8
1.4
Data from 96,518 adult relatives of 7625 Utah students reported by Hunt et al. [4]. Source: Reprinted with permission from AP Simopoulos, B Childs eds, Genetic Variation and Nutrition, World Rev Nutr Diet, Vol 63, RR Williams et al., p 120. Basel Karger, 1990.
46894_CH12_311_360.qxd
320
2/25/10
3:40 PM
Page 320
CHAPTER 12 HIGH BLOOD PRESSURE
Stamler,30 a National Heart, Lung and Blood Institute Workshop on Sodium and High Blood Pressure,31 a broad-based review emphasizing the exceptional experience of Finland,32 and a major report from the World Health Organization released in 2007.33 This history was also reviewed by Blackburn and Prineas from an evolutionary-anthropological perspective, including public health implications of the evidence assembled at that time, 25 years ago.34 Dietary aspects addressed were calorie balance and obesity; sodium, potassium, and alcohol intake; and water supply. Relationships of diet to blood pressure in several populations, and results of trials reducing salt intake in individuals, led to the conclusion that “The observational and experimental data suggest the potentially important influence in prevention of hypertension in the first place of multiple strategies achieving a diminution of salt use and frequency of overweight.”34, p 62 Key recommendations focused on reducing salt intake, with an “ideal” population median intake of 3 g/day, “desirable” intake of 5 g/day, and immediately “feasible” intake of 8 g/day. An extensive body of evidence regarding relationships between salt intake and blood pressure establishes answers to several distinct but interrelated questions, for example: • Does salt intake cause a rise in blood pressure in individuals, either within the desirable range or from desirable to undesirable levels? • Does salt intake cause persistence or progression of high blood pressure after its initiation by other, unknown, causes? • Are population differences in the increase of blood pressure with age (blood pressure “slope”) associated with average salt intake of populations? • Does reduction of salt intake result in lower blood pressure levels in persons with either high blood pressure or blood pressure within the desirable range? • Will reduction in salt intake at the population level reduce the incidence of high blood pressure in the first place? The last two questions are of greatest importance from the public health perspective because judgment on this issue forms the basis of policy recommendations, including regulatory action on production and labeling of processed foods—the major source of dietary sodium in most industrialized societies. Here lies a major reason for controversy, which has less to do with scientific than with policy conflict, as witnessed in mid-1996 in what might be termed the “salt
issue” of the British Medical Journal (see Godlee35 and companion articles). That issue presents an unusual view of open conflict between scientific and commercial interests, given that a series of articles and commentaries exchange allegations of withholding of data, improper secondary analysis of published data, and failure of government agencies to resist pressure from industry in setting health policies. These issues are addressed more recently through advocacy groups in the United States, the United Kingdom, and elsewhere (see Prevention and Control, as follows). Dietary intake of salt, or sodium, is difficult to determine by interview methods because it is contained in many food sources; content in processed foods varies according to particular manufacturing techniques and in restaurant or home cooking according to particular recipes or habits; and salt may or may not be added by the consumer at the table. However, urinary excretion of sodium can serve as an estimate of intake. This is because equilibrium in the body’s content of sodium is maintained mainly through renal excretion; on the assumption of a steady physiologic state, sodium excretion provides an estimate of sodium intake on a day-to-day basis. Therefore, 24hour urinary collections are conducted for this purpose in population studies, ideally on more than a single occasion to improve reliability of the estimate. Note: Dietary intake of “salt” is expressed as quantities of either sodium or sodium chloride. Other sodium-containing compounds are often present in foods or medications, usually in relatively small amounts. Various measures of sodium (Na) and salt (NaCl) are in common use. The atomic weight of Na 23; 1 grammolecular weight (mole) Na 23 grams (g); 1 millimole (mmol) Na 23 mg; approximate daily requirement for intake of Na 10 mmol or 230 mg29; recommended daily intakes36 2300 mg (100 mmol) or less, or up to 10 times the daily requirement, excepting people with hypertension, African Americans, and middleaged or older adults with recommended intake 1500 mg or less, or up to 6.5 times the daily requirement. Atomic weight of chlorine (Cl) 35; molecular weight of NaCl 23 35 58; equivalent amounts of NaCl for the recommended daily intakes of 2300 and 1500 mg Na 5800 and 3750 mg, or 5.8 and 3.75 g, respectively, of NaCl or approximately 1 teaspoon (tsp) and 2/3 tsp of salt. Populations have been found with lower average intakes of sodium than the cited 8–10 mmol/day requirement. Intakes usually exceed this requirement
46894_CH12_311_360.qxd
2/25/10
3:40 PM
Page 321
DETERMINANTS 321
by many times, with the median intake among 48 centers of the INTERSALT Study being 160 mmol/ day (3680 mg Na).18 The major source of dietary sodium in the United States, perhaps 75% of the total intake, is in commercially processed foods.37 Therefore, the potential for reduction in average population intakes of sodium depends on the availability and use of lower-sodium products. The food industry is said to resist reduction in salt content because of loss of flavor, and therefore marketability, of its products.35 Salt taste is remarkably modifiable, however, and adaptation in only a matter of weeks can make many previously desirable foods seem unpalatable because of saltiness. The INTERSALT Study, described previously, estimated the public health benefit of reduced sodium intake by calculating the contribution of sodium intake to the cross-sectional slope of blood pressure increase with age in its many study populations.30 If habitual sodium intake were reduced by 100 mmol/ day, as from 170 to 70 mmol (approximately from 3900 to 1600 mg Na), the age-related increase in systolic blood pressure would be 10.2 mmHg less and in diastolic pressure 6.3 mmHg less, from age 25 to 55 years. The effect of these reductions in average later adult blood pressure levels would be a substantial reduction in incidence and prevalence of high blood pressure in the population. The further impact of a reduction in hypertension on mortality from cardiovascular diseases and all causes is shown in Table 12-5. For each of several degrees of reduction in systolic blood pressure—2, 5, or 9 mm Hg as population averages (all less than that projected above for a de-
crease of sodium consumption by 100 mmol/day, or 2300 mg Na/day)—the percentage reduction in mortality and number of deaths averted in the US male population aged 35–54 years are presented. The decrease of 9 mm Hg corresponds to mortality reductions of 18.0% for coronary heart disease, 26.2% for stroke, 19.4% for all cardiovascular disease, and 13.7% for all-cause mortality. The impact at older ages, where rates are much higher, would be even greater. Other Nutrients In a 2003 review of macronutrients and dietary patterns in relation to hypertension, Appel and Elliott concluded that a number of changes in dietary patterns could lower blood pressure levels, whereas for others evidence was insufficient.38 Approaches that were considered to be well supported were the DASH diet or a vegetarian diet, high intake of omega-3 polyunsaturated fat, and perhaps increased intake of vegetable protein. To study the relation of blood pressure to multiple macronutrients, micronutrients, and protein, the INTERMAP Study was undertaken by extension of the INTERSALT model of international collaboration and study organization.39 On the basis of four repeated 24-hour dietary recall interviews for each participant, blood pressure differences among more than 4500 individuals from China, Japan, the United States, and the United Kingdom are being studied in relation to intake of 76 nutrients. From two 24-hour urine samples collected for each participant, analyses include urinary excretion of sodium, potassium, calcium, magnesium, creatinine and amino
Table 12-5
Reduction in Population Average Systolic Blood Pressure and Expected Reductions in Deaths in Six Years by Cause, US Men Aged 35–54 Expected Decreases in Death Ratesa (and Number of Deaths) for 26,024,000b US Men Aged 35–54 Reduction in Population Average Systolic Blood Pressure Disease (mmHg) Coronary Heart Disease Stroke Cardiovascular All Causes 2 4.3% 6.5% 4.7% 3.2% (6,870) (1,093) (10,279) (16,603) 5 10.5% 15.5% 11.3% 7.9% (16,785) (2,602) (24,723) (40,962) 9 18.0% 26.2% 19.4% 13.7% (28,757) (4,398) (42,471) (71,019) a
Based on multiple logistic coefficient (with baseline systolic blood pressure, age, serum cholesterol, cigarettes per day, and diabetes in the model) for the cohort of 347,978 men screened for the Multiple Risk Factor Intervention Trial, from 6-year follow-up data. b US male population aged 35–54 in 1984. Source: Reprinted with permission from J Stamler, Dietary Salt and Blood Pressure, Annals of the New York Academy of Science, Vol 676, p 145, © 1993, New York Academy of Science.
46894_CH12_311_360.qxd
322
2/25/10
3:40 PM
Page 322
CHAPTER 12 HIGH BLOOD PRESSURE
acids, and microalbuminuria. Multiple nutrients and metabolites are being studied in urinary samples by nuclear magnetic resonance and high-pressure liquid chromatography. One of many results published to date indicates a consistent inverse association between vegetable protein intake and blood pressure, adding weight to the tentative conclusion Appel presented previously.40 Migration From a social science perspective, interest in cardiovascular conditions has led to studies of migrant populations, or the process of migration, at least since the 1960s. This literature is discussed in Chapter 16, “Social and Physical Environment,” and has been of particular interest in connection with blood pressure. In general, surveys of cross-sectional design have been conducted in which migrant groups have been surveyed, often in comparison with nonmigrant groups at the place of origin. Such studies suffer the limitation that effects of self-selection to migrate cannot be taken adequately into account. Longitudinal studies are much preferred, in which persons may be examined before and after migration, perhaps in comparison with nonmigrants in both the origin and destination locations. This approach has been applied, for example, in studies of Tokelau Islanders in the southwestern Pacific and the Yi population in Sichuan, China.41 Study of migration within the Luo population of Kenya exemplifies the ideal design, in which at the outset both rural- and urban-living individuals were examined in cross-sectional comparisons, revealing higher prevalence of hypertension in the urban area.42 Then a longitudinal component of the study was implemented with repeated examination at 3, 6, 12, 18, and 24 months for persons actively migrating into the urban area as well as members of a cohort remaining in the rural area.43 Multiple 24-hour dietary recall histories and 24-hour urine collections were included with well-standardized blood pressure measurements. Migrants developed increased urinary Na/K ratio, pulse rate, weight, and systolic blood pressure. The blood pressure changes were evident from the first measurement onward. A consistent impression from such studies is that the personal, social, and broader environmental changes associated with migration can result in higher blood pressure among migrants than among nonmigrants and can develop within weeks to months. Dietary changes are strongly implicated, including salt intake and energy balance, with resulting weight gain. Psychosocial stress may also play a role.
MECHANISMS Increase in Blood Pressure Among factors leading to increase in blood pressure beyond optimum levels, nutrition and specifically sodium intake have been emphasized, and other determinants—or more generally, associated characteristics—have also been recognized as described previously. Regarding mechanisms by which these factors or exposures increase blood pressure levels, explanations of two kinds are commonly offered. First, imbalances between intakes and physiologic requirements for sodium and potassium affect cellular function in the kidneys, which are critical for regulatory systems controlling blood pressure. The precise mechanism of these effects is unknown. Other dietary factors investigated in the past include calcium, alcohol, polyunsaturated to saturated fat ratio, protein, fiber, and others.28,29 The broad scope of the INTERMAP study regarding nutrition and blood pressure was noted previously.39 Second, psychosocial factors may be mediated by their known effects on endocrine function, by which heightened catecholamine activity results in increased arterial resistance and consequently elevated blood pressure. This can occur in response to acute stimuli, although it has not been established that sustained elevation of blood pressure results from this process.44 This mechanism is cited as relevant to observations of increased blood pressure after migration, but the multiple concurrent changes in dietary and other factors preclude clearly distinguishing such an effect.29 Regarding the first line of explanation, the concept of salt sensitivity has been proposed to explain variation among individuals in their response to experimental increase or decrease in sodium intake. Although animal models of genetic susceptibility to sodium-induced hypertension have been investigated extensively, less convincing evidence has resulted from human studies—generally addressing only acute changes in short-term experiments. This evidence is not persuasive of an important or practical distinction to be made for clinical or public health purposes.29 Effects of High Blood Pressure How does the presence of successively higher blood pressure levels or, in the extreme, high blood pressure, lead to serious complications such as myocardial infarction, stroke, congestive heart failure, and other outcomes? The Hypertension Primer is a compendium of short essays on laboratory, clinical, and epidemiologic aspects of blood pressure, including several con-
46894_CH12_311_360.qxd
2/25/10
3:40 PM
Page 323
DISTRIBUTION 323
tributions on mechanisms that collectively produce these results.45 Prominent among these are the mechanical effects of sustained high pressure in the arterial and capillary circulation, with direct consequences damaging blood vessels particularly in the major “target organs”: heart, brain, kidneys, and eyes, including, for example, accelerated atherosclerosis, cardiac enlargement, or cerebrovascular hemorrhage. In addition, a regulatory dysfunction affecting blood pressure may have other effects. For example, the catecholamines epinephrine and norepinephrine may act to increase arterial resistance, contributing to high blood pressure, and may also have a direct deleterious effect on heart muscle cells.44 Alternatively, structural and functional change to the heart resulting from high blood pressure may in turn impair its ability to respond to increased oxygen demand and exacerbate the risk of muscle injury in the presence of a relative decrease in oxygen supply, as in an acute thrombosis or plaque rupture.46
DISTRIBUTION High blood pressure is the commonest cardiovascular condition in the United States, Europe, and many other populations throughout the world. Blood pressure distributions differ by age, race/ethnicity, and sex within a population, as illustrated in Figure 12-2a and b. Prevalence of high blood pressure varies correspondingly by whatever criterion is applied. Prevalence may or may not be estimated on the basis of blood pressure measurement alone. Both systolic and diastolic pressure are usually taken into account, as shown in Table 12-1, in which “hypertension” includes either or both of systolic pressures of 140 mm Hg or greater or diastolic pressures of 90 mm Hg or greater. However, persons who report current use of antihypertensive medication may not have elevated blood pressure by measurement—that is, their blood pressure is “controlled.” Because such treatment has become common in many populations, it is important to include such persons in estimates of prevalence. Therefore, “high blood pressure” is often defined as meeting blood pressure criteria and/or reporting use of antihypertensive medication. Recently, persons who report having been told by health professionals of having high blood pressure on two or more occasions may also be included as hypertensive, regardless of their actual blood pressure levels or medication status. Interpretation and comparison of prevalence estimates or trends must take these differences in definition, as well as consistency of blood pressure measurement procedures, into account.
Often, prevalence is reported not only in terms of the proportion of the total population that is affected but also the proportions of that total who are aware of having high blood pressure, are receiving treatment for it (specifically with medications), and are controlled—that is, have blood pressure levels below the criterion for high blood pressure. These are the proportions of those with high blood pressure who are classified as “aware, treated, and controlled.” Apart from changes in measurement, blood pressure criteria, or definitions, trends in prevalence of high blood pressure over time may reflect changes in underlying determinants of blood pressure levels or in frequency and effectiveness of policies and practices regarding prevention, treatment, and control. Incidence of high blood pressure, the rate at which new cases of high blood pressure are arising in a population, is assessed much less often than prevalence. Prevalence and Trends United States Prevalence of high blood pressure among adults in the United States is presented in Table 12-6.47 For three periods of NHANES, data are shown by sex, race/ethnicity, poverty status, and age for prevalence of both hypertension, defined by blood pressure level or reported medication, and elevated blood pressure, based on measurement alone. For example, in 2003–2006, 31.3% of adults aged 20 years or older (more than 70 million persons) had hypertension, and a subset of 17.9% had elevated blood pressure; by inference, the difference of 13.4% is accounted for by persons who were classified as hypertensive but were controlled (systolic pressure below 140 mm Hg and diastolic pressure below 90 mm Hg). This means that nearly one-fifth of adults have elevated blood pressure that is not being treated or is treated insufficiently to be controlled. Data for hypertension in 2003–2006, ageadjusted for comparability across groups, indicate that nearly one-third of all adults had high blood pressure, about equally between males and females. Prevalence was greatest among African Americans and least among persons of Mexican origin. Prevalence was notably higher for groups below 100% or 200% of poverty level according to family size and income. The age distribution of hypertension shows sharp increases in prevalence, already approaching 10% by age 20–34 years and increasing to 65% or more at age 75 or older; prevalence was greater for males younger than age 45 and greater for females at ages 55 years and older.
20–34 years 35–44 years
Male
Both sexes5 Male Female Not Hispanic or Latino: White only, male White only, female Black or African American only, male Black or African American only, female Mexican male Mexican female Percent of poverty level:6 Below 100% 100%–less than 200% 200% or more 30.3 34.8 28.2
25.7 26.7 22.2
*8.1 17.1
28.3 32.8 35.9 41.9 16.5 18.8
24.3 24.6 31.1 32.5 16.4 15.9
7.1 17.1
30.2 27.6 32.7
33.9 33.5 28.2
31.7 26.6 23.9
24.1 23.8 24.4
27.6 28.5 40.6 43.5 26.8 27.9
25.6 23.0 37.5 38.3 26.9 25.0
9.2 21.1
28.8 36.8 31.1
32.4 33.4 38.8 42.8 16.6 20.0
32.1 31.3 32.9
35.0 34.1 30.3
31.2 28.3 42.2 44.1 24.8 28.6
31.3 31.8 30.3
6.6 15.2
18.7 19.8 16.2
18.7 16.4 25.5 22.2 13.9 12.7
17.6 18.7 16.5
22.5 19.3 17.5
19.7 15.1 30.3 26.4 22.2 20.4
18.5 20.6 16.4
*7.3 12.1
21.1 24.1 17.8
17.8 21.6 25.2 27.2 14.1 13.8
19.9 18.2 21.6
23.3 23.0 18.2
17.6 18.5 28.2 28.8 21.5 21.2
19.9 19.1 20.2
7.6 13.2
18.3 22.5 16.8
17.9 18.8 24.8 22.4 10.9 13.0
18.2 17.9 18.6
22.6 21.1 16.6
17.4 15.9 26.5 23.9 15.3 19.2
17.9 18.2 17.3
3:40 PM
20 years and over, crude
30.0 28.8 30.6
25.5 26.4 24.4
Elevated Blood Pressure2 1999–2002 2003–2006
2/25/10
Both sexes5 Male Female Not Hispanic or Latino: White only, male White only, female Black or African American only, male Black or African American only, female Mexican male Mexican female Percent of poverty level:6 Below 100% 100%–less than 200% 200% or more
Sex, Age, Race and Hispanic Origin, and Percent of Poverty Level 20 years and over, age-adjusted4
Hypertension2.3 (Elevated Blood Pressure and/or Taking Antihypertensive Medication) 1988–1994 1999–2002 2003–2006 1988–1994 Percent of Population
324
1
Hypertension and Elevated Blood Pressure Among Persons 20 Years of Age and Over, by Selected Characteristics: United States, 1988–1994, 1999–2002, and 2003–2006 [Data are based on interviews and physical examinations of a sample of the civilian noninstitutionalized population]
Table 12-6
46894_CH12_311_360.qxd Page 324
CHAPTER 12 HIGH BLOOD PRESSURE
See footnotes at end of table.
2.9 11.2 23.9 42.6 56.2 73.6
*2.7 15.1 31.8 53.9 72.7 83.1
31.0 45.0 59.6 69.0 *2.2 12.6 36.2 54.4 70.8 80.2
36.2 50.2 64.1 65.0 *2.4 6.4 13.7 27.0 38.2 59.9
21.9 28.4 39.9 49.7 *1.4 8.5 19.1 31.9 53.0 64.4
20.4 24.8 34.9 50.6
* 5.8 20.0 28.6 40.8 55.4
21.0 26.4 29.2 38.2
Data from Health, United States, 2008, National Center for Health Statistics, pp 312–313.
Source: CDC/NCHS, National Health and Nutrition Examination Survey.
3:40 PM
*Estimates are considered unreliable. Data received by an asterisk have a relative standard error of 20%–30%. Data not shown have an RSE greater than 30%. 1 Persons of Mexican origin may be of any race. Starting with 1999 data, race-specific estimates are tabulated according to the 1997 Revisions to the Standards for the Classification of Federal Data on Race and Ethnicity and are not strictly comparable with estimates for earlier years. The two non-Hispanic race categories shown in the table conform to the 1997 Standards. Starting with 1999 data, race-specific estimates are for persons who reported only one racial group. Prior to data year 1999, estimates were tabulated according to the 1977 Standards. Estimates for single-race categories prior to 1999 included persons who reported one race or, if they reported more than one race, identified one race as best representing their race. See Appendix II. Hispanic origin, Race. 2 Hypertension is defined as having measured elevated blood pressure and/or taking antihypertensive medication. Elevated blood pressure is defined as having a measured systolic pressure of at least 140 mmHg or diastolic pressure of at least 90 mmHg. Those with elevated blood pressure also may be taking prescribed medicine for high blood pressure. Those taking antihypertensive medication may not have measured elevated blood pressure but are still classified as having hypertension. See Appendix II, Blood pressure, elevated. 3 Respondents were asked. “Are you now taking prescribed medicine for your high blood pressure?” 4 Age-adjusted to the 2000 standard population using five age groups: 20–34 years, 35–14 years, 45–54 years, 55–64 years, and 65 years and over (65–74 years for estimates for 20–74 years). Age-adjusted estimates may differ from other age-adjusted estimates based on the same data and presented elsewhere if different age groups are used in the adjustment procedure. See Appendix II, Age adjustment. 5 Includes persons of all races and Hispanic origins, not just those shown separately. 6 Percent of poverty level is based on family income and family size. Persons with unknown percent of poverty level are excluded (5% in 2003–2006). See Appendix II, Family income; Poverty. Notes: Percents are based on the average of blood pressure measurements taken. In 2003–2006, 81% of participants had three blood pressure readings. See Health, United States, 2003, Table 66 for a longer trend based on a single blood pressure measurement, which provides comparable data across five time periods (1960–1962 through 1999–2000). Excludes pregnant women. Estimates for persons 20 years and over are used for setting and tracking Healthy People 2010 objectives. Standard errors are available in the spreadsheet version of this table. Available from: http://www.cdc.gov/nchs/hus.htm. Data have been revised and differ from previous editions of Health, United States. Data for additional years are available. See Appendix III.
Female
29.2 40.6 54.4 60.4
2/25/10
20–34 years 35–44 years 45–54 years 55–64 years 65–74 years 75 years and over
45–54 years 55–64 years 65–74 years 75 years and over
46894_CH12_311_360.qxd Page 325
DISTRIBUTION 325
46894_CH12_311_360.qxd
326
2/25/10
3:40 PM
Page 326
CHAPTER 12 HIGH BLOOD PRESSURE
Persistent elevated blood pressure was increasingly common for both older males and females. Data regarding awareness, treatment, and control of hypertension from the NHANES examinations of 1999–2002 indicate that prevalence was 28.6% overall (Table 12-7).48 Of those with hypertension, only two-thirds were aware of having the condition, fewer than one-half were currently treated, and fewer than one-third had it controlled. Proportions of persons aware, under treatment, and controlled were lower for men than women at each stage. Non-Hispanic Blacks had the highest prevalence, awareness, and treatment, but no better frequency of control than Whites; Mexican Americans had lower prevalence, but among hypertensives much lower awareness, treatment, and control than other groups. Overall, regardless of sex, race/ethnicity, or age, the great majority of US adults with high blood pressure do not have it controlled. Prehypertension (defined in Table 12-1) was estimated from NHANES 2005–2006 data to affect approximately 25% of the population at ages 20 and older, corresponding to some 32 million men and 21 million women. Estimates have varied among sources depending on exclusion or inclusion of persons who have high blood pressure treated and controlled, whose blood pressures fall in the prehypertension
range.1,49 In keeping with JNC VII criteria, such persons have hypertension, not prehypertension. Special attention has been given to high blood pressure in older adults, for whom prevalence is highest and systolic hypertension (with diastolic pressure below 90 mm Hg) becomes common. Lloyd-Jones and colleagues focused on persons aged 80 years and older and found stage 1 or 2 hypertension in 69.1% of men and 76.5% of women at this age, with only 38% of men and 23% of women having it controlled.50 The Coordinating Committee of the National High Blood Pressure Education Program issued a special Clinical Advisory Statement on this topic calling for greater attention to systolic hypertension in the elderly population.51 The statement sought to dispel the once-popular view that the standard of normal blood pressure should be “your age 100,” in favor of adherence to Joint National Committee guidelines that apply at any adult age. Optimistically, the statement concluded that “The vast majority of hypertensive individuals can achieve recommended BP targets without significant difficulty.”51, p 1024 Blood pressure increases throughout childhood and adolescence—except for a decrease in girls in the late teens—from values low at birth to those found by
Table 12–7
Percentage of Noninstitutionalized US Adults with Hypertension* and Among Those with Hypertension, Estimated Percentage of Persons Who Are Aware of,† Treated for,§ and in Control of Their Condition, by Sex, Race/Ethnicity, and Age Group—United States, 1999–2002 Hypertension Awareness of Under Current Condition Prevalence Condition Treatment Controlled Characteristic** % (95% C1††) % (95% C1) % (95% C1) % (95% C1) Sex Men 27.8 (24.9–29.7) 59.4 (55.8–63.1) 45.2 (40.9–49.6) 27.5 (23.7–31.3) Women 29.0 (27.3–30.8) 69.3 (61.7–77.0) 56.1 (29.2–63.1) 35.5 (28.4–42.7) Race/Ethnicity White, non-Hispanic Black, non-Hispanic Mexican American
27.4 40.5 25.1
(25.3–29.5) (38.2–42.8) (23.1–27.1)
62.9 70.3 49.8
(57.3–68.5) (64.9–75.8) (40.4–59.2)
48.6 55.4 34.9
(44.1–53.1) (51.2–59.6) (27.5–42.3)
29.8 29.8 17 3
(25.7–34.0) (25.2–34.5) (10.7–23.8)§§
Age group (yrs) 20–39 40–59 60
6.7 29.1 65.2
(5.3–8.2) (25.9–32.4) (62.4–68.0)
48.7 73.5 72.4
(38.8–58.7) (69.1–77.9) (70.0–74.7)
28.1 61.2 65.6
(20.1–36.1) (57.1–65.2) (61.9–69.3)
17.6 40.5 31.4
(11.6–23.7) (36.4–44.5) (28.7–34.2)
Total¶¶
28.6
(26.8–30.4)
63.4 (59.4–67.4)
29.3
(26.0–32.7)
45.3 (45.3–52.8)
*Had a blood pressure measurement 140 mm Hg systolic or 90 mm Hg diastolic or took antihypertensive medication. † Told by a healthcare professional that blood pressure was high. § Took antihypertensive medication. ¶ Hypertension levels 140 mm Hg systolic and 90 mm Hg diastolic. **All characteristic estimates (excluding age group) are age adjusted. †† Confidence interval. §§ Estimate should be used with caution; relative standard error is 20%–29%. ¶¶ Total population estimates (including sex and age group) include only non-Hispanic Whites, non-Hispanic Blacks, and Mexican Americans. Source: Reprinted from MMWR, Vol 54, January 14, 2005, p 7.
46894_CH12_311_360.qxd
2/25/10
3:40 PM
Page 327
DISTRIBUTION 327
early adulthood. Criteria for levels of concern before adulthood, according to JNC VII, are indicated in Table 12-2. Reference percentiles are based on aggregated surveys to provide large numbers of observations for each year of age from birth to 19 years, separately by sex. These values are adjusted for height, which is strongly associated with growthrelated increase in blood pressure especially in adolescence. Data from NHANES pooled for the years 1999–2006 showed prevalence of hypertension defined as exceeding the 95th percentile—although based on the single examination and not multiple ones—to be 3.7% for boys and girls together at ages 8–12 years and 20.1% for those aged 13–17 years (unpublished data). A large survey in Quebec, Canada, found “high-normal or elevated SBP” to be present at ages 9, 13, and 16 years in 12, 22, and 30% of boys and 14, 19, and 17% of girls, respectively.52 The frequency of elevated blood pressure at these ages is often unrecognized, as shown by a study in northeast Ohio, in the United States, which found the condition coded in the electronic medical record of only 26% of those in whom it was present in the examination data.53 This would correspond to a much lower proportion aware of the condition than in most surveys of adults. Trends in population-wide blood pressure levels over several years or longer may indicate change in population determinants of blood pressure. Measurement comparability is a requirement, and increasingly since the 1970s, in the United States and in many other countries, the possible influence of treatment must also be considered. Goff and others analyzed data from national surveys in the United States from 1960–1962 to 1988–1994 for successive birth cohorts of survey participants (Figure 12-4a and b).54 (This analysis parallels the study of cholesterol trends addressed earlier.) The 90th, 50th, and 10th percentile levels of the blood pressure distribution for each birth cohort and age were determined and presented both as raw values and as estimated from a modeling procedure. Neither the 50th nor 10th percentile levels would be influenced by treatment, which at least for later cohorts would become common at the 90th percentile. Figure 12-4a shows the 50th percentile levels of systolic blood pressure for persons examined at age 70 years (left panel, far right) who would have been born between 1887–1899 if examined in 1960–1962, between 1900–1909 if examined in 1970–1974, and so on. Blood pressure at age 70 years declined from the earliest to the latest birth cohort examined at that age, and similar declines are generally shown for other ages as well. This means that at each attained age from 20 to 70 years, men examined more recently had lower systolic
pressures than their predecessors. The right panel represents the data modeled as though each birth cohort were observed at every age throughout adulthood. Figure 12-4b presents the corresponding information for diastolic blood pressure. Together, these trends indicate that, free of any influence of treatment of high blood pressure, the distribution of both systolic and diastolic pressure was shifting downward at all ages, from the early 1960s to the mid-1990s. There is some question regarding comparability of blood pressure measurements over the full span of these surveys. When NHANES data are compared from 1971–1975 to 2005–2006, mean values of both systolic and diastolic pressure decreased sharply for nearly every age, sex, and race/ethnic group between 1976–1980 (NHANES II) and 1988–1994 (NHANES III), with generally much less variation thereafter (unpublished data). However—partly on the basis of medication status and not blood pressure levels alone—prevalence of high blood pressure increased from 1988–1994 to 1999–2002, as reported in the Healthy People 2010 Mid-Course Review.55 This finding stood out as showing change opposite to the direction of the target, which was to reduce prevalence at age 20 years and above from 26 to 14%; instead, it increased by one-third of the intended distance of 12 percentage points, to 30%. It may be that increased survival of persons reporting medication use was the greater contributor to this change. Available data do not permit distinguishing between these explanations. Trends in awareness, treatment, and control of high blood pressure have been reported on the basis of NHANES data by several investigators in recent years, with one or more new reports forthcoming as new NHANES data become available. Three reports address trends in prevalence, awareness, treatment, and control of hypertension through 2004.56–58 There is variation among reports as to the proportion of hypertension controlled, ranging from 35.1 to 43%. Similarly, trends in blood pressure levels or prevalence of hypertension in children and adolescents are addressed in multiple reports. Muntner and others reported a marked increase in mean values of systolic and diastolic blood pressure in 8–12- and 13–17-yearolds from 1988–1994 to 1999–2000 (Table 12-8a and b).59 Others extended analysis to 2003–2006 and concluded that prevalence of elevated blood pressure increased for girls but decreased for boys in this period relative to 1988–1994.60 Systolic pressure was higher and diastolic pressure the same in 2003–2006 as in 1988–1994, but both were lower (diastolic pressure much lower) than in 1971–1975. Comparability among these multiple and sometimes conflicting
90
10
10
10
20
1887–99 1900–09 1910–19 1920–29 1930–39 1940–49 1950–59 1960–69
10th
20
1887–99 1900–09 1910–19 1920–29 1930–39 1940–49 1950–59 1960–69
50th
20
1887–99 1900–09 1910–19 1920–29 1930–39 1940–49 1950–59 1960–69
90th
30
30
30
40 50 60 Attained Age
40 50 60 Attained Age
40 50 60 Attained Age
70
70
70
80
80
80
90
100
110
120
130
140
150
160
110
120
130
140
150
160
170
180
130
140
150
160
170
180
190
200
10
10
10
20
10th
20
50th
20
90th
30
30
30
40 50 60 Attained Age
40 50 60 Attained Age
40 50 60 Attained Age
70
70
70
1890 1900 1910 1920 1930 1940 1950 1960
80
80
1890 1900 1910 1920 1930 1940 1950 1960
80
1960
1950
1940
1930
1920
1910
1900
1890
B
50
60
70
80
90
60
70
80
90
100
70
80
90
100
110
10
10
10
20
1887–99 1900–09 1910–19 1920–29 1930–39 1940–49 1950–59 1960–69
10th
20
1887–99 1900–09 1910–19 1920–29 1930–39 1940–49 1950–59 1960–69
50th
20
1887–99 1900–09 1910–19 1920–29 1930–39 1940–49 1950–59 1960–69
90th
30
30
30
40 50 60 Attained Age
40 50 60 Attained Age
40 50 60 Attained Age
70
70
70
80
80
80
50
60
70
80
90
60
70
80
90
100
70
80
90
100
110
10
10
10
20
10th
20
50th
20
90th
30
30
30
40 50 60 Attained Age
40 50 60 Attained Age
40 50 60 Attained Age
70
70
70
Figure 12-4 Observed (Left) and Estimated (Right) 50th Percentile Curves for Systolic Blood Pressure (Panel A) and Diastolic Blood Pressure (Panel B) Over the Age Range of 18 Through 74 Years by Birth Cohort. In the panels depicting estimated patterns, the solid lines reflect the ranges for which data were observed and the dotted lines reflect the ranges for which the values were extrapolated from the observed data. The models used to derive these estimates included the following independent variables: age, age2, birth-year, age*birth-year, and age2*birth-year. Source: Adapted and printed with permission from Annals of Epidemiology, Vol 11, DCC Goff Jr, G Howard, GB Russell, DR Labarthe, pp 275–276. © Elsevier Science.
A
100
110
120
130
140
150
160
110
120
130
140
150
160
170
180
130
140
150
160
170
180
190
200
DBP (mm/Hg) DBP (mm/Hg)
SBP (mm/Hg)
SBP (mm/Hg)
SBP (mm/Hg)
SBP (mm/Hg) SBP (mm/Hg) SBP (mm/Hg)
80
1890 1900 1910 1920 1930 1940 1950 1960
80
1960
1950
1930 1940
1920
1900 1910
1890
80
1960
1950
1940
1930
1920
1910
1900
1890
3:40 PM
DBP (mm/Hg)
DBP (mm/Hg) DBP (mm/Hg)
328
2/25/10
DBP (mm/Hg)
46894_CH12_311_360.qxd Page 328
CHAPTER 12 HIGH BLOOD PRESSURE
46894_CH12_311_360.qxd
2/25/10
3:40 PM
Page 329
DISTRIBUTION 329
Table 12-8a
Total Boys Girls Non-Hispanic White Black Mexican American Age group, y 8–12 13–17
Systolic Blood Pressure by Age, Race/Ethnicity, and Sex for Children and Adolescents Aged 8 Through 17 Years* Mean (SE) Systolic Blood Pressure, mm Hg 1988–1994† 1999–2000‡ Difference Between Years p Value 104.6 (0.36) 106.0 (0.26) 1.4 (0.43) 0.001 106.2 (0.48) 107.6 (0.37) 1.4 (0.55) 0.03 102.9 (0.42) 104.4 (0.47) 1.5 (0.62) 0.001 104.3 (0.50) 105.6 (0.43) 104.8 (0.39)
105.3 (0.36) 107.5 (0.52) 107.1 (0.43)
1.0 (0.58) 1.9 (0.59) 2.3 (0.60)
0.06 0.001 0.001
100.6 (0.46) 108.4 (0.42)
102.5 (0.54) 109.4 (0.34)
1.9 (0.68) 1.0 (0.52)
0.001 0.09
*Standardized to the age (by year), race/ethnicity, and sex distribution of children and adolescents in the third National Health and Nutrition Examination Survey. † From the third National Health and Nutrition Examination Survey. ‡ From the National Health and Nutrition Examination Survey 1999–2000. Source: Reprinted with permission from Journal of the American Medical Association, Vol 291, No 17, P Munter, J He, JA Cutler, RP Wildman, PK Whelton, p 2111. © 2004 American Medical Association.
reports is compromised by differences in measurement protocols, data selection, and definitions and criteria. Global The global dimensions of high blood pressure were reviewed by Hajjar and others in 2006, who concluded that an estimated 972 million persons worldwide have hypertension.61 Prevalence estimates from 20 to 48% of adults were reported from published sur-
Table 12-8b
Total Boys Girls Non-Hispanic White Black Mexican American Age Group, y 8-12 13-17
veys in North America, Europe, South and East Asia, Australia, and the Middle East. Again, comparisons across surveys were limited by differences in age span, sampling design, and blood pressure measurement. Regardless of these limitations, the many studies indicated broad concern about blood pressure as a public health problem throughout the world. The magnitude of between-population differences in the prevalence of high blood pressure, as measured under a well-standardized protocol and study
Diastolic Blood Pressure by Age, Race/Ethnicity, and Sex for Children and Adolescents Aged 8 Through 17 Years* Mean (SE) Diastolic Blood Pressure, mm Hg 1988–1994† 1999–2000‡ Difference Between Years p Value 58.4 (0.40) 61.7 (0.46) 3.3 (0.61) 0.001 58.9 (0.50) 61.2 (0.59) 2.3 (0.77) 0.001 57.9 (0.45) 62.1 (0.62) 3.2 (0.77) 0.001 58.8 (0.51) 57.6 (0.50) 57.6 (0.69)
61.6 (0.62) 61.7 (0.80) 62.0 (0.40)
2.8 (0.80) 4.1 (0.94) 4.4 (0.80)
0.001 0.001 0.001
54.9 (0.54) 61.8 (0.50)
59.7 (0.76) 63.5 (0.49)
4.8 (0.93) 1.7 (0.70)
0.001 0.02
*Standardized to the age (by year), race/ethnicity, and sex distribution of children and adolescents in the third National Health and Nutrition Examination Survey. † From the third National Health and Nutrition Examination Survey. ‡ From the National Health and Nutrition Examination Survey 1999–2000. Source: Reprinted with permission from Journal of the American Medical Association, Vol 291, No 17, P Munter, J He, JA Cutler, RP Wildman, PK Whelton, p 2111. © 2004 American Medical Association.
46894_CH12_311_360.qxd
330
2/25/10
3:40 PM
Page 330
CHAPTER 12 HIGH BLOOD PRESSURE
design, is apparent in data from the World Health Organization MONICA Project, in which 40 reporting population units are represented (see Appendix 4-A for population locations).62 Figures 12-5A and B shows for 35- to 64-year-old men and women, re-
spectively, the proportions of the total population among four categories from the left to the right segments of the graph: those who were treated and controlled (systolic pressure less than 160 mm Hg, diastolic pressure less than 95 mm Hg, and on treat-
Men (35–64 Years)
Population Catalonia Ghent Glostrup Luxembourg Province Vaud/Fribourg Ticino Charleroi Auckland Northern Sweden Augsburg (rural) Rhein-Neckar Region Novi Sad Stanford Belfast Beijing Perth Newcastle Haute-Garonne Bremen Moscow (intervention) Augsburg (urban) Tarnobrzeg Voivodship Novosibirsk (control) Kaunas Brianza Area Glasgow Czechoslovakia Malta Halle County Friuli Turku/Loimaa Novosibirsk (intervention) Moscow (control) Berlin-Lichtenberg Warsaw DDR MONICA (other surveys) Karl-Marx-Stadt County North Karelia Bas-Rhin Kuopio Province
0 Categories
25
50
75
I
II
III
IV
No. observations
2.1 5.3 3.7 5.2 4.4 3.6 6.5 4.4 5.2 2.7 9.3 5.8 7.7 2.8 1.7 5.3 7.9 3.8 3.0 1.5 3.0 4.6 1.7 4.2 2.0 4.5 6.5 4.0 3.3 3.1 4.2 2.0 2.2 6.4 1.7 6.0 3.8 4.7 2.3 2.9
1.2 3.1 2.8 3.3 3.1 3.1 2.4 3.7 3.8 3.2 4.5 6.6 4.0 3.4 5.4 4.5 5.3 5.6 2.7 3.7 4.1 5.3 4.3 7.8 6.4 3.7 7.0 7.3 6.5 6.3 7.2 5.4 6.9 12.9 7.3 9.2 6.9 8.7 7.8 8.8
5.1 4.5 8.5 7.0 10.6 12.2 10.5 12.1 11.4 15.7 8.2 10.3 11.8 17.6 17.5 15.4 12.5 16.3 20.3 21.2 20.6 17.9 23.2 18.4 22.7 23.8 18.9 21.8 23.8 24.5 24.1 28.3 27.5 17.9 28.4 22.7 28.2 26.1 32.2 33.6
91.6 87.0 85.1 84.5 82.0 81.1 80.5 79.9 79.6 78.4 78.0 77.4 76.6 76.3 75.4 74.9 74.3 74.3 74.0 73.6 72.3 72.2 70.7 69.6 68.9 68.0 67.6 66.9 66.4 66.1 64.5 64.3 63.4 62.9 62.5 62.2 61.2 60.6 57.8 54.7
396 390 1380 971 602 745 275 1012 635 846 559 536 431 927 618 631 1217 622 633 1093 712 1191 1060 728 613 492 942 656 982 708 1162 601 775 526 1309 529 796 1115 660 948
100
Proportion (%) I. Systolic blood pressure <160 and diastolic blood pressure <95; on treatment for hypertension. II. Systolic blood pressure >159 and/or diastolic blood pressure >94; on treatment for hypertension. III. Systolic blood pressure >159 and/or diastolic blood pressure >94; not on treatment for hypertension. IV. Systolic blood pressure <160 and diastolic blood pressure <95; not on treatment for hypertension.
Figure 12-5A Age-Standardized Proportions of Categories of Blood Pressure Index, Men, WHO MONICA Project. Source: Reprinted with permission from The WHO MONICA Project, World Health Statistics Quarterly, Vol 41, p 128, © 1998, World Health Organization.
46894_CH12_311_360.qxd
2/25/10
3:40 PM
Page 331
DISTRIBUTION 331
Women (35–64 Years)
Population Glostrup Catalonia Vaud/Fribourg Charleroi Ghent Stanford Ticino Rhein-Neckar Region Haute-Garonne Auckland Northern Sweden Perth Augsburg (rural) Augsburg (urban) Luxembourg Province Belfast Beijing Bremen Turku/Loimaa Brianza Area Newcastle Glasgow Novi Sad Moscow (intervention) Friuli Kaunas Warsaw Czechoslovakia Novosibirsk (control) Bas-Rhin Berlin-Lichtenberg Tarnobrzeg Voivodship North Karelia Karl-Marx-Stadt County Halle County Malta Moscow (control) Kuopio Province Novosibirsk (intervention) DDR MONICA (other surveys) 0
25
50
75
I
II
III
IV
No. observations
6.0 9.4 4.8 10.3 12.0 8.1 5.4 8.2 4.6 9.5 6.1 7.6 5.1 4.1 8.2 4.7 5.2 3.3 4.9 4.4 11.6 3.9 10.3 6.5 4.6 7.3 3.2 9.1 4.7 6.0 9.8 7.7 5.0 7.1 6.0 6.9 6.2 6.5 5.8 8.0
1.9 1.2 3.0 1.8 2.6 2.4 4.0 5.6 5.4 2.3 5.1 5.0 4.5 4.6 5.0 4.0 4.4 5.1 5.4 8.7 6.0 2.1 10.9 8.6 8.2 13.1 9.4 10.2 8.3 8.3 12.0 9.1 10.2 9.2 9.1 13.3 10.8 9.0 13.3 14.3
4.7 2.4 6.2 3.6 1.9 6.2 7.6 3.5 7.6 6.4 7.6 6.5 9.8 10.8 6.3 11.6 11.9 14.9 14.0 11.5 7.4 19.4 6.5 12.9 16.4 10.3 18.3 12.0 18.5 17.8 10.5 17.6 19.5 18.4 20.4 16.0 20.7 22.1 21.2 18.2
87.4 86.9 86.0 84.4 83.4 83.2 83.0 82.6 82.4 81.9 81.3 80.9 80.7 80.6 80.4 79.7 78.5 76.7 75.8 75.4 74.9 74.6 72.3 72.1 70.7 69.4 69.1 68.7 68.5 67.9 67.6 65.6 65.3 65.3 64.5 63.8 62.4 62.3 59.7 59.5
1339 389 551 247 311 518 759 608 626 566 608 661 857 679 937 923 641 650 1243 630 1245 475 519 1130 720 734 1337 988 1054 713 565 1428 1196 889 1055 687 649 964 646 592
100
Proportion (%) Categories I. Systolic blood pressure <160 and diastolic blood pressure <95; on treatment for hypertension. II. Systolic blood pressure >159 and/or diastolic blood pressure >94; on treatment for hypertension. III. Systolic blood pressure >159 and/or diastolic blood pressure >94; not on treatment for hypertension. IV. Systolic blood pressure <160 and diastolic blood pressure <95; not on treatment for hypertension.
Figure 12-5B Age-Standardized Proportions of Categories of Blood Pressure Index, Women, WHO MONICA Project. Source: Reprinted with permission from The WHO MONICA Project, World Health Statistics Quarterly, Vol 41, p 128, © 1998, World Health Organization.
ment for hypertension, I); treated and not controlled (II); hypertensive but not treated (III); and not hypertensive (IV). The criterion for hypertension and control was set at systolic pressure of 160 mm Hg and
diastolic pressure of 95 mm Hg. The proportions of the population in each class and the number of observations are indicated for each population. The populations are ranked in descending order of freedom
46894_CH12_311_360.qxd
332
2/25/10
3:40 PM
Page 332
CHAPTER 12 HIGH BLOOD PRESSURE
from hypertension (category IV)—for men, greatest in Catalonia (91.6%) and least in Kuopio Province (54.7%) and for women, greatest in Glostrup (87.4% and least in DDR (59.5%). Total prevalence of hypertension is the sum of categories I–III and the proportion of hypertension controlled is the ratio of I to I–III—for example, for men in Catalonia, 2.1/(2.1 1.2 5.1) 25%. The corresponding value for the highest-prevalence population, Kuopio, is 6.4%. Even within the scope of the primarily European populations in the MONICA Project, high blood pressure control was achieved for only a small proportion of affected persons as of the mid-1980s. The limited achievement of blood pressure control in various countries is further illustrated in Figure 12-6.63 It appears, subject to the qualifications regarding comparability noted above, that among the countries represented here, the great majority of per-
sons known to have high blood pressure do not have it controlled, whether the criterion of control is at the levels of systolic/diastolic blood pressure below 140/90 or 160/95 mm Hg. Other reports on populations in North America, Europe, Asia, and Africa support this view.64,65 Canadian experience illustrates perhaps an exceptional level of success in controlling high blood pressure, with a report from the 2006 Ontario Survey on the Prevalence and Control of Hypertension indicating that 65.7% of persons identified with high blood pressure were on treatment and controlled.66 Whether self-selection among survey participants (about one-third of eligible individuals did not participate) biased this estimate in favor of better control is uncertain, although this was not considered likely by the authors of the report. Perhaps most remarkable is the great improvement from a 1992 survey: total
⬍140/90 mmHg 11 %
25 %
Belgium[8]
2%
Bulgaria[8]
Cameroon[9] 3%
9.3 %
16 %
Canada[5]
China[6]
England[4] 27 %
33 %
France[7]
34 %
Italy[8]
USA[3]
⬍160/95 mmHg 20 %
19 %
Australia[2]
22 %
Finland[2]
Germany[2]
9%
India[2]
20 %
Spain[2]
Figure 12-6 Percentage of Patients with Controlled Blood Pressure in Different Countries ( 140/90 mm Hg or 160/95 mm Hg). Source: Reprinted with permission from Clinical and Experimental Hypertension, Vol 26, Nos 7&8, S Erdine, SN Aran, p 732. © 2004 by Marcel Dekker.
46894_CH12_311_360.qxd
2/25/10
3:40 PM
Page 333
DISTRIBUTION 333
proportion treated, from 32.9 to 80.4%; treated and not controlled, from 20.8 to 14.7%; and treated and controlled, from 12.1 to 65.7%. Also notable is the apparent lack of increase in prevalence of high blood pressure (defined by blood pressure levels of 140/90 mm Hg or greater or treatment) in this population, remaining at about 20% of adults on the basis of these surveys. On a global scale, prevalence of high blood pressure has been estimated for the year 2000 and projected to 2025 for world regions in accordance with the World Bank geographic definitions.67 Figure 12-7a and b presents the results first as the rate (prevalence) of hypertension for men and women at the two time points and second as numbers of persons affected, all based on the criteria of systolic/diastolic blood pressure of 140/90 mm Hg or greater or use of antihypertensive medication. Age ranges varied in the available data sources but were statistically adjusted to the 1990 world population for persons aged 20 years or older. The overall world prevalence in 2000 was estimated to be 26.4%. In general across the regions, prevalence varied similarly for men and women, being least in the region defined as other Asia and islands (about 16%) and greatest in established market economies (about 37%). Estimates for 2025 were based on projected population size and age structure for each region, without consideration of possible changes in incidence over the interval. Prevalence in 2025 would thus range from about 18 to 42% at the low and high extremes. In numbers of persons, reflecting the public health burden, the aggregate total worldwide was estimated to be 972 million in 2000, increasing to 1.56 billion in 2025. Trends in prevalence beyond those discussed above for the United States and including periodic assessment of awareness, treatment, and control have been studied most rigorously in the WHO MONICA Project.68,69 Across all MONICA populations, measured changes in mean systolic and diastolic blood pressure from the initial to final surveys, in the mid1980s and mid-1990s, respectively, were evaluated through independent probability sample surveys. Changes were predominantly downward, although with some striking exceptions. Overall decreases in mean systolic and diastolic (phase 5) blood pressure, respectively, were 2.2 and 1.4 mm Hg for men and 3.3 and 2.2 mm Hg for women were shown (Table 12-9). Wide variation among study populations was evident, for example, in changes in mean values of systolic pressure: from 10.6 mm Hg in Warsaw to 4.5 mm Hg in Halifax; the same populations were at the extremes for women: 13.5 mm Hg in Warsaw and 8.1 mm Hg in Halifax.
A focus of this assessment was on the question of whether decreases in population means were attributable to treatment effects alone, at the upper extreme of the blood pressure distribution, or to populationwide influences that would affect the whole distribution. Accordingly, changes for the pooled populations and for each population in values of the 20th, 50th, and 80th centiles and the average of these specific centile changes were evaluated. For the pooled populations, decreases were generally as great or greater at the 20th and 50th centiles as at the 80th. For the extreme populations, decreases for Warsaw and increases for Halifax in centile values of both systolic and diastolic pressure were generally consistent across the distribution. These observations indicate, as did the analysis by Goff and others for the United States,54 that changes in whole distributions occurred that were not attributable to treatment effects but must reflect population-wide influences, either favorable or, in some cases, unfavorable. The authors note that these findings may be specific to the decade of the MONICA Project. Further, they caution that improvements in blood pressure control coincident with increased treatment, observed in other studies, might reflect similar but unexamined changes in the distribution overall and not be due to treatment. Trends in awareness, treatment, and control were similarly found to vary across the MONICA populations and improved in many but not all instances.69 Both blood pressure levels of 160/95 mm Hg and 140/90 mm Hg were used in evaluating these trends. For example, data for women in relation to the criterion of 140/90 mm Hg for prevalence and control from the final survey in the mid-1990s showed control to range from 12% of those treated in Gothenburg to 63% in Ghent. Generally higher proportions of treated persons were controlled by the 160/95 criterion, but the range was still wide—from 33 to 90% for men and 45 to 92% for women. Once again the evidence indicated strongly that on average, with few exceptions among studied populations, the majority of persons with high blood pressure do not have it controlled. Incidence The incidence of high blood pressure, or progression from blood pressure levels below a specified criterion value to those above, has been investigated less often than prevalence because of the necessity for years-long observation and the difficulty of classifying individuals reliably at any given occasion due to the moment-to-moment variability of each person’s blood pressure. Langford characterized this problem
M
ar
39.1
35.3
45.9
39.1
a
44.5 40.2
34.8
24.0
27.0
22.0 23.7
n a ic er t er ean ast en Am ibb e E esc tin ar ddl Cr a L e C Mi th nd
di
a
In
22.9 23.6
20.6 20.9
40.7
19.7
na
14.5
18.8 17.1
17.0
27.0 28.2
n d ra a an s a nd aha fric C i S A As Isla ber h t Su O hi
27.7 27.0
22.6
26.9 28.3
Men Women
B
a M
161.8 147.9
2025
116.2123.3
2000
44.0
40.6
59.7
52.5
t d lis he il s ies ocia ies b ta nom r S nom e o Es co m c r E For E e rk
180 160 140 120 100 80 60 40 20 0
180 160 140 120 100 80 60 40 20 0 60.0 54.3
ia
72.2
80.4
35.9 37.9
a n ic er er ean ast ent m c A ibb e E es tin ar ddl Cr a L e C Mi th d an d In
107.3 106.2 102.1 98.5
60.4 57.8
83.1
O
67.3 62.1
38.4 33.0
73.6 77.1
38.2 41.6
Men Women
d n ra a an s a nd aha fric i S A As Isla ber th Su na hi C
151.7147.5
98.5
Figure 12-7 Frequency of Hypertension (Panel A) and Numbers of People with Hypertension (Panel B) in People Aged 20 Years and Older by World Region and Sex in 2000 (Upper) and 2025 (Lower). Source: Reprinted with permission from The Lancet, Vol 365, PM Kearney, M Whelton, J He, pp 219 and 221.
A
41.6 42.5
2025
37.4 37.2
2000
t d lis he is ies ocia ies l b ta nom r S nom e o Es co rm Ec E o r F ke
0
10
20
30
40
50
0
10
20
30
40
50
3:40 PM
Rate of Hypertension (%)
334
2/25/10
Number of People with Hypertension (millions)
46894_CH12_311_360.qxd Page 334
CHAPTER 12 HIGH BLOOD PRESSURE
46894_CH12_311_360.qxd
2/25/10
3:40 PM
Page 335
DISTRIBUTION 335
Table 12-9 Population Men Canada-Halifax Poland-Warsaw Pooled mean Women Canada-Halifax Poland-Warsaw Pooled mean
Differences in Systolic Blood Pressure* in mmHg in Men and Women Between the Initial and Final MONICA Surveys: Age 35–64 (Age Standardized) Mean SD 20th Centile 50th Centile 80th Centile Centile Average 4.5 10.6 2.2
2.3 0.3 0.1
2.0 11.0 2.1
3.0 10.0 2.4
7.0 11.0 2.1
4.0 10.7 2.2
8.1 13.5 3.3
0.8 2.4 0.3
8.0 12.0 2.9
6.0 15.0 3.3
10.0 15.0 3.3
8.0 14.0 3.2
*Values listed are differences, derived by subtracting the initial survey statistic from that from the final survey. This explains, for example, why SD is not the standard deviation of the mean and has negative, and zero, as well as positive values. Source: Data from BMJ Online First, H Tunstall-Pedoe, J Connaghan, M Woodward, H Tolonen, K Kuulasmaa.
as like that of “counting fireflies” (personal communication, Herbert Langford, 1973). An early opportunity presented itself in a study of “sustained” and “transient” hypertension and health outcomes in 22,741 US military officers over an average of 9.8 years of follow-up.70 One report in a series that appeared between 1944 and 1947 described predictors of sustained hypertension, defined as persistent readings of greater than 150 mm Hg systolic or 90 mm Hg diastolic pressure within an examination and without lower values on subsequent examinations. Sustained hypertension was predicted by transient elevations of either blood pressure or heart rate (over, then below 100 beats per min) or by overweight (20 pounds or more above the Army standard by age and height for each man). The frequency of sustained hypertension during follow-up was shown to increase 12-fold in the presence of these three factors. Another example is the study of more than 7000 participants age 25–74 years at first observation in the National Health and Nutrition Examination Survey (NHANES I) Epidemiologic Follow-Up Study (1971–1984).71–73 The criterion for the new appearance of high blood pressure was an increase from below 160 mm Hg systolic and 95 mm Hg diastolic to above one or both of these values or from a negative to a positive history of using blood pressure-lowering medications. The factors most strongly predictive of this change were the initial BMI in Black and White men and women, and low educational attainment, especially in White women.71 Further analysis refined this observation to demonstrate an interaction of age on the effect of educational attainment on incidence of high blood pressure.72 Among White men and women aged 25–44 years, incidence differed between those with 12 years of education or less (12–14/1000 person-years) and those with more than 12 years
(6.7/1000 person-years). Incidence was greater (21–23/1000 person-years) among Whites aged 45–64 years but less strongly related to education. Among Black men and women aged 25–44 years, a similar gradient was apparent but not statistically significant (incidence 20–24/1000 person-years at 12 years or less versus 16.8/1000 person-years with more than 12 years of education). The difference in incidence for young Blacks in contrast to Whites is striking. These data were also explored to address the hypothesis that greater incidence would be found in the southeastern United States, but the evidence from this study was not judged by the authors to be convincing.73 Incidence of hypertension was studied in Ontario, Canada, in another population-based study, not by remeasurement after a follow-up interval but by linkage of records for the total population of Ontario in a universal database covering all physician and hospital services.74 From 1995 through 2005, data were monitored for identifying entries for hypertension for each year among men and women aged 20 years or older. By this method, prevalence was found to increase substantially in every age-sex group studied— overall, from 153.1 to 244.8/1000 population from 1995 to 2005. This was interpreted as a 60% relative increase in prevalence. The number of new diagnoses of hypertension in a given year was divided by the size of the population without a hypertension-related entry in the previous year. By this calculation, incidence of hypertension was determined to be 25.5/ 1000 in 1997 and 32.1/1000 in 2004, a relative increase of 25.7%. The authors project that prevalence will increase far beyond the estimate for 2025 and that prevention and management of hypertension and its consequences are an urgent public health priority. It is notable that the Ontario survey cited previously66 found the prevalence of hypertension by direct
46894_CH12_311_360.qxd
336
2/25/10
3:40 PM
Page 336
CHAPTER 12 HIGH BLOOD PRESSURE
measurement in a sample survey to be approximately 20% and stable over recent decades, whereas the present study found cases through clinical record linkage to increase from about 15 to 24% over the decade from 1995–2005. It is interesting to speculate that the health system of Ontario was catching up with the constant population burden, an interpretation supported by the demonstrated improvement in treatment and control from the survey. If so, perhaps the forecast of mounting prevalence is erroneous, as the upper limit of true population prevalence has become nearly reached with the recent increase of clinical services for this population. Disparities Disparities by race/ethnicity regarding determinants, prevalence, awareness, treatment, and control in high blood pressure have been discussed briefly earlier. Especially prominent in the United States is the excess prevalence of high blood pressure among African Americans, followed by the especially poor levels of treatment and control among Mexican Americans. For many groups of concern, insufficient data are available through current surveillance systems to determine the nature and extent of such disparities. The implications of these disparities, recognized and unrecognized, are important for successful and equitable prevention, detection, and control of high blood pressure in the United States. In addition, geographic disparities in cardiovascular diseases and conditions were addressed in general terms in Chapter 2, “Distributions and Disparities.” Work of Murray and colleagues on the Eight Americas project, introduced there, was extended to examine geographic aspects of hypertension specifically.75 Interest in state-level variation in blood pressure and its control stimulated an innovative approach to analysis to utilize both the national-level measurement data from NHANES and the state-level self-report data from the Behavioral Risk Factor Surveillance System (BRFSS) to generate estimates at the state level. As a result, the investigators concluded that “Lifestyle factors and pharmacological interventions for lowering blood pressure are particularly needed in the South and Appalachia, and with emphasis on control among women.”75, p 905
RELATION TO RATES AND RISKS The major contribution of blood pressure to overall population burden and mortality from ischemic heart disease and stroke is illustrated in Table 1-8, as estimated by the Global Burden of Disease and Risk
Factors Study for high-income and low- and middleincome regions and the world.76 The worldwide population-attributable fraction was 45% for ischemic heart disease and 54% for stroke. For stroke, this far exceeded the contribution of any other factor, whereas for ischemic heart disease, cholesterol was slightly more strongly related. Other sources demonstrate that blood pressure levels are strongly associated with both population differences in cardiovascular disease rates and variation in individual risks within a population. Population Differences The relation of population levels of blood pressure to corresponding death rates from coronary heart disease was shown most clearly by the Seven Countries Study, described in Chapter 4.77 Figure 12-8 illustrates the relation of the median values of systolic blood pressure to the respective 10-year coronary death rates across the 16 study cohorts. Population median values below 136 mm Hg clustered with generally low coronary mortality, whereas higher median values were generally characteristic of the populations with highest mortality. The regression coefficient of 2.1 shown in the figure represents slightly more than a doubling of the coronary death rate for every upward change of 10 mm Hg in the population median systolic pressure. Similar results were reported for median diastolic pressure. These findings underscore the influence of indicators of the blood pressure distribution, as a population phenomenon, in predicting differences in cardiovascular mortality among populations. Individual Differences Among the earliest investigations of individual risks of mortality related to blood pressure were those by the life insurance industry. In the United States, studies of policyholders were reported since 1925, with the Blood Pressure Study of the Society of Actuaries and Association of Life Insurance Medical Directors of America appearing in 1980.78 The report was based on some 4 million policies (disregarding the possibility that some individuals held multiple policies) and an average of 6.6 years of follow-up from the date of issuance. As in earlier reports, death was strongly related to blood pressure. For men, it increased from less than average (0.84 times the overall rate) at pressures below 128 mm Hg systolic and 83 mm Hg diastolic to 2.69 times the overall rate at pressures over 167 mm Hg systolic and 97 mm Hg diastolic. Overweight compounded the risk: overweight of only 15–25% fully offset the benefit of the lowest blood pressure category and increased relative risk in the highest cate-
46894_CH12_311_360.qxd
2/25/10
3:40 PM
Page 337
RELATION TO RATES AND RISKS 337
70 E
Y = 10-Year Deaths per 1,000 Men
60
r = 0.64
50
Y = –259 + 2.1X R
40
30
N B
W I
20
C
S
Z
G M
10 V
D
T U
0
K 124
126
128
130
132
134
136
138
140
142
144
146
X = Median Systolic Blood Pressure
Note: B, Belgrade; C, Crevalcore; D, Dalmatia; E, East Finland; G, Corfu; I, Italian Railroad; K, Crete; M, Montegiorgio; N, Zutphen; R, American Railroad; S, Slavonia; T, Tanushimaru; U, Ushibuka; V, Velika Krsna; W, West Finland; Z, Zvenjanin.
Figure 12-8 Ten-Year Mortality in Relation to Median Levels of Systolic Blood Pressure. Seven Countries Study. Source: Reprinted with permission of the publisher from Seven Countries by A Keys. Cambridge, Mass: Harvard University Press, Copyright © 1980 by the President and Fellows of Harvard College.
gory from 2.69 to 3.19 times the overall risk. For women, the same general pattern was reported: Risk in the lowest blood pressure category at standard weight was 0.90 times the overall risk; it was 2.16 times the overall risk for the highest category; and 15–25% overweight increased these relative risks to 0.95 and 2.94, respectively. As in previous insurance industry reports, the data indicated increased risks of death from the lowest to the second category of blood pressure, at levels approximately equivalent to the current criteria for prehypertension. At pressures from 128 to 137 mm Hg systolic and 78 to 87 mm Hg diastolic, relative risks were 1.11 for men and 1.08 for women. Accuracy of blood pressure values obtained under the circumstances of insurance examination is questionable, but the consistency of findings from these and other studies lends credence to their indication of overall patterns of risk. They have also been well enough validated through experience to provide a
basis for life insurance companies to adjust premiums to the blood pressure of an applicant and thereby assure financial benefit to the insurer.79 Risks specific for ischemic heart disease and stroke are also closely related to blood pressure levels, as shown in Figures 12-9a and b and 12-10a and b, respectively, from the Prospective Studies Collaboration.80 These figures summarize a meta-analysis of 61 prospective studies representing 12.7 million person-years of risk and more than 34,000 deaths from ischemic heart disease and 12,000 due to stroke. Studies that included repeated measurement of blood pressure permitted adjustment for individual variability in the baseline values of both systolic and diastolic pressure to estimate “usual” levels. “Usual” levels were considered to be free of regression-dilution bias that would lead to underestimation of risks. Results are presented separately for four age strata in each analysis. The risks for each age blood pressure category of experience, termed “floating risks,” are
46894_CH12_311_360.qxd
338
2/25/10
3:40 PM
Page 338
CHAPTER 12 HIGH BLOOD PRESSURE
Systolic Blood Pressure
Diastolic Blood Pressure Age at Risk:
256
80–89 Years
256
128
70–79 Years
128
64
60–69 Years
32
50–59 Years
16 40–49 Years 8 4
80–89 Years IHD Mortality (Floating Absolute Risk and 95% CI)
IHD Mortality (Floating Absolute Risk and 95% CI)
Age at Risk:
64
60–69 Years
32
50–59 Years
16
4 2
1
1
140 160 180 Usual Systolic Blood Pressure (mmHg) (a)
40–49 Years
8
2
120
70–79 Years
70
80 90 100 Usual Diastolic Blood Pressure (mmHg)
110
(b)
Figure 12-9 Ischemic Heart Disease (IHD) Mortality Rate in Each Decade of Age Versus Usual Blood Pressure at Start of That Decade. Source: Reprinted with permission from The Lancet, Vol 360, Prospective Studies Collaboration, p 1908.
calculated relative to the value 1 set for those with the lowest blood pressure at the youngest age represented in the data set (see each figure). For ischemic heart disease mortality, the absolute risks rose steeply, on a multiplicative scale, in relation to baseline levels of both systolic and diastolic pressure. They were higher at every level of blood pressure for successively older subjects. Only for diastolic pressure, in contrast with systolic pressure, was there a more gradual initial increase in risk from the first to the second level of pressure. There was no evidence of a threshold below which risk did not increase. For stroke mortality, the picture was strikingly similar, except that risks rose more steeply than for ischemic heart disease mortality. The age-specific associations were described as similar for women and men. Public Health Impact The public health impact of high blood pressure can be expressed in various terms—numbers of deaths, DALYs, economic costs, and others. In the United States, an analysis of contributions to mortality from 12 “modifiable dietary, lifestyle, and metabolic fac-
tors,” reported in 2009, included high blood pressure. This was defined as systolic pressure above the “theoretical-minimum-risk exposure distribution,” a value that is based on the lowest related mortality rate observed in epidemiologic studies, reported to be 115 mm Hg. The distribution of risks resulting from systolic blood pressure greater than 115 mm Hg was considered to account for 395,000 deaths annually. Reference to 115 mm Hg as a threshold between “optimum” and higher systolic pressure implies that efforts to reduce blood pressure to this level for persons at any higher level would confer benefit. That is, the attributable burden is not confined to those levels conventionally categorized as “hypertension” but extends to much lower levels. Only tobacco smoking contributed a greater number of deaths, estimated at 467,000. The authors concluded that “Smoking and blood pressure, which both have effective interventions, are responsible for the largest number of deaths in the US.”81, p 1 The global impact of blood pressure on coronary heart disease was estimated in terms of attributable deaths and DALYs for WHO-defined subregions
46894_CH12_311_360.qxd
2/25/10
3:40 PM
Page 339
RELATION TO OTHER FACTORS 339
Diastolic Blood Pressure
Systolic Blood Pressure
Age at Risk:
Age at Risk: 256
80–89 Years
128 64
60–69 Years 32
50–59 Years
16 8 4 2 1
80–89 Years
128
70–79 Years Stroke Mortality (Floating Absolute Risk and 95% CI)
Stroke Mortality (Floating Absolute Risk and 95% CI)
256
70–79 Years
64
60–69 Years
32
50–59 Years
16 8 4 2 1
120
140 160 180 Usual Systolic Blood Pressure (mm Hg) (a)
70
80 90 100 Usual Diastolic Blood Pressure (mm Hg)
110
(b)
Figure 12-10 Stroke Mortality Rate in Each Decade of Age Versus Usual Blood Pressure at Start of That Decade. Source: Reprinted with permission from The Lancet, Vol 360, Prospective Studies Collaboration, p 1906.
throughout the world (Table 12-10).82 Subregions represent gradations of mortality considered separately for children and adults—from “A” for low mortality in both groups to “E” for mortality that is high in children and very high in adults. The attributable fractions of CHD burden, attributable deaths, and attributed DALYs represent effects of systolic blood pressures exceeding 115 mm Hg, as in the preceding example. Measures of attributable burden ranged from 41 to 64% across subregions. In Southeast Asia, the most heavily afflicted area, this represents 650,000 deaths and nearly 7.1 million DALYs. A more comprehensive analysis of bloodpressure-related disease, on a global level, considered the impact in 2001 of systolic blood pressure above 115 mm Hg on stroke, ischemic heart disease, hypertensive disease, and other cardiovascular disease.83 It was estimated that 7.6 million premature deaths and 92 million DALYs were attributable to elevated blood pressure that year, comprising 13.5% of deaths and 6.0% of all DALYs worldwide. Further emphasis was given to the point that a large proportion of this bur-
den was due to blood pressure not considered hypertensive, but in the prehypertensive range. Further, some 80% of the burden occurred in low- and middle-income countries. The majority occurred among people in the 45–69 year age range, largely within the working age population.
RELATION TO OTHER FACTORS Nature of the Relations The relation of high blood pressure to other factors in determining rates and risks of cardiovascular consequences warrants at least brief comment. In the case of some types of stroke, heart failure, and renal failure, blood pressure is the clearly dominant factor. It may be considered the sine qua non for development of hemorrhagic stroke, hypertensive heart disease, hypertensive renal disease, and some other vascular disorders. These are more than tautological relations in the sense that, in the absence of uncontrolled high blood pressure, a substantial subset of cardiovascular
Western Pacific A B 52 41 51 291 400 2664
49 2991 28,201
All
Source: Reprinted with permission from Coronary Heart Disease Epidemiology: From Aetiology to Public Health, M Marmot, P Elliott, eds, CMM Lawes, S Vander Hoorn, MR Law, P Elliott, S MacMahon, A Rodgers, p 160. © 2005 Oxford University Press.
Data from the Global Burden of Disease 2000 study (Ezzati et al. 2002; Lawes et al. 2004, WHO 2002)
C 63 603 5239
South East Asia B D 46 41 100 650 1050 7080
3:40 PM
CHD-Attributable Deaths and DALYs for Non-Optimal Blood Pressure by Subregion Eastern Africa Americas Mediterranean Europe D E A B D B D A B Attributable fraction 57 46 44 50 45 55 51 54 64 Attributable deaths (000s) 82 67 203 129 13 71 176 290 263 Attributable DALYs (000s) 893 760 1548 1303 132 811 1922 2079 2320
340
2/25/10
Table 12-10
46894_CH12_311_360.qxd Page 340
CHAPTER 12 HIGH BLOOD PRESSURE
46894_CH12_311_360.qxd
2/25/10
3:40 PM
Page 341
PREVENTION AND CONTROL 341
disease—the hypertensive diseases—would not occur. The contribution of high blood pressure to the atherosclerotic diseases is more complex. Here the role is understood to be one of potentiating the effects of other factors, principally adverse blood lipids. Its role as an independent risk factor in determining population rates and individual risks of atherosclerosis, beginning in childhood, and the clinical events of advanced disease in later life is extensively documented in many populations. But this is statistical independence in the context of multiple interrelated factors, each of which makes its contribution to the adverse cardiovascular outcomes. Implications for Prevention and Control As discussed previously, high blood pressure is itself a reflection of other factors addressed in earlier or subsequent chapters—dietary imbalance, including excess intake of sodium, alcohol, and calories (with inadequate intake of potassium); physical inactivity and lack of physical fitness; overweight and obesity; and the social and environmental circumstances that contribute to these underlying influences. This intermediate position of high blood pressure in the causal pathway of the hypertensive and atherosclerotic diseases implies that, in principle, high blood pressure itself can be prevented, as a high public health priority. However, failing prevention of high blood pressure itself, effective control can greatly diminish its major consequences. Both aspects are addressed in the following section. One further note regarding the relation of high blood pressure to other factors concerns the longestablished understanding of “multi-factorial” causation and its more recent representation in the concept of “global risk.” At the individual level, this concept takes multiple factors into account to determine who shall be treated to reduce cardiovascular risk. This topic is addressed in some detail in Part IV, “Causation and Prevention: Theory, Practice, and Further Research.”
PREVENTION AND CONTROL Relative to longstanding recognition of hypertension as a clinical problem, understanding blood pressure as a population-level phenomenon appears to be quite recent—from only a few decades ago. The work of Pickering, noted previously, and others who conducted early surveys to examine the distribution of blood pressure in populations, gave impetus to what has become an accepted public health perception of the problem of high blood pressure—a serious con-
dition in terms of its consequences, widespread in the population with marked disparities in its occurrence, and in principle preventable by interventions at both individual and community or population-wide levels. Over these same recent decades, interventions to reduce already elevated blood pressure became vastly more effective than earlier therapies. Pharmacologic advances provide a wide array of effective bloodpressure-lowering medications, and changes in individual behaviors especially regarding nutrition and physical activity have been shown to be effective as well. Despite this progress in clinical management, actual blood pressure control is not being achieved for the great majority of persons with high blood pressure, as clearly shown in the preceding. The general failure to achieve control makes prevention of high blood pressure a practical necessity if the potential for preventing major consequences is to be achieved. Other considerations add to make prevention— and not sole reliance on control—a compelling public health priority, including the following:84 Significant vascular disease may develop before high blood pressure is detected; treatment does not reduce risk to the level of persons without high blood pressure; mass use of drugs by individuals requiring decades-long treatment is unlikely to be free of risk; and neglect of prevention implies continuing incidence of new cases and persisting need for detection and control. “Primary prevention of hypertension” distinguishes strategies aimed at averting development of high blood pressure in the first place from strategies to control it or to prevent its cardiovascular consequences. This emphasis was underscored in a 1983 report from WHO, Primary Prevention of Essential Hypertension.85 That report reviewed the natural history of blood pressure elevations; genetic aspects (chiefly family history as evidence of heritability); and several environmental influences, including body weight, salt, and other dietary factors, alcohol, physical activity, and psychological and social influences. Recommendations focused on research, including controlled community trials, that were needed to determine feasibility and effectiveness of interventions. Only in populations with high salt intake or high prevalence of overweight were immediate interventions recommended, and these were limited to public information on the connection of these factors with high blood pressure and individual counseling of those at increased risk of developing it. In the United States, a decade later, the National High Blood Pressure Education Program presented a report on primary prevention of hypertension.86 Elaborating on the rationale for this approach, the Working Group wrote:86, pp 188–189
46894_CH12_311_360.qxd
342
2/25/10
3:40 PM
Page 342
CHAPTER 12 HIGH BLOOD PRESSURE
. . . identification and treatment of the hypertensive patient represents an important but insufficient response to the problem of BP-related cardiovascular disease. Primary prevention is part of the long-term solution to the problem. It is a logical extension of our current efforts directed at detection and treatment of patients with established hypertension. It also provides opportunities to interrupt and reduce the continuing costly cycle of managing hypertension and its complications. The report advocated use of “intervention strategies that have shown promise” based on review of evidence from trials to modify excessive intake of sodium, calories, and alcohol; deficient intake of potassium; and physical inactivity. It was judged premature to conclude whether other dietary interventions or stress management were effective. Interventions to modify these factors in the general population (the population strategy) would have the objective of shifting the blood pressure distribution downward. Applied to groups most likely to develop hypertension—African Americans and persons with high normal blood pressure, a family history of hypertension, or lifestyle factors contributing to age-related increase in blood pressure aim of the “target strategy,” would be to lower blood pressure specifically in these groups. An update of this report in 2002 reviewed new evidence in support of the original recommendations and reiterated the dual population and target strategies.87 More specific recommendations were made for weight loss, dietary sodium reduction, increased physical activity, and moderation of alcohol intake. In addition, use of potassium supplementation was recommended, as was modification of whole diets, in accordance with experience of the Dietary Approaches to Stop Hypertension (DASH)—Sodium Trial. Supplementation with calcium, fish oil, and herbal or botanical products was discussed but considered to have uncertain or less proven efficacy; psychological interventions were not addressed. These reports focused nearly exclusively on adults. Concurrently, prevention of high blood pressure beginning in childhood received attention from WHO and from others, including the (US) National High Blood Pressure Education Program and American Heart Association.88–90 The 1985 WHO Study Group report, Blood Pressure Studies in Children, addressed preventive measures to be taken at both community and individual levels.88 Recommendations were similar to those for adults, adding specific reference to legislation to reduce sodium content of baby food and to the roles of teachers and
others who work with children in addition to health professionals. Individual intervention should focus on children with blood pressure that is persistently high, over several years, especially those with a family history of hypertension, or those with obesity, or “an unusually high” intake of salt. A second WHO report focusing on early intervention, discussed earlier, appeared in 1990: Prevention in Childhood and Youth of Adult Cardiovascular Diseases: Time for Action.89 Regarding hypertensive diseases, current data on related factors were summarized and recommendations were presented for both action and research. These were directed to prevention of atherosclerotic and hypertensive diseases together, not specifically to high blood pressure. The main components, however, were similar: improvement of eating patterns and exercise habits, and elimination of tobacco use. Recommendations from the National High Blood Pressure Education Program since the late 1970s and from the American Heart Association since the early 1990s have addressed approaches to detection and management of high blood pressure in children and adolescents, either alone or in the context of promoting cardiovascular health generally. These several reports are summarized elsewhere.90 Their intent has been to call attention to the potential for intervention in childhood and adolescence to prevent or detect increased risk of cardiovascular disease later in life. In addition to details of measurement and classification of blood pressure in the young (Table 12-2), these reports provide recommendations for action at the population or individual level, or both, that are generally consistent with those described previously. It is noteworthy that these longstanding recommendations extend well beyond those of the United States Preventive Services Task Force (USPSTF), which has not found evidence sufficient to justify screening for high blood pressure at ages younger than 18 years.91 (See discussion of USPSTF and other guidelines in Chapter 20, “Recommendations, Guidelines, and Policies.”) Individual Measures Prevention of high blood pressure is, as argued above, primarily dependent on habits of diet and physical activity that promote health generally and cardiovascular disease in particular. The expression “lifestyle modification” is widely accepted as representing the need for improvement in these and other behaviors, including tobacco use. Alternative terms include “behavior modification” and “therapeutic lifestyle change.” At the individual level, these lifestyle factors
46894_CH12_311_360.qxd
2/25/10
3:40 PM
Page 343
PREVENTION AND CONTROL 343
are typically considered as requiring reversal, from currently adverse to favorable patterns of behavior. The concept of promoting and maintaining favorable behaviors in the first place is less often encountered. From this perspective, several components of diet, as well as physical activity, weight reduction, and biofeedback and relaxation were reviewed in Lifestyle Modification for the Prevention and Treatment of Hypertension.92 To realize the potential impact of weight reduction was judged to depend upon identifying and implementing effective approaches to obesity prevention and control.93 Dietary salt reduction depends upon individual awareness, health providers’ advice to patients, and effective action by the food manufacturing and food service industries.94 Routine assessment of alcohol intake is proposed. When excessive intake is found, it is further proposed that a trial period of substantial reduction be observed before concluding that persistent, treatable hypertension is present.95 Maintaining recommended levels of physical activity on a continuous basis is judged to reduce the risk of developing high blood pressure.96 From these and other sources discussed previously, multiple targets of intervention are available to reduce individual risk of developing high blood pressure, or to reduce blood pressure levels once elevated. Extensive evidence supports these interventions. Of special significance is the demonstration of reduced blood pressure, in persons with or without hypertension, through dietary modification in the Dietary Approaches to Stop Hypertension (DASH) Trial.97 Figure 12-11 demonstrates the effect of a diet rich in vegetables, fruits, and low-fat dairy products
DASH-Sodium Trial: Effects of Dietary Changes on Blood Pressure* Systolic Blood Pressure
135
Control Diet 2.1
130
6.7 p.0001
4.6
125
1.7
1.3
3.0 p.0001 DASH Diet
120 1.5 (65)
2.4 (106)
3.3 (143)
Sodium Level: gm/d (mmol) per Day *Adapted from: Sacks, 2001 (412 prehypertensive and hypertensive adults)
Figure 12-11 DASH-Sodium Trial: Effects of Dietary Changes on Blood Pressure. Source: Adapted with permission from New Engl J Med, Vol 344, Sacks FM , Svetkey LP, Vollmer WM et al., pp 3–10. © 2001 Massachusetts Medical Society.
(the DASH diet) at three levels of sodium intake. This adapted representation of the results portrays the low-sodium DASH diet as the reference condition and demonstrates the effects of the “control diet”— typical for the United States—and successively higher sodium intakes in increasing blood pressure above the DASH diet. At the lowest sodium level, the control diet was associated with systolic blood pressure higher by 2.0 mm Hg. Increases in sodium content added another 6.7 mm Hg in average systolic pressure, for a total increase of 8.7 mm Hg for the highestsodium “control” diet versus the lowest-sodium DASH diet. This trial, with 412 participants in three successive 30-day periods of varying sodium content in the diet, has had a major impact in promotion of dietary intervention to prevent and control high blood pressure. The Trials of Hypertension Prevention (TOHP) sought to reduce sodium intake to 80 mmol/day (approximately 1800 mg Na) in adults aged 30–54 years with initial diastolic blood pressure 80–89 mm Hg.98 Sodium reductions in the 314 participants randomly allocated to active intervention were from 154.6 to 99.4 mmol/day (about 3500 to 2300 mg Na/day) at the 18-month follow-up visit, with significant reductions in both systolic and diastolic blood pressure. In addition, a recent review has considered the safety of dietary sodium reduction in view of concerns raised from time to time.36 The authors concluded, “Overall, we identified extensive data supporting the safety of public health recommendations for moderate Na reduction and none suggesting cause for concern.”36, p 192 Importantly, evidence that sodium reduction can have a long-term impact on cardiovascular outcomes and not only on reduction of blood pressure has been presented from long-term follow-up of the TOHP participants.99 Studies specifically of reduced-salt dietary interventions have received great interest and have been reviewed on several occasions. For example, Cutler and colleagues recently conducted an overview of 32 trials of blood pressure reduction with such interventions.100 The studies were mixed in design and selection criteria but showed significant reductions in systolic and diastolic pressure in the several subgroups of studies examined. Their overall estimates of effects per 100 mmol of sodium intake per day were reductions (systolic/diastolic) of 5.8/2.5 mm Hg among hypertensive subjects and 2.3/1.4 mm Hg among normotensives. They noted no evidence of safety hazards in the context of these trials and projected that substantial lowering of sodium intake in the US population at large could reduce cardiovascular morbidity and mortality. (However, published
46894_CH12_311_360.qxd
344
2/25/10
3:40 PM
Page 344
CHAPTER 12 HIGH BLOOD PRESSURE
reviews on a given topic can come to different conclusions.101,102) Related observations adding to the case for reducing sodium intake include a consistent finding in short-term trials in infants and children of blood pressure reduction with decreased sodium intake.103 For focusing specifically on management of high blood pressure once established, the treatment algorithm presented in JNC VII begins with reference to lifestyle modifications (Figure 12-12):7, p 25 “Adoption of healthy lifestyles by all persons is critical for the prevention of high BP and is an indispensable part of the management of those with
hypertension.” Very limited guidance is presented as to how behavior change is to be achieved or how long a trial period is needed to determine success. But if the treatment goal of 140/90 mm Hg (or lower for persons with diabetes or chronic kidney disease) is not reached, drug treatment is to be initiated and progressively intensified until the blood pressure goal is attained. Still considering lifestyle approaches, a meta-analysis of 105 trials, 8 weeks to 1 year in duration and having aggregate numbers of participants ranging from 400–1500 per intervention, reconfirmed efficacy of improved diet, aerobic exercise, and alcohol and sodium restriction.104
LIFESTYLE MODIFICATIONS
Not at Goal Blood Pressure (,140/90 mm Hg) (,130/80 mm Hg) for patients with diabetes or chronic kidney disease)
INITIAL DRUG CHOICES
Without Compelling Indications
With Compelling Indications
Stage 1 Hypertension (SBP 140–159 or DBP 90–99 mmHg)
Stage 2 Hypertension (SBP $ 160 or DBP $ 100 mmHg)
Drugs(s) for the compelling Indications (See table 8)
Thiazide-type diuretics for most. May consider ACEI, ARB, BB, CCB, or combination.
Two-drug combination for most (usually thiazidetype diuretic and ACEI, or ARB, or BB, or CCB).
Other antihypertensive drugs (diuretics, ACEI, ARB, BB, CCB) as needed.
NOT AT GOAL BLOOD PRESSURE Optimize dosages or add additional drugs until goal blood pressure is achieved. Consider consultation with hypertension specialist. DBP - Diastolic Blood Pressure; SBP - Systolic Blood Pressure. Drug Abbreviations: ACEI - Angiotensin Converting Enzyme Inhibitor; ARB - Angiotensin Receptor Blocker; BB - Beta Blocker; CCB - Calcium Channel Blocker.
Figure 12-12 Algorithm for Treatment of Hypertension. Source: JNC Express. The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation and Treatment of High Blood Pressure, National Heart, Lung and Blood Institute, National Institutes for Health, NIH Publication No. 03-5233, May 2003, p 13.
46894_CH12_311_360.qxd
2/25/10
3:40 PM
Page 345
PREVENTION AND CONTROL 345
Another aspect of lifestyle change concerns people who report they are taking personal actions to control their blood pressure after having been told on two or more visits to a health professional of having the condition. The Behavioral Risk Factor Surveillance System (BRFSS) in 2005 included a questionnaire module in 20 states asking whether the respondent was changing any of five behaviors to control blood pressure.105 Results by age, sex, and race/ethnicity are summarized in Table 12-11, showing overall frequencies of “changing your eating habits,” 70.9%; “cutting down on salt,” 79.5%; “reducing alcohol use,” 79.2%; “exercising,” 68.6%; and “taking antihypertensive medicine,” 73.2%. These responses suggest a high level of awareness of appropriate actions, especially among older persons, women, and nonHispanic Blacks. If these actions are being practiced effectively, they may be contributing to the population-wide distribution and trends of blood pressure. With respect to pharmacotherapy, few effective drugs were available before the 1960s, when new agents began to be evaluated through the first randomized control trials in this field. The Veterans Administration hospital system in the United States provided the institutional framework for landmark trials establishing the efficacy and safety of antihypertensive medications.106 Reports in 1967 and 1970 demonstrated that patients with baseline diastolic blood pressures from 115–129 mm Hg, then from 90–114 mm Hg, benefited from treatment in terms of reduced morbidity and mortality.107 The question remained of whether these results, based on the experience of highly selected patients with sustained in-hospital elevation of blood pressure, were applicable to the general population. The Hypertension Detection and Follow-Up Program (HDFP) was designed to address this question by screening for high blood pressure in 14 communities throughout the United States; enrolling persons with qualifying levels of blood pressure in the trial; referring half of them to available sources of treatment in their communities (the “referred care” group); and providing the other half (the “stepped care” group) with an intensified treatment and followup program.108,109 Mortality was reduced by 17–20% overall in stepped versus referred care, and benefit was shown in most age and sex–race subgroups. The HDFP findings were consistent with other studies of “mild hypertension” undertaken in the United Kingdom, Europe, and Australia during the same period of the 1970s. A further unanswered question was whether older persons with high blood pressure, being at especially high risk of stroke, would benefit or be harmed by
blood pressure lowering. The Systolic Hypertension in the Elderly Program (SHEP) was conducted in the 1980s and found the stroke rate to be reduced by 36% among persons aged 60 years and older with isolated systolic hypertension.110 There were no offsetting serious adverse effects of reducing blood pressure in these patients. Again, consistent findings of studies elsewhere added to the growing consensus that antihypertensive drug therapy was beneficial across a wide range of blood pressure levels and across major segments of the general population. As knowledge regarding treatment of high blood pressure continued to develop with new agents becoming available, practice guidelines evolved in the United States, the United Kingdom, Europe, and elsewhere. These have generally taken the form of recommending assessment of each of the major risk factors for atherosclerotic and hypertensive diseases, use of a multifactor prediction equation or analogous procedure to estimate risk of a coronary or cardiovascular event within 10 years, taking each of these factors into account, and treatment of those at high levels of risk. (See Chapter 17, “Recommendations, Guidelines, and Policies.”) A countervailing approach is also advocated—to disregard the absolute level of blood pressure altogether on the premise that “lower is better” and provide combination drug therapy to all adults older than age 55 years to lower blood pressure and cholesterol and reduce blood platelet aggregation to prevent clot formation. (See Chapter 18, “The Case for Prevention.”) Regardless of these developments, it remains important to ensure that persons for whom intervention is recommended are in fact receiving the intervention and the expected benefit. This raises again the question of high blood pressure control. The proportion of respondents to the BRFSS questions reporting use of medication to lower their blood pressure was much greater (overall, 73.2%) than those reported in NHANES only 5 years earlier as having high blood pressure and being treated (overall, 45.3%).105 The typical finding is much closer to this earlier estimate, being one key factor in the inadequate rate of high blood pressure control in nearly all populations. Even in the more nearly ideal conditions of a clinical trial, treatment and control are not universally achieved. In the HDFP, whose participants were drawn from the general population, over the 5 years of the trial, at most 86% of stepped care and 58% of referred care participants were on treatment.108 The treatment goal was a diastolic pressure below 90 mm Hg, or 10 mm Hg below a baseline pressure from 90–99 mm Hg. The proportion of participants at goal blood pressure at year 5 of the trial
State Alabama Arizona Arkansas Connecticut Florida Georgia Hawaii Kansas Kentucky Louisiana Maryland
70.7 73.1 65.9 63.8 75.3
1045 374 89 303 162 76.1 70.2 64.4 76.1 73.5 70.6 74.5 68.6 74.4 75.9 75.3
77.5
2769
912 993 1407 1065 2026 1568 558 1000 1771 739 989
69.5
66.7 76.4
71.9–80.3 62.1–78.3 57.7–71.0 71.3–80.8 66.1–80.8 65.6–75.7 65.6–83.4 61.1–76.0 69.1–79.8 69.9–81.9 68.4–82.2
57.1–93.6
53.5–74.0
63.5–77.9 61.2–85.0 47.3–84.5
72.7–82.3
66.8–72.1
63.8–69.6 74.2–78.6
32.3–62.1 73.3–78.2 75.5–78.0 65.7–68.5
—†† 73.7 75.8 82.2 83.2 —†† 83.8 75.8 81.4 81.4 81.1
71.5
76.8
73.9 80.9 78.5
90.0
79.0
77.0 82.0
68.2 78.9 82.1 85.0
— 69.1–78.2 69.4–82.1 77.6–86.7 79.3–87.0 — 77.8–89.8 67.0–84.5 76.3–86.5 75.9–87.0 76.5–85.8
52.9–90.0
69.6–83.9
66.8–80.9 69.1–92.8 57.0–100.0
86.5–93.5
76.7–81.2
73.6–80.3 79.3–84.7
52.4–84.0 76.1–81.6 80.8–83.4 83.8–86.1
88.3 77.0 83.0 70.3 81.3 79.1 71.6 83.4 80.3 85.0 78.5
82.5
79.6
84.3 79.3 74.5
86.9
76.6
75.5 82.5
77.2 79.9 79.6 78.6
85.5–91.0 68.6–85.4 77.2–88.9 64.4–76.1 77.8–84.7 73.1–85.0 63.1–80.1 76.4–90.5 75.6–84.9 80.2–89.7 73.8–83.1
70.2–94.8
71.5–87.7
79.0–89.5 69.0–89.6 63.6–85.5
83.2–90.5
74.6–78.6
73.2–77.8 80.2–84.8
66.8–87.6 77.6–82.1 78.4–80.9 77.3–79.8
66.7 73.0 72.1 74.0 74.6 65.8 63.2 79.9 59.9 69.7 70.8
74.8
76.0
66.8 72.0 77.1
67.5
69.4
68.8 69.0
65.9 70.9 68.6 65.2
62.3–71.1 64.7–81.4 66.1–78.0 69.0–79.0 71.2–77.9 59.6–71.9 53.8–72.6 76.5–83.2 54.7–65.1 63.6–75.8 63.7–77.8
62.1–87.5
66.8–85.1
60.1–73.5 60.9–83.0 62.6–91.5
62.2–72.7
67.2–71.5
65.4–72.2 66.6–71.5
50.1–81.8 68.4–73.4 67.2–70.0 63.8–66.6
81.4 68.2 72.3 69.9 73.7 74.0 76.5 76.0 78.3 85.8 76.7
75.0
61.3
62.5 77.4 63.1
75.2
75.9
71.1 76.3
35.3 64.6 88.7 96.2
72.8–90.1 59.2–77.1 65.8–78.8 65.1–74.7 65.7–81.7 69.4–78.6 66.6–86.4 67.1–84.8 73.4–83.2 80.4–91.2 69.6–83.8
57.1–92.9
49.2–73.4
57.1–67.8 65.4–89.4 52.5–73.8
71.2–79.2
73.1–78.7
67.7–74.5 73.8–78.8
20.4–50.1 61.7–67.6 87.8–89.6 95.6–96.8
3:40 PM
19,705
9077 15,370
Sex Men Women
47.2 75.7 76.7 67.1
2/25/10
Race/Ethnicity White, non-Hispanic Black, non-Hispanic Hispanic** Asian Native Hawaiian/ Pacific Islander American Indian/ Alaska Native Other
129 2694 10,889 10,735
Number and Percentage of Respondents Taking Selected Actions to Control High Blood Pressure (HBP) Among Adults Told on Two or More Visits to a Health Professional That They Have HBP, by Selected Characteristics—Behavioral Risk Factor Surveillance System, 20 States, 2005 Action taken to control HBP Changing Reducing Use of Reducing Use of Taking Antihypertensive Eating Habits or Not Using Salt or Not Drinking Alcohol Exercising Medication Total No. of Respondents† %§ 95% CI¶ % 95% CI % 95% CI % 95% CI % 95% CI
346
Characteristic Age group (yrs) 18–24 25–44 45–64 65
Table 12-11
46894_CH12_311_360.qxd Page 346
CHAPTER 12 HIGH BLOOD PRESSURE
24,447
70.9
68.7–73.1
58.9–68.8 77.6–83.4 53.5–65.8 56.7–73.7 64.4–74.2 60.4–73.2 57.9–73.0 63.0–75.3 60.0–76.4 79.5
69.5 —†† 73.1 82.8 80.9 77.5 76.9 73.7 75.4 77.1–81.9
64.4–74.5 — 64.8–81.5 78.5–87.0 76.2–85.6 70.5–84.6 71.2–82.5 68.4–78.9 67.7–83.2 79.2
61.4 75.5 69.5 75.6 75.4 79.2 74.3 85.1 84.5 77.6–80.9
56.0–66.7 69.4–81.6 60.8–78.2 67.8–83.3 70.5–80.2 75.0–83.3 66.7–81.9 80.0–90.2 77.2–91.9 68.6
60.8 69.5 76.7 66.5 70.0 65.8 75.4 75.5 57.6
66.3–70.9
55.3–66.2 66.1–72.9 71.6–81.7 58.0–75.1 65.7–74.2 60.2–71.4 69.7–81.0 69.9–81.1 49.3–65.9
73.4
78.9 81.0 58.1 72.7 71.7 66.2 74.3 68.9 70.9
71.2–75.7
73.2–84.5 74.5–87.6 52.9–63.3 64.1–81.2 67.5–75.8 60.9–71.5 68.5–80.2 62.8–74.9 67.1–74.7
Source: Adapted from MMWR, Vol 56, 2007, p 421.
*Respondents were asked the following five questions: “Are you changing your eating habits to help lower or control your high blood pressure?” “Are you cutting down on salt to help lower or control your high blood pressure?” “Are you reducing alcohol use to help lower or control your high blood pressure?” “Are you exercising to help lower or control your high blood pressure?” “Are you currently taking medicine for your high blood pressure?” † The number of respondents in the salt-use column is lower because of missing values for three states. § Weighted percentages, except for age groups, are age standardized to the 2000 US standard population. ¶ Confidence interval. **Might be of any race. †† Data not comparable for this question because of different response categories.
Total
63.9 80.5 59.7 65.2 69.3 66.8 65.5 69.1 68.2
3:40 PM
603 1338 1002 1012 2978 1677 851 915 1043
2/25/10
Minnesota Mississippi Montana Nebraska New Jersey New York North Dakota Utah West Virginia
46894_CH12_311_360.qxd Page 347
PREVENTION AND CONTROL 347
46894_CH12_311_360.qxd
348
2/25/10
3:40 PM
Page 348
CHAPTER 12 HIGH BLOOD PRESSURE
was 64.9% in stepped care and 43.6% in referred care. The Antihypertensive and Lipid-Lowering Heart Attack Trial (ALLHAT) was based in 623 clinical practice centers, where 90% of participants were already on treatment for high blood pressure at entry to the trial (beginning in 1994).111 Even here, the final proportion of participants controlled ( 140/90 mm Hg) was only 66%, although it was greatly increased from 27.4% at baseline. Data from NHANES III (1988–1994) showed that greater frequency of high blood pressure control (at best, 20.5–28.2% among non-Hispanic Whites) was associated with indicators of regular health care (private health insurance, same healthcare facility and provider, semi-annual, or annual blood pressure checks).112 Self-reported adoption of lifestyle changes (weight loss, sodium reduction, and exercise) was even more strongly related (at best, 37.8–39.7% control among non-Hispanic Whites). The HealthStyles survey is a commercial postal survey conducted annually among US adults aged 18 years and older.113 The 2005 survey included more than 1400 persons for whom antihypertensive medications had been prescribed. About 28% of respondents indicated having difficulties in taking this medication. Major factors were “not remembering,” cost, and lack of insurance; other factors were side effects, thinking it was not needed, and having no healthcare provider. For assessing evidence for interventions to improve high blood pressure control, a Cochrane Review (see Chapter 19, “Evidence and Decision Making”) was conducted that identified 56 randomized control trials of several approaches: patient selfmonitoring, patient or provider education, care led by a nonphysician health professional, organizational improvements in care delivery, or appointment reminder systems.114 The HDFP was prominent in this review, contributing to the observation that “An organized system of registration, recall and regular review allied to vigorous stepped care approach to antihypertensive drug treatment appears the most likely way to improve the control of high blood pressure.”114, p 2 Results of other interventions were less consistent. Studies of interventions to control increasing blood pressure in childhood and adolescence have had very mixed results, in part reflecting differences in selection of participants, interventions, and study designs. Most trials have been of nonpharmacologic interventions. One drug trial was conducted in 8- to 18-year-olds who had greater than 90th percentile values of blood pressure repeatedly over a 4-month period.115 Low-dose medication was combined with
education on diet and physical activity. With evaluation at the close of the treatment period of 30 months, significant reductions in systolic (3.59 mm Hg) and diastolic (1.73 mm Hg) pressure were reported. Overall it is not yet established, however, that longterm change in blood pressure is influenced by these interventions applied in childhood and adolescence. This is an important area for research if prevention of high blood pressure in the first place is to be achieved. Community and Population-Wide Measures Effective and sustained management of high blood pressure in individuals reduces the prevalence of the highest levels of risk. But only intervention for the population as a whole, beginning in childhood, can reduce the excessive rise in blood pressure with age and the public health burden of increased cardiovascular risk and the need for treatment. The goals of community or population-wide interventions are both to control the incidence of new cases of high blood pressure, resulting in fewer new cases, and to reduce its prevalence. Success therefore depends both on influencing modifiable causes of progression in blood pressure levels and on reinforcing behavior by health professionals and affected individuals to ensure detection and control of already-established high blood pressure. Interventions for these purposes have been advocated and implemented at least since the early 1970s. For example, beginning in 1973, the National High Blood Pressure Education Program (NHBPEP) has emphasized public and professional education as a means to promote the detection, evaluation, and management of high blood pressure.116 The World Health Organization has also recommended national policies for prevention of high blood pressure, both in its original report on primary prevention of hypertension and in a subsequent Expert Committee Report, Hypertension Control, with discussion of population-level measures and policies.85,117 The central concept in this discussion is that of the whole distribution of blood pressure in the population and the prospect of favorable or unfavorable changes in the distribution:117, p 44 The fact that the distribution of blood pressure in many developing countries is to the left of that in developed countries makes the task principally one of preventing a shift to the right. This will probably be easier than achieving a shift to the left in developed countries. This makes primary prevention the major goal in developing countries, using a population-based strategy encouraging changes in lifestyle.
46894_CH12_311_360.qxd
2/25/10
3:40 PM
Page 349
PREVENTION AND CONTROL 349
Recommended measures for lifestyle change, based on evidence reviewed in the report, were weight reduction, reduction of alcohol intake, increased physical activity, and reduced sodium intake. These were to be pursued through public, professional, and patient education. More broadly, a comprehensive hypertension control policy should be developed in each country to address both whole communities and individuals with established high blood pressure. Elements of such national policies could include:117, p 51 • comprehensive health education programs for communities including children; • agricultural policies that help ensure that potassium-rich natural foods, like fresh fruit and vegetables, are readily available and inexpensive; • regulation of the food industry to promote the availability of prepared food items with a lowsalt and low-fat content, and labeling of marketed items for salt and fat content; • providing facilities for outdoor recreational sports and leisure time; • control of tobacco smoking which is an important additional risk factor for cardiovascular disease; • ensuring the availability of inexpensive but effective drugs for lowering blood pressure; • integration of programs for blood pressure detection, treatment, and control into the various levels of healthcare services, especially primary care. Attention is called to nongovernmental national and regional organizations whose purpose is to foster high blood pressure prevention and control. The International Society of Hypertension (http://www .ish-world.com) and World Hypertension League (http://www.worldhypertensionleague.org) are prominent examples. Organizations specifically addressing salt include the Center for Science in the Public Interest (CSPI, http://www.cspinet.org) and World Action on Salt & Health (WASH, www.worldactionsalt.org). Community intervention to improve high blood pressure control has been undertaken in a variety of settings and with differing approaches. Beginning in the early 1970s, the North Karelia Project, in Finland, brought about reduction in blood pressure and other cardiovascular risk factors through a comprehensive program of changes in policy, environment, healthcare practices, and individual behaviors.61 These included reduction of sodium intake and other changes in diet and achieved a mean decrease of systolic and diastolic blood pressure of 7/7 mm Hg, accompanied
by decreases in mortality from stroke and coronary heart disease. Three US community studies conducted mainly in the 1980s (with an earlier phase in the Stanford study) observed generally favorable blood pressure changes in intervention communities relative to controls, principally through communications and community organization.118 Some blood pressure changes in intervention communities were substantial, but blood pressure was decreasing in control communities as in the nation as a whole during this period, and few of the differences were statistically significant. Also in the 1980s, in two Portuguese communities, education about the benefits of reducing salt in cooking (especially bakery breads) and limiting consumption of particularly high-salt foods led to marked changes in salt intake and levels of blood pressure between the intervention and the control communities.119 Sodium consumption was reduced from 364 to 202 mmol/day (from nearly 8400 to 4600 mg Na/day) in the intervention community. In the control community, sodium consumption increased from 352 to 371 mmol/day (more than a 400 mg Na/day increase). The corresponding net decreases of systolic and diastolic blood pressure in the intervention community relative to the control community were 13.3 and 6.1 mm Hg, respectively. In Georgia (US), the state-funded Stroke and Heart Attack Prevention Program (SHAPP) was implemented to improve high blood pressure control among medically indigent patients. SHAPP provided needed medical services, including screening, lifestyle counseling and education, low- or no-cost medications, and follow-up.120,121 In two of the most effective clinics, 68 and 60% of persons on medication were controlled, respectively, with a range from 41 to 68% across the entire program. The above-average rate of control was estimated to result in substantial cost savings due to averted strokes and heart attacks. The Asheville Project, in North Carolina, was a “medication therapy management” program to address high blood pressure and cholesterol.122 Participants were city or hospital employees or their spouses or dependents. Pharmacists were reimbursed by health plans to provide patient consultations, monitoring, and recommendations to physicians regarding patient management. Substantial reductions in systolic and diastolic blood pressure (11/5 mm Hg) were observed, with an increase in the proportion controlled from 40.2 to 67.4%. Despite increased medication costs for the program, total medical costs decreased by about 30% relative to previous experience. A model community organization to implement effective population-level and individual-level measures
46894_CH12_311_360.qxd
350
2/25/10
3:40 PM
Page 350
CHAPTER 12 HIGH BLOOD PRESSURE
for prevention and control of high blood pressure was presented by McClellan and Wilber, based on the experience of numerous programs in the United States (Figure 12-13).123 Their message was that an effective community-oriented agency must have linkages of several kinds with affected individuals, providers, and political units and their agencies concerned with the problem.
sons with diastolic pressure 95 mm Hg by 7%, by medication.125 The effect of each of the two strategies would be to reduce stroke and coronary heart disease deaths by 16% and 5%, respectively, and preventing 1 million deaths per year throughout Asia in 2020. Two million deaths would be prevented if both strategies were implemented. A cost-effectiveness analysis of interventions to lower systolic blood pressure and cholesterol in order to reduce disease burden (in DALYs) found that “nonpersonal health interventions” such as government action to lower salt content of processed foods, could avert more than 21 million DALYs worldwide every year.126 Combination drug therapy for persons with greater than 35% risk of a cardiovascular event within 10 years was also found cost-effective, with potential to avert 63 million DALYs per year worldwide. Policy decisions regarding each of these approaches, or combinations of them, are needed on a country-bycountry basis, depending on resources and other considerations of feasibility. But the potential magnitude of benefit is great. A policy to reduce salt intake in the population of 23 developing countries that account for 80% of the world chronic disease burden was evaluated as to its cost and its impact on chronic disease mortality.127 It was found that, by reducing salt intake by 15%, 8.5
Global Strategies Recommendations of the World Health Organization are directed to all Member States and are therefore in a sense global in their intent. The 2007 WHO guidelines, Prevention of Cardiovascular Diseases, are presented with the purpose “to scale up cost-effective, integrated approaches for prevention of CVD” as part of the Global Strategy for the Prevention and Control of Noncommunicable Diseases.124, pp 2–3 Hypertension is addressed from a clinical perspective, with target levels for treatment of individual cases and description of the various classes of antihypertensive agents, including their clinical indications, grading of evidence for their use, contraindications, and cautions. An analysis of strategies for blood pressure reduction in Eastern Asia estimated the potential impact of reducing diastolic blood pressure population-wide by 2%, by diet and lifestyle changes, and among per-
Community Health System
service Aware, motivated hypertensive patient
1. 2.
3.
Community resources
Regional, state, and federal health systems
Inf or
ion
ma
at
tio
m or nf
n
I
Sc
re
en Re ing fer r a Ed l uc ati on
Community hypertension control program
Information
Interface
Prepared provider
utilization
1. Providerpatient interaction
Initiation Resources Information
l na sio n s e o n of ati io Pr uc at d e m r 1. fo In 2. Community assessment
2. Community program— local catalyst
3. Regional/ state/federal initiatives— catalyst to local action
Figure 12-13 Schematic View of a Model of Community Hypertension Control Program. Source: From Chronic Disease Epidemiology and Control. Copyright 1993 by the American Public Health Association. Reprinted with permission.
46894_CH12_311_360.qxd
2/25/10
3:40 PM
Page 351
CURRENT ISSUES 351
million deaths could be averted over the 10 years from 2006–2015. Reducing intake to 5 g salt/day (2000 mg Na/day, the current WHO recommendation) would avert 28 million deaths over the same period. Costs for reducing salt consumption were estimated at $0.04–$0.32 per person per year across the 23 countries. Elements of cost would be awareness campaigns and management and supervision of saltreduction programs. In addition to WHO, other international organizations are addressing hypertension on a global level. The International Society of Hypertension and its member organizations are dedicated to detection and treatment of high blood pressure, with emphasis on research and on physician and patient education.128 World Action on Salt & Health has “the aim of bringing about a gradual reduction in salt intake throughout the world.”129, p 1 Although the missions differ among these and other particular organizations in the cardiovascular and chronic disease prevention arena, there is potential synergy in their activities and a growing base of evidence on which to build a more effective global effort.
CURRENT ISSUES Prevention of High Blood Pressure in the First Place The greatest public health challenge in relation to blood pressure is the prevention of this risk factor in the first place. Longstanding recommendations for intervention to prevent the unwanted rise of blood pressure with age remain to be implemented adequately. Meanwhile, overweight and obesity in childhood and adolescence, as well as in adults at all ages, are contributing to increasing levels of blood pressure. More effective efforts to reverse the obesity epidemic are needed, and changes in dietary patterns specifically to reduce cardiovascular risk (such as the DASH-low sodium pattern) are needed as well to have the full potential benefit of preserving optimum blood pressure levels throughout the life course. By this approach, disparities in the burden of high blood pressure and its consequences, especially for African Americans, could be overcome before they begin. Control of Established High Blood Pressure in the Whole Population Control of already-established high blood pressure is currently attained for only a minority of affected persons in virtually every general population. Even in clinical settings with insured patient populations, at least 25% of persons with high blood pressure do
not have it controlled, and these are the most favorable circumstances. In the United States, marked disparities in control make the picture even worse for Mexican Americans than others for whom data are available, and for some groups the extent of awareness, treatment, and control remains unknown. For many countries, the effort to achieve control of high blood pressure is at a very early stage. These are disturbing observations in view of the decades-old recognition of the problem, development of effective interventions, and knowledge of the failure to achieve agreed-upon national goals for improvement. Learning from the most successful control efforts, assessing the factors that limit success, and implementing policies and programs with the greatest feasible impact are steps that require effective public health action. This point has been expressed previously— notably by Richard Remington, celebrating the 50th anniversary of the Epidemiology Section of the American Public Health Association. His title was “High blood pressure control: what are the next steps?”:130, p 461 Certainly no landmark passed in these 50 years justifies complacency, a reduction of investigative efforts, or a decreased allocation of resources to improve the health of the people through high blood pressure control. Public health demands such efforts, and epidemiology can contribute to the efforts. That the message remains valid today—after still another 30 years—indicates that something more is needed than has been invested to date to bring about the needed further improvement in high blood pressure control. Dietary Salt Evidence is abundant that excessive consumption of salt increases blood pressure and reducing consumption lowers blood pressure at both elevated and nonelevated starting levels. Reducing salt consumption is a key to prevention of high blood pressure in the first place and to reducing it when elevated. In the United States and most industrial countries, the bulk of sodium in the diet is in processed or manufactured foods and in restaurant and fast food products. It is necessary to reduce salt intake to prevent and control high blood pressure, and the commercial food supply is the principal source to be addressed. In the United States, national guidelines specify for several subgroups of the adult population a lower level of intake—no more than 1500 mg Na/day—than for the population at large—2300 mg Na/day. These special
46894_CH12_311_360.qxd
352
2/25/10
3:40 PM
Page 352
CHAPTER 12 HIGH BLOOD PRESSURE
groups comprise persons with high blood pressure, African Americans, and middle-aged and older adults. It has been shown that these “subgroups” together constitute more than two-thirds of the adult population.131 Therefore this lower level recommendation applies to the great majority of US adults. It is unlikely to be attained by most people without substantial reduction in sodium content of the most widely consumed foods. Achievement of this goal in the United States and globally is a major issue for current attention. REFERENCES 1. Lloyd Jones D, Adams R, Carnethon M, et al. Heart disease and stroke statistics 2009 update. A report from the American Heart Association Statistics Committee and Stroke Statistics Committee. Circulation. 119:e21-e181. doi: 10.1161/CIRCULATIONAHA.108.191261. http://circ.ahajournals.org. Accessed December 18, 2008. 2. He FJ, MacGregor GA. Blood pressure is the most important cause of death and disability in the world. Eur Heart J Suppl. 2007;(suppl B): B23–B28.
Bethesda, MD: US Department of Health and Human Services. National Institutes of Health. National Heart, Lung and Blood Institute. National High Blood Pressure Education Program; August 2004. 8. Stamler J, Stamler R, Neaton JD. Blood pressure, systolic and diastolic, and cardiovascular risks. Arch Intern Med. 1993;153:598–615. 9. Mancia G, De Backer G, Dominiczak A, et al. 2007 Guidelines for the management of arterial hypertension. The Task Force for the Management of Arterial Hypertension of the European Society of Hypertension (ESH) and of the European Society of Cardiology (ESC). Eur Heart J. 2007;28:1462–1536. 10. US Department of Health and Human Services. Fourth Report on the Diagnosis, Evaluation, and Treatment of High Blood Pressure in Children and Adolescents. NIH Publication No. 05-5267. US Department of Health and Human Services, National Institutes of Health, National Heart, Lung and Blood Institute; May 2005.
3. US Department of Health, Education, and Welfare. Report of the Hypertension Task Force Volume One—General Summary and Recommendations. NIH Publication No. 79-1623. Bethesda, MD: US Department of Health, Education, and Welfare. Public Health Service. National Institutes of Health; September 1979.
11. Korotkov NS. A contribution to the problem of methods for the determination of the blood pressure. (English translation). In: Ruskin A, ed. Classics in Arterial Hypertension. Springfield, IL: Charles C Thomas; 1956:127–133.
4. Epstein FH. 1. The epidemiology of hypertension. In: Robertson JIS, ed. Handbook of Hypertension, Vol 1: Clinical Aspects of Essential Hypertension. Elsevier Science Publishers BV; 1983:1–20.
12. National High Blood Pressure Education Working Group on Hypertension Control in Children and Adolescents. Update on the 1987 Task Force Report on High Blood Pressure in Children and Adolescents: a Working Group report from the National High Blood Pressure Education Program. Pediatr. 1996;98:649–658.
5. Pickering GW. The Nature of Essential Hypertension. New York: Grune & Stratton; 1961. 6. Hypertension Detection and Follow-Up Cooperative Group. The Hypertension Detection and Follow-Up Program. A progress report. Circ Res. 1977;40(suppl I):I-107–I-109. 7. US Department of Health and Human Services. The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure. NIH Publication No. 04-5230.
13. Jones DW, Appel LJ, Sheps SG, Roccella EJ, Lenfant C. Measuring blood pressure accurately. New and persistent challenges. JAMA. 2003;289:1027–1030. 14. Rose G. Standardization of observers in blood pressure measurement. Lancet. 1965;1: 673–674. 15. Curb JD, Labarthe DR, Cooper SP, et al. Training and certification of blood pressure observers. Hypertension. 1983;5:610–614.
46894_CH12_311_360.qxd
2/25/10
3:40 PM
Page 353
REFERENCES 353
16. White WB. Clinical assessment of early morning blood pressure in patients with hypertension. Prev Cardiol. 2007;10:210–214. 17. Burt VL, Whelton P, Rocella EJ, et al. Prevalence of hypertension in the US adult population: results from the Third National Health and Nutrition Examination Survey, 1988–1991. Hypertension. 1995;25:305–313. 18. Epstein FH, Eckoff RD. The epidemiology of high blood pressure—geographic distributions and etiological factors. In: Stamler J, Stamler R, Pullman TN, eds. The Epidemiology of Hypertension. New York: Grune & Stratton; 1967:155–166. 19. The INTERSALT Co-operative Research Group. INTERSALT Study. An international co-operative study on the relation of blood pressure to electrolyte excretion in populations, I: design and methods. Hypertension. 1986;4:781–787. 20. Rodriguez BL, Labarthe DR, Huang B, LopezGomez J. Rise of blood pressure with age: new evidence of population differences. Hypertension. 1994;24:779–785. 21. Labarthe DR. Emergence of risk factors in children. In: Marmot M, Elliott P, eds. Coronary Heart Disease Epidemiology: From Aetiology to Public Health. 2nd ed. Oxford (England): Oxford University Press; 2005:591–605. 22. McDonough JR, Garrison GE, Hames CG. Blood pressure and hypertensive disease among Negroes and Whites in Evans County, Georgia. In: Stamler J, Stamler R, Pullman TN, eds. The Epidemiology of Hypertension. New York: Grune & Stratton; 1967:167–187. 23. Falkner B, Hulman S, Kushner H. Birth weight versus childhood growth as determinants of adult blood pressure. Hypertension. 1998; 31(part 1):145–150. 24. Cruickshank JK, Mzayek F, Liu L, et al. Origins of the “black/white” difference in blood pressure. Roles of birthweight, postnatal growth, early blood pressure, and adolescent body size. The Bogalusa Heart Study. Circulation. 2005;111:1932–1937.
25. Mzayek F, Hassig S, Sherwin R, et al. The association of birth weight with developmental trends in blood pressure from childhood through mid-adulthood. Am J Epidemiol. 2007;166:413–420. 26. Arnett DK, Baird AE, Barkley RA, et al. Relevance of genetics and genomics for prevention and treatment of cardiovascular disease: a scientific statement from the American Heart Association Council on Epidemiology and Prevention, the Stroke Council, and the Functional Genomics and Translational Biology Interdisciplinary Working Group. Circulation. 2007;115:2878–2901. 27. Williams RR, Hunt SC, Hasstedt SJ, et al. Hypertension: genetics and nutrition. In: Simopoulos AP, Childs B, eds. Genetic Variation and Nutrition. World Rev Nutr Diet. 1990;63:116–130. 28. MacGregor GA, de Wardener HE. Salt, Diet & Health. Cambridge, UK: Cambridge University Press; 1998. 29. Committee on Diet and Health, Food and Nutrition Board, Commission on Life Sciences, National Research Council. Diet and Health: Implications for Reducing Chronic Disease Risk. Washington, DC: National Academy Press; 1989. 30. Stamler J. Dietary salt and blood pressure. Ann NY Acad Sci. 1993;676:122–156. 31. Chobanian AV, Hill M. National Heart, Lung, and Blood Institute Workshop on Sodium and Blood Pressure. A critical review of current scientific evidence. Hypertension. 2000;35: 858–863. 32. Karppanen H, Mervaala E. Sodium intake and hypertension. Prog Cardiovasc Dis. 2006;49: 59–75. 33. World Health Organization. Reducing Salt Intake in Populations. Report of a WHO Forum and Technical Meeting 5–7 October 2006, Paris, France. Geneva (Switzerland): World Health Organization, 2007. http:// www.who.int/dietphysicalactivity/reducingsalt/ en/index.html. Accessed April 4, 2007.
46894_CH12_311_360.qxd
354
2/25/10
3:40 PM
Page 354
CHAPTER 12 HIGH BLOOD PRESSURE
34. Blackburn H, Prineas R. Diet and hypertension: anthropology, epidemiology, and public health implications. Prog Biochem Pharmacol. 1983;19:31–79. 35. Godlee F. The food industry fights for salt. Br Med J. 1996;312:1239–1240. 36. Dietary Guidelines for Americans. US Department of Health and Human Services (HHS) and the US Department of Agriculture (USDA); 2005. http://www.health.gov/dietary guidelines/dga2005/document/pdf/DGA2005 .pdf. Accessed December 7, 2006. 37. Mattes RD, Donnelly D. Relative contributions of dietary sodium sources. J Am Coll Nutr. 1991;10:383–393. 38. Appel LJ, Elliott P. Macronutrients, fiber, cholesterol, and dietary patterns. In: Whelton PK, He J, Louis GT, eds. Lifestyle Modification for the Prevention and Treatment of Hypertension. New York, NY: Marcel Dekker, Inc.; 2003. 39. Stamler J, Elliott P, Dennis B, et al. INTERMAP: background, aims, design, methods, and descriptive statistics (nondietary). J Hum Hypertension. 2003;17:591–608. 40. Elliott P, Stamler J, Dyer AR, et al. Association between protein intake and blood pressure. Arch Int Med. 2006;166:79–87. 41. He J, Klag MJ, Whelton PK, et al. Migration, blood pressure pattern, and hypertension: the Yi Migrant Study. Am J Epidemiol. 1991;134: 1085–1101. 42. Poulter N, Khaw KT, Hopwood BEC, et al. Blood pressure and its correlates in an African tribe in urban and rural environments. J Epidem Commun Health. 1984;38:181–186. 43. Poulter N, Khaw KT, Hopwood BEC, et al. The Kenyan Luo migration study: observations on the initiation of a rise in blood pressure. BMJ. 1990;300:967–972. 44. Keiser HR. Catecholamines and target organ damage. In: Izzo JL, Jr, Black HR, eds. Hypertension Primer. Dallas, TX: American Heart Association; 1993:123–124.
45. Izzo JL Jr, Black HR, eds. Hypertension Primer. Dallas, TX: American Heart Association; 1993. 46. Sullivan JM. Coronary artery disease: pathophysiology. In: Izzo JL, Jr, Black HR, eds. Hypertension Primer. Dallas, TX: American Heart Association; 1993:133–134. 47. US Department of Health and Human Services. Health, United States, 2008 with Special Feature on the Health of Young Adults. Washington DC: US Department of Health and Human Services. Centers for Disease Control and Prevention. National Center for Health Statistics; 2008. 48. Centers for Disease Control and Prevention. Racial/ethnic disparities in prevalence, treatment, and control of hypertension––United States, 1999–2002. MMWR. 2005;54:7–9. 49. Wang Y, Wang QJ. The prevalence of prehypertension and hypertension among US adults according to the new Joint National Committee guidelines. Arch Int Med. 2004;164: 2126–2134. 50. Lloyd-Jones DM, Evans JC, Levy D. Hypertension in adults across the age spectrum. Current outcomes and control in the community. JAMA. 2005;294:466–472. 51. Izzo JL, Levy D, Black HR. Importance of systolic blood pressure in older Americans. Hypertension. 2000;35:1021–1024. 52. Paradis G, Lambert M, O’Loughlin J, et al. Blood pressure and adiposity in children and adolescents. Circulation. 2004;110: 1832–1838. 53. Hansen ML, Gunn PW, Kaelber DC. Underdiagnosis of hypertension in children and adolescents. JAMA. 2007;298:874–879. 54. Goff Jr DC, Howard G, Russell GB, Labarthe DR. Birth cohort evidence of population influences on blood pressure in the United States, 1887–1994. Ann Epidemiol. 2001;11: 271–279. 55. US Department of Health and Human Services. Healthy People 2010 Midcourse Review:
46894_CH12_311_360.qxd
2/25/10
3:40 PM
Page 355
REFERENCES 355
Washington, DC: US Government Printing Office; December 2006. 56. Cutler JA, Sorlie PD, Wolz M, Thom T, Fields LE, Roccella EJ. Trends in hypertension prevalence, awareness, treatment, and control rates in United States adults between 1988–1994 and 1999–2004. Hypertension. 2008;52; 818–827. 57. Ong KL, Cheung BMY, Man YB, Lau CP, Lam KSL. Prevalence, awareness, treatment and control of hypertension among United States Adults 1999–2004. Hypertension. 2007;49: 69–75. 58. Ostchega Y, Dillon CF, Hughes JP, Carroll M, Yoon S. Trends in hypertension prevalence, awareness, treatment, and control in older U.S. adults: data from the National Health and Nutrition Examination Survey 1988 to 2004. J Am Geriatr Soc. 2007;55:1056–1065. 59. Muntner P, He J, Cutler JA, Wildman RP, Whelton PK. Trends in blood pressure among children and adolescents. JAMA. 2004;291: 2107–2113. 60. Ostchega Y, Carroll M, Prineas RJ, McDowell MA, Louis T, Tilert T. Trends of elevated blood pressure among children and adolescents: data from the National Health and Nutrition Examination Survey 1988–2006. Am J Hypertens. 2009;22:59–67. 61. Hajjar I, Kotchen JM, Kotchen TA. Hypertension: trends in prevalence, incidence, and control. Annu Rev Public Health. 2006;27: 465–490. 62. The WHO MONICA Project. Geographical variation in the major risk factors of coronary heart disease in men and women aged 35–64 years. World Health Stat Q. 1988;41:115–140. 63. Erdine S, Aran SN. Current status of hypertension control around the world. Clin Exper Hypert. 2004;26:731–738. 64. Wolf-Maier K, Cooper RS, Banegas JR, et al. Hypertension prevalence and blood pressure levels in 6 European countries, Canada, and the United States. JAMA. 2003;289:2363–2369.
65. Whelton PK, He J, Muntner P. Prevalence, awareness, treatment and control of hypertension in North America, North Africa and Asia. J Human Hypert. 2004;18:545–551. 66. Leenen FHH, Dumais J, McInnis NH, et al. Results of the Ontario Survey on the Prevalence and Control of Hypertension. Can Med Assoc J. 2008;178:1441–1449. 67. Kearney PM, Whelton M, Reynolds K, Muntner PK, He J. Global burden of hypertension: analysis of worldwide data. Lancet. 2005;365:217–223. 68. Tunstall-Pedoe H, Connaghan J, Woodward M, Tolonen H, Kuulasmaa K. Pattern of declining blood pressure across replicate population surveys of the WHO MONICA project, mid-1980s to mid-1990s, and the role of medication. BMJ. 2006;332:629–635. doi: 10.1136/bmj.38753.779005.BE. 69. Antikainen RL, Moltchanov VA, Chukwuma Sr C, et al., for the WHO MONICA Project. Trends in the prevalence, awareness, treatment and control of hypertension: the WHO MONICA Project. Eur J Cardiovasc Prev Rehabil. 2006;13:13–29. 70. Levy RL, White PD, Stroud WD, Hillman CC. Sustained hypertension: predisposing factors and causes of disability and death. JAMA. 1947;135:77–80. 71. Ford ES, Cooper RS. Risk factors for hypertension in a national cohort study. Hypertension. 1991;18:598–606. 72. Vargas C, Ingram DD, Gillum RF. Incidence of hypertension and educational attainment. The NHANES I Followup Study. Am J Epidemiol. 2000;152:272–278. 73. Gillum RF, Mussolino ME, Madans JH. Relation between region of residence in the United States and hypertension incidence––the NHANES I Epidemiologic Follow-up Study. J Natl Med Assoc. 2004;96:625–634. 74. Tu K, Chen Z, Lipscombe LL for the Canadian Hypertension Education Program Outcomes Research Task Force. Prevalence and incidence
46894_CH12_311_360.qxd
356
2/25/10
3:40 PM
Page 356
CHAPTER 12 HIGH BLOOD PRESSURE
of hypertension from 1995 to 2005: a population-based study. Can Med Assoc J. 2008;178: 1429–1435. 75. Ezzati M, Oza S, Danaei G, Murray CJL. Trends and cardiovascular mortality effects of state-level blood pressure and uncontrolled hypertension in the United States. Circulation. 2008;117:905–914. 76. Mazzati E, Vander Hoorn S, Lopez AD, et al. Comparative quantification of mortality and burden of disease attributable to selected risk factors. In: Lopez, AD, et al., eds. Global Burden of Disease and Risk Factors. Washington, DC: The International Bank for Reconstruction and Development/The World Bank; 2006:241–396. 77. Keys A. Seven Countries: A Multivariate Analysis of Death and Coronary Heart Disease. Cambridge, MA: Harvard University Press; 1980. 78. Society of Actuaries and Association of Life Insurance Medical Directors of America. Blood Pressure Study 1979. Chicago: Society of Actuaries and Association of Life Insurance Medical Directors of America; 1980. 79. National Heart and Lung Institute, for the National High Blood Pressure Education Program. The underwriting significance of hypertension for the life insurance industry. DHEW Publication No. (NIH) 75-426. Bethesda, MD: National Heart and Lung Institute; 1974. 80. Prospective Studies Collaboration. Age-specific relevance of usual blood pressure to vascular mortality: a meta-analysis of individual data for one million adults in 61 prospective studies. Lancet. 2002;360:1903–1913.
In: Marmot M, Elliott P, eds. Coronary Heart Disease Epidemiology: From Aetiology to Public Health. 2nd ed. Oxford (England): Oxford University Press; 2005;152–173. 83. Lawes CMM, Vander Hoorn S, Rodgers A for the International Society of Hypertension. Global burden of blood-pressure-related disease, 2001. Lancet. 2008;371:1513–1518. 84. Stamler R. The primary prevention of hypertension and the population blood pressure problem. In Marmot M, Elliott P, eds. Coronary Heart Disease Epidemiology: From Aetiology to Public Health. Oxford: Oxford University Press; 1992:415–434. 85. World Health Organization. Primary prevention of hypertension. Report of a WHO Scientific Group. Technical Report Series 686. Geneva: World Health Organization; 1983. 86. National High Blood Pressure Education Program Working Group report on primary prevention of hypertension. Arch Int Med. 1993;153:186–208. 87. Whelton PK, He J, Appel LJ. Primary prevention of hypertension. Clinical and public health advisory from the National High Blood Pressure Education Program. JAMA. 2002; 288: 1882–1888. 88. World Health Organization. Blood pressure studies in children. Report of a WHO Study Group. Technical Report Series 715. Geneva: World Health Organization; 1985. 89. World Health Organization. Prevention in childhood and youth of adult cardiovascular diseases: time for action. Report of a WHO Expert Committee. Technical Report Series 792. Geneva: World Health Organization; 1990.
81. Danaei G, Ding EL, Mozaffarian D, et al. The preventable causes of death in the United States: comparative risk assessment of dietary, lifestyle, and metabolic factors. PLoS Med. 2009;6(4): e1000058. doi:10.1371/journal .pmed.1000058.
90. Labarthe DR, Dai S, Day S, Fulton JE, Grunbaum JA for the Project HeartBeat! Writing Group. Findings from Project HeartBeat! Their importance for CVD prevention. Am J Prev Med. 2009;37(suppl 1):S105–S115.
82. Lawes CMM, Vander Hoorn S, Law MR, Elliott P, MacMahon S, Rodgers A. Blood pressure and the burden of coronary heart disease.
91. US Preventive Services Task Force. The Guide to Clinical Preventive Services 2006. Recommendations of the U.S. Preventive
46894_CH12_311_360.qxd
2/25/10
3:40 PM
Page 357
REFERENCES 357
Services Task Force. Washington, DC: Agency for Healthcare Research and Quality. 2006. www.ahrq.gov/clinic/uspstf/uspstbac.htm. Accessed October 14, 2007. 92. Whelton PK, He J, Louis GT, eds. Lifestyle Modification for the Prevention and Treatment of Hypertension. New York: Marcel Dekker, Inc.; 2003. 93. Kumanyika S, Iqbal N. Weight reduction, In Whelton PK, He J, Louis GT, eds. Lifestyle Modification for the Prevention and Treatment of Hypertension. New York: Marcel Dekker, Inc.; 2003:107–138. 94. Cutler JA, Obarzanek E, Roccella EJ. Dietary salt reduction. In Whelton PK, He J, Louis GT, eds. Lifestyle Modification for the Prevention and Treatment of Hypertension. New York: Marcel Dekker, Inc.; 2003:139–160. 95. Puddey IB, Cushman WC. Moderation of alcohol consumption. In Whelton PK, He J, Louis GT, eds. Lifestyle Modification for the Prevention and Treatment of Hypertension. New York: Marcel Dekker, Inc.; 2003: 161–190. 96. Ishikawa-Takata K, Ohta T. Physical activity. In Whelton PK, He J, Louis GT, eds. Lifestyle Modification for the Prevention and Treatment of Hypertension. New York: Marcel Dekker, Inc.; 2003:191–212. 97. Sacks FM, Svetkey LP, Vollmer WM, et al. Effects on blood pressure of reduced dietary sodium and the Dietary Approaches to Stop Hypertension (DASH) Diet. N Engl J Med. 2001;344:3–10. 98. Kumanyika SK, Hebert PR, Cutler JA, Lasser VI, et al. Trials of Hypertension Prevention Collaborative Research Group: feasibility and efficacy of sodium reduction in the trials of hypertension prevention, phase I. Hypertension. 1993;22:502–511. 99. Cook NR, Cutler JA, Obarzanek E, et al. Long term effects of dietary sodium reduction on cardiovascular disease outcomes: observational follow-up of the trials of hypertension prevention (TOHP). BMJ. 2009;334(7599):885–888. doi:10.1136/bmj.38147.604896.55.
100. Cutler JA, Follman D, Allender PS. Randomized trials of sodium reduction: an overview. Am J Clin Nutr. 1997;65(suppl): 643S–651S. 101. Midgley JP, Matthew AG, Greenwood CMT, Logan AG. Effect of reduced dietary sodium on blood pressure: a meta-analysis of randomized controlled trials. JAMA. 1996;275: 1590–1597. 102. Staessen JA, Lijnen P, Thijs L, Fagard R. Salt and blood pressure in community-based intervention trials. Am J Clin Nutr. 1997; 65(suppl):661S–670S. 103. He J, MacGregor GA. Importance of salt in determining blood pressure in children. Metaanalysis of controlled trials. Hypertension. 2006;48:861–869. 104. Dickinson HO, Mason JM, Nicolson DJ, et al. Lifestyle interventions to reduce raised blood pressure: a systematic review of randomized controlled trials. J Hypertens. 2006;24: 215–233. 105. Centers for Disease Control and Prevention. Prevalence of actions to control high blood pressure—20 states, 2005. MMWR. 2007; 56:420–423. 106. Freis ED. Long-term treatment. Organization of a long-term multiclinic therapeutic trial in hypertension. In Gross F, ed. Antihypertensive Therapy: Principles and Practice, an International Symposium. New York: SpringerVerlag, New York, Inc.; 1966:345–354. 107. Freis ED. Reminiscences of the Veterans Administration trials of the treatment of hypertension. Hypertension. 1990;16:472–475. 108. Hypertension Detection and Follow-up Program Cooperative Group. Five-year findings of the Hypertension Detection and Follow-up Program. I. Reduction of mortality of persons with high blood pressure, including mild hypertension. JAMA. 1979;242: 2562–2571. 109. Hypertension Detection and Follow-up Program Cooperative Group. Five-year findings of the Hypertension Detection and
46894_CH12_311_360.qxd
358
2/25/10
3:40 PM
Page 358
CHAPTER 12 HIGH BLOOD PRESSURE
Follow-up Program. II. Mortality by race-sex and age. JAMA. 1979;242:2572–2577. 110. SHEP Cooperative Research Group. Prevention of stroke by antihypertensive drug treatment in older persons with isolated systolic hypertension. Final results of the Systolic Hypertension in the Elderly Program (SHEP). JAMA. 1991;265:3255–3264. 111. Cushman WC, Ford CE, Cutler JA, et al. for the ALLHAT Collaborative Research Group. Success and predictors of blood pressure control in diverse North American settings: the Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT). J Clin Hypertens. 2002;4:393–404. 112. He J, Muntner P, Chen J, Roccella EJ, Streiffer RH, Whelton PK. Factors associated with hypertension control in the general population of the United States. Arch Int Med. 2002;162: 1051–1058. 113. Vawter L, Tong X, Gemilyan M, Yoon PW. Barriers to antihypertensive medication adherence among adults––United States, 2005. J Clin Hypertens. 2008;10:922–929. 114. Fahey T, Schroeder K, Ebrahim S. Interventions used to control blood pressure in patients with hypertension. Cochrane Database Syst Rev. 2006 Oct 18;(4):CD005182. doi: 10.1002/14651858.CD005182.pub3. 115. Berenson GS, Shear CL, Chiang YK, et al. Combined low-dose medication and primary intervention over a 30-month period for sustained high blood pressure in childhood. Am J Med Sci. 1990;299:79–86. 116. Roccella EJ, Horan MJ. The National High Blood Pressure Education Program: measuring progress and assessing its impact. Health Psychol. 1988;7(suppl):297–303. 117. World Health Organization. Hypertension Control: Report of a WHO Expert Committee. Technical Report Series 862. Geneva (Switzerland): World Health Organization; 1996. 118. Lamar Welch VL, Hill MN. Effective strategies for blood pressure control. Cardiol Clin. 2002;20:321–333.
119. Forte JG, Pereira Miguel JM, Pereira Miguel MJ, de Pádua F, et al. Salt and blood pressure: a community trial. J Hum Hypertens. 1989;3: 179–184. 120. Rein DB, Constantine RT, Orenstein D, et al. A cost evaluation of the Georgia Stroke and Heart Attack Prevention Program. Prev Chronic Dis. 2006;3(1):A12. http://www .cdc.gov/pcd/issues/2006/jan/05_0143.htm. 121. Constantine R, Brownstein JN, Hoover S, et al. Strategies for controlling blood pressure among low-income populations in Georgia. Prev Chron Dis. 2008;5(2). http://www .cdc.gov/pcd/issues/2008/apr/07_0200.htm. Accessed May 1, 2008. 122. Bunting BA, Smith BH, Sutherland SE. The Asheville Project: clinical and economic outcomes of a community-based long-term medication therapy management program for hypertension and hyperlipidemia. J Am Pharm Assoc. 2008;48:23–31. 123. McClellan W, Wilber JA. A decade’s experience with hypertension control programs in the United States: the empirical basis for a model of community control programs. In: Rosenfeld JB, Silverberg DS, Viskiper R, eds. Hypertension Control in the Community. London: John Libbey; 1985:1–16. 124. World Health Organization. Prevention of Cardiovascular Disease. Guidelines for assessment and management of cardiovascular risk. Geneva: World Health Organization; 2007. 125. Rodgers A, Lawes C, MacMahon S. Reducing the global burden of blood pressure-related cardiovascular disease. Hypertension. 2000; 18(suppl 1):S3–S6. 126. Murray CLJ, Lauer JA, Hutubessy RCW, et al. Effectiveness and costs of interventions to lower systolic blood pressure and cholesterol: a global and regional analysis on reduction of cardiovascular disease risk. Lancet. 2003; 361:717–725. 127. Asaria P, Chisholm D, Mathers C, Ezzati M, Beaglehole R. Chronic disease prevention: health effects and financial costs of strategies to reduce salt intake and control tobacco abuse. Lancet. 2007;370:2044–2053.
46894_CH12_311_360.qxd
2/25/10
3:40 PM
Page 359
REFERENCES 359
128. International Society of Hypertension. http://www.ish-world.com. Accessed May 12, 2009.
130. Remington RD. High blood pressure control: what are the next steps? Pub Health Rep. 1980;95:456–461.
129. World Action on Salt and Health. Newsletter Issue No. 4, March 2009. http://www .worldactiononsalt.com. Accessed April 1, 2009.
131. Centers for Disease Control and Prevention. Application of lower sodium intake recommendations for adults—United States, 1999–2006. MMWR. 2009;58:281–283.
46894_CH12_311_360.qxd
2/25/10
3:40 PM
Page 360
46894_CH13_361_394.qxd
2/25/10
3:41 PM
Page 361
C H A P T E R
13 Diabetes and the Metabolic Syndrome mension of public health policy and clinical practice in diabetes, and the targets for intervention—weight reduction through improved nutrition and increased physical activity—are common to cardiovascular disease prevention more generally. Further, criteria defining IFG and IGT have been aligned so as to identify a consistent level of risk for progression to T2DM. The precursor range of blood glucose levels is now identified as “prediabetes” and calls attention to this broad stratum of the population as representing a new focus for prevention. Data from the National Health and Nutrition Examination Surveys (NHANES), 2003–2006, indicate that more than 23 million US adults, or 7.8%, have diabetes of one or the other type. An additional 57 million are estimated to have prediabetes. The global estimate is 330 million people with diabetes, predominantly type 2. Concurrent with the increasing prevalence of obesity in childhood and adolescence (as well as in adulthood), T2DM—once termed “adultonset diabetes”—has become recognized as occurring before adulthood, although estimates of its frequency remain uncertain. The cardiovascular disease risks associated with T2DM have led to the concept of diabetes as a “CHD equivalent” for purposes of risk assessment and management to prevent future cardiovascular events. This means that presence of diabetes warrants equally aggressive risk-reduction approaches as for persons with recognized coronary heart disease at first evaluation. Current issues in epidemiology and prevention of diabetes include further understanding of its development beginning in childhood and adolescence (or earlier), reconsidering policies against population screening in view of recognition of prediabetes and primary prevention, resolving different perspectives
SUMMARY A spectrum of conditions, from elevated blood glucose concentration to severe disturbance of insulinglucose regulation, is denoted by such terms as hyperglycemia, impaired fasting glucose (IFG), impaired glucose tolerance (IGT), insulin resistance (IR), and insulin-dependent diabetes mellitus (IDDM) or non-insulin-dependent diabetes mellitus (NIDDM). The latter form is now commonly termed “T2DM mellitus” (T2DM) and is the predominant form in its contribution to the global burden of chronic diseases. Together, these terms refer to disorders that are closely related to risks of atherosclerotic and hypertensive cardiovascular diseases. Elevated glucose or insulin are associated with these cardiovascular conditions and cluster with other cardiovascular risk factors such as adverse blood lipid profiles, high blood pressure, and obesity. Such clustering has come to be viewed as constituting a syndrome, the “metabolic syndrome,” addressed here because of its close intersection with glucose, insulin, and T2DM. Advances in the epidemiology of diabetes and related disorders followed from adoption of standard criteria for diagnosis and classification in population-based and clinical studies. Major impetus to international collaboration in the study of diabetes has come from the World Health Organization (WHO) and the International Diabetes Federation (IDF). Comprehensive reviews of the epidemiologic and related literature have made the field more accessible. A major development in understanding of T2DM and its prevention follows from several clinical trials demonstrating that lifestyle intervention is effective in preventing progression from IGT to T2DM. Thus, primary prevention of T2DM has become a new di-
361
46894_CH13_361_394.qxd
362
2/25/10
3:41 PM
Page 362
CHAPTER 13 DIABETES AND THE METABOLIC SYNDROME
on the clustering of risk factors through a more holistic approach, and adopting as strategies of diabetes prevention the well-established policies for improving nutrition, increasing physical activity, achieving and maintaining healthy weight, and promoting prevention or cessation of tobacco use.
INTRODUCTION Diabetes mellitus—often referred to simply as diabetes—may be characterized as a disorder of glucose transport or metabolism due to reduced production or effectiveness of insulin, a hormone produced by specialized cells in the pancreas and having multiple regulatory functions. In its classic form, marked by passage of large quantities of sugarcontaining urine (this distinguishes diabetes mellitus from another condition, diabetes insipidus), diabetes mellitus is said to have been recognized for more than 2000 years. Diabetes mellitus occurs mainly in two distinct forms, whose typical onset is either early in life or in middle to later adulthood. Vascular manifestations occur in both forms. In adult-onset diabetes, it is these vascular complications that constitute the leading cause of death. Because the presence of diabetes increases the risk of coronary heart disease and other consequences of atherosclerosis, emphasis has increased on prevention of these cardiovascular conditions among people with diabetes. Elevated blood glucose concentration is a defining characteristic of diabetes. Blood glucose concentration is continuously distributed in the population, without a distinct threshold to distinguish diabetic from nondiabetic individuals. However, critical values of blood glucose have been adopted to define diabetes, analogous to those for blood lipids or blood pressure. The importance of blood glucose levels below the diagnostic criterion, but above optimum levels, have become recognized and categorized as “prediabetes.” Disturbances of insulin production or function and attendant impairment of glucose utilization, including “insulin resistance,” have also been recognized. These disorders are also related to increased cardiovascular risk. Public health concern therefore includes a wide spectrum of disturbances of the glucose-insulin regulatory system and related conditions. In Epidemiology of Diabetes and Its Vascular Lesions, published in 1978, West presented an extensive historical review of work in this field. He dated population research on diabetes from the earliest collection of death certificate information in Europe and the United States in 1850.1 Other early epidemiologic observations included geographic dif-
ferences and secular changes in the prevalence of diabetes that were reported well before the 20th century. He noted at the same time, however, that until the 1960s, very little systematic epidemiologic investigation of diabetes had been accomplished. This fact was attributed to lack of epidemiologic training and experience on the part of those studying diabetes in the earlier years. Adoption of methods and criteria for population studies accounts for a dramatic change in this field over the most recent three to four decades. Epidemiologic studies have contributed importantly to understanding the relation of insulin resistance to both diabetes and atherosclerosis. Obesity, most specifically visceral adiposity, and physical inactivity are found to be prominent factors common to both disorders. Advances in this field have added support to the “common soil hypothesis,” which proposes a single set of antecedent conditions underlying these disorders.2 Links of multiple factors with both diabetes and atherosclerosis were suggested in the schematic representation in Figure 10-7. One outgrowth of this concept was establishment in 2004 of the International Society of Diabetes and Vascular Disease to foster collaboration between diabetologists and cardiologists. Research on the prognostic importance of diabetes for coronary heart disease has also had an important implication for clinical practice. It has been recognized that control of cardiovascular risk factors among people with diabetes is essential for prevention of major cardiovascular complications. Accordingly, guidelines of the Adult Treatment Panel III (ATP III) of the National Cholesterol Education Program included presence of diabetes as equivalent to existing coronary heart disease (“CHD equivalent”) in determining the recommended intensity of lipid-lowering therapy for such patients.3 Also addressed in ATP III is the “metabolic syndrome,” a term that is variously defined but generally refers to presence of multiple cardiovascular risk factors including diabetes or elevated blood glucose or insulin concentration. The “syndrome” is not identical with diabetes, and the validity of the concept is challenged by some, tending to divide the diabetes and cardiovascular communities. 4 However, the concept has generated considerable interest and is appropriately discussed in conjunction with diabetes. Other developments adding to the prominence of diabetes in the context of cardiovascular disease prevention include evidence that rising prevalence of diabetes in the United States has partially offset the effect of risk factor improvements in prevention of coronary heart disease;5 recognition that the formerly
46894_CH13_361_394.qxd
2/25/10
3:41 PM
Page 363
CONCEPTS AND DEFINITIONS 363
adult-onset form of diabetes is occurring with increasing frequency in childhood and adolescence and is closely linked with the epidemic of childhood (and adult) obesity;6 strong evidence that diabetes can be prevented by the same lifestyle interventions to achieve improvement in diet and physical activity, with weight loss, as are promoted for prevention and control of cardiovascular risk factors generally;7 and the substantial and projected increasing prevalence of diabetes in developing countries, adding significantly to its global health burden.8
CONCEPTS AND DEFINITIONS Terminology The International Textbook of Diabetes Mellitus presents an extensively annotated “Classification of Diabetes Mellitus and Other Categories of Glucose Intolerance.”9 This classification represents the deliberations of an international working group convened in the United States, a WHO Expert Committee, several national diabetes societies, and other organizations. It reflects the view that diabetes mellitus is a heterogeneous group of disorders both in causation and in clinical manifestations, with elevated blood glucose concentration as the major unifying characteristic. Non-insulin-dependent diabetes mellitus (NIDDM, type 2, T2DM) is distinguished from insulin-dependent diabetes mellitus (IDDM, type 1); impaired glucose tolerance (IGT) is identified as a separate class. It is useful to note some of the terms used in the evolving literature of this field. Several terms have been used synonymously with NIDDM (e.g., T2DM, adult-onset diabetes) and IDDM (e.g., type 1 diabetes, juvenile-onset diabetes). Because T2DM is far more prevalent than IDDM, it contributes more to the relation between diabetes and cardiovascular diseases; reference hereafter to “diabetes” is therefore to T2DM unless otherwise noted. IGT and impaired fasting glycemia (IFG) reflect distinct aspects of glucose metabolism and have different epidemiologic patterns and prognostic relations. IFG is more clearly related to development of diabetes and IGT to cardiovascular risk. Details of their definition were reviewed by a workshop of the IDF and published in 2002.10 The term “prediabetes” is now used to refer to the intermediate categories of blood glucose levels determined by either fasting plasma glucose (FPG) or oral glucose tolerance test (OGTT), in the ranges from 100–126 mg/dl or 140–200 mg/dl, respectively.
Natural History The natural history of T2DM is depicted in Figure 13-1, from the 1994 WHO Study Group report, Prevention of Diabetes Mellitus.11 It indicates the broad time course, usually over many years, of progression from onset to development of complications, disability, and death. Prior to the onset of clinical signs, genetic susceptibility and environmental factors, especially unfavorable nutrition, obesity, and physical inactivity, result in progressive IGT. This is detectable by measures of insulin resistance (impairment of glucose uptake in tissues despite insulin loading), hyperinsulinemia (increased blood concentration of insulin), and decrease in concentration of highdensity lipoprotein (HDL) cholesterol. Both hyperglycemia and hypertension are characteristic of the clinical phase of T2DM, with eventual development of complications that affect the eyes, kidneys, and peripheral nerves and promote atherosclerosis. Progression of these conditions may result in disability and death. Type 1 diabetes mellitus differs from this course, especially by the absence of obesity and physical inactivity, presence of certain immunologic markers as risk factors in the preonset phase, absence of an IGT phase, and postonset dependency on exogenous insulin. Hypertension is not typical of type 1. The later phases are otherwise similar to T2DM in clinical features, but their development and progression are accelerated in type 1. International Statistical Classification Another approach to classification is that of the International Statistical Classification of Diseases and Related Health Problems, Tenth Revision, which specifies five main categories for diabetes: IDDM is coded as E10, NIDDM as E11, malnutrition-related diabetes mellitus as E12, and other specified or unspecified diabetes mellitus as E13 or E14, respectively.12 Subcategories for each code identify the absence, or the presence and type, of complications. The Metabolic Syndrome Insulin resistance is one dimension of the insulin– glucose relationship that has received greatly increased attention. It was recognized decades ago that hyperglycemia could occur in the presence of normal or even increased blood concentrations of insulin, a circumstance considered to be due to some aberration of the insulin molecule itself or its site of action. Therefore, diabetes is not necessarily a consequence of failure of the pancreas to secrete insulin. This understanding formed part of the basis for distinguishing
46894_CH13_361_394.qxd
364
2/25/10
3:41 PM
Page 364
CHAPTER 13 DIABETES AND THE METABOLIC SYNDROME
Onset of diabetes Environmental factors
Genetic susceptibility
nutrition obesity physical inactivity
Disability
Death
eg,
Complications
IGT
Insulin resistance Hyperinsulinemia HDL
Hyperglycemia Hypertension
Retinopathy Nephropathy Atherosclerosis Neuropathy
Blindness Renal failure Coronary heart disease Amputation
Note: HDL, high-density lipoprotein; IGT, impaired glucose tolerance.
Figure 13-1 The Natural History of Non-Insulin-Dependent Diabetes Mellitus. Source: Reprinted with permission from Report of a WHO Study Group, WHO TRS 844, p 25, © 1994, World Health Organization.
both T2DM and IGT from IDDM. However, the broader concept of insulin resistance, or decreased insulin sensitivity, as constituting part of a syndrome of physiologic and metabolic disorders is a more recent development. What has been variously termed the metabolic syndrome or the insulin resistance syndrome was originally dubbed “syndrome X” by Reaven in 1988.13 Its components, according to Reaven, were resistance to insulin-stimulated glucose uptake, glucose intolerance, hyperinsulinemia, increased very-low-density lipoprotein (VLDL) triglyceride, and decreased HDLcholesterol concentrations, and hypertension. By 1999, a WHO consultation proposed a definition of metabolic syndrome represented in Figure 13-2. T2DM, IGT or IGF, or insulin resistance must be present, with at least two of four other conditions— hypertension, obesity (as measured by either body mass index or waist to hip ratio), raised triglycerides or low HDL cholesterol, or microalbuminuria.14 ATP III, by contrast, identified six factors as characteristic of the metabolic syndrome: abdominal obesity, atherogenic dyslipidemia, raised blood pressure, insulin resistance with or without glucose intolerance, prothrombotic state, and proinflammatory
state.3 It was noted that there were “no well-accepted criteria for the diagnosis of the metabolic syndrome.”3, p II-27 However, for clinical identification of the syndrome, five factors were to be measured: abdominal obesity, triglycerides, HDL-cholesterol, blood pressure, and fasting glucose. A more recent conceptualization by Grundy emphasizes abdominal obesity as the main underlying condition, with the metabolic syndrome represented as including borderline and elevated cardiovascular risk factors together with T2DM (Figure 13-3).4 Risks of cardiovascular disease and its complications are intensified by higher levels of risk factors and by presence of diabetes, which also leads to specific complications of diabetes itself. Grundy attributed to diabetologists the view that insulin resistance is the fundamental disorder, progressing to prediabetes and T2DM, while also causing metabolic syndrome (MetS) (Figure 13-4).4 From this perspective, the term “insulin resistance syndrome” would be more appropriate, and the cardiologist’s preference for “metabolic syndrome” is problematic. Others have observed that insulin resistance or hyperinsulinemia may not be present in persons with the syndrome; that these conditions may
46894_CH13_361_394.qxd
2/25/10
3:41 PM
Page 365
CONCEPTS AND DEFINITIONS 365
Hypertension: 140/90 mmHg Obesity: BMI 30 kg/m2, or waist to hip ratio > 0.90 for males and > 0.85 for females
Type 2 diabetes Impaired glucose tolerance or impaired fasting glucose
Raised triglycerides ( 1.7 mmol/l) and/or low HDL (< 0.9 mmol/l for men; < 1.0 mmol/l for women)
Insulin resistance
Microalbuminuria: urinary albumin excretion rate 20 g/min or albumin creatinine ratio 30 mg/g
At least 1 of
At least 2 of
Metabolic syndrome
Figure 13-2 Metabolic Syndrome as Defined by the WHO (WHO Consultation 1999). Source: Reprinted with permission from Coronary Heart Disease Epidemiology from Aetiology to Public Health, G Hu, Q Qiao, J Tuomilehto, p 312. © 2005 Oxford University Press.
simply be additional manifestations of another, yet unidentified, underlying cause; and that the inclusion of some weakly associated factors and exclusion of others more strongly related to insulin resistance further cloud the picture.15 In addition to the fundamental issue of definition, the meaning of this syndrome for risks of atherosclerotic and hypertensive diseases is debated as to whether it represents simply the aggregation of risk factors that often coexist or signifies a single underlying pathophysiologic entity as a potential target for prevention and treatment. In contrast, a still
Abdominal Obesity
broader interpretation of this complex set of interrelated conditions linked with insulin resistance has been suggested by Keen, that it results from “some much more general disturbance of adaptation to the conditions of modern life.”16, p xxviii The question of meaning and utility of the concept of the metabolic syndrome in children and adolescents was examined in an American Heart Association (AHA) Scientific Statement, which noted the lack of a universally accepted definition.17 Varying criteria limited comparability of the 15 identified studies in this age group.
Metabolic Syndrome Multiple Borderline Risk Factors
Cardiovascular Disease & Complications
Multiple Categorical Risk Factors Type 2 Diabetes
Diabetic Complications
Figure 13-3 Progression and Outcomes of the Metabolic Syndrome. The metabolic syndrome arises largely out of abdominal obesity. With aging and increasing obesity, metabolic risk factors worsen. Many persons with the metabolic syndrome eventually develop type 2 diabetes. As the syndrome advances, risk for cardiovascular disease and its complications increase. Once diabetes develops, diabetic complications other than cardiovascular disease often develop. The metabolic syndrome encompasses each stage in the development of risk factors and type 2 diabetes. Source: Reprinted with permission from the Journal of the American College of Cardiology, Vol 47, No 6, SM Grundy, p 1094. © 2006 by the American College of Cardiology Foundation.
46894_CH13_361_394.qxd
366
2/25/10
3:41 PM
Page 366
CHAPTER 13 DIABETES AND THE METABOLIC SYNDROME
Insulin Resistance Metabolic Syndrome (MetS)
PreDiabetes (75% MetS) Type 2 Diabetes (86% MetS)
Cardiovascular Disease
Figure 13-4 Interrelations and Overlap of Metabolic Syndrome with Insulin Resistance, Prediabetes, and Type 2 Diabetes. According to the insulin resistance hypothesis, the metabolic syndrome is caused predominantly by insulin resistance. The latter also contributes to prediabetes and, ultimately, to type 2 diabetes. About 75% of people with prediabetes and 86% of people with type 2 diabetes have the metabolic syndrome. Both metabolic syndrome and type 2 diabetes are known to predict cardiovascular disease. Source: Reprinted with permission from the Journal of the American College of Cardiology, Vol 47, No 6, SM Grundy, p 1095. © 2006 by the American College of Cardiology Foundation.
MEASUREMENT Diagnostic Criteria The unifying element of all diagnostic categories in diabetes mellitus is elevated blood glucose concentration, or hyperglycemia. Measurement of blood glucose concentration and the criteria for classification on this basis are the foundation of population studies as well as comparability of case identification in clinical research. Because blood glucose concentration varies in Table 13-1
relation to the timing of food intake, values based on casual samples, uncontrolled for food intake, are of little use. Reliable screening or diagnostic testing requires standardization, either through assurance of fasting for a fixed minimum period or through feeding a known quantity of glucose. In the latter approach, blood samples can be obtained in the fasting state and at one or more fixed intervals, such as 2 hours, after ingestion of a known glucose load. This approach is indicated in the diagnostic values for the oral glucose tolerance test as recommended by the WHO, shown in Table 13-1.18 Values in mmol/L are given for classification of diabetes mellitus, IGT, or IFG in either fasting or postload status and whether obtained as venous or capillary (finger stick) samples of whole blood or plasma. Even under these standardized conditions, classification for clinical purposes requires confirmatory testing because of intraindividual and laboratory variation and the prognostic importance of the diagnosis. As noted earlier, prediabetes as determined by either fasting plasma glucose (FPG) or oral glucose tolerance test (OGTT), corresponds to blood glucose levels of 100–126 mg/dl or 140–200 mg/dl, respectively. Screening Tests for Type 2 and Type 1 Diabetes Mellitus Additional measures are available for classifying individuals with respect to type 1 and T2DM.11 The oral glucose tolerance test is equal or superior in performance to other tests for T2DM, although at intermediate cost, whereas most tests to distinguish type 1 are higher in cost. These considerations enter into the design of population surveys to estimate prevalence of diabetes as well as other epidemiologic investigations. Population screening for purposes of case detection has previously been considered unjus-
Values for Diagnosis of Diabetes Mellitus and Other Categories of Hyperglycemia Glucose Concentration (mmol/L) Whole Blood Venous Capillary
Diabetes mellitus Fasting or 2 h post glucose load IGT Fasting (if measured) and 2 h post glucose load IFG Fasting and (if measured) 2 h post glucose load
Plasma Venous
ⱖ 6.1 ⱖ 10.0
ⱖ 6.1 ⱖ 11.1
ⱖ 7.0 ⱖ 11.1
6.1 and ⱖ 6.7
6.1 and ⱖ 7.8
7.0 and ⱖ 7.8
ⱖ 5.6 and 6.1 6.7
ⱖ 5.6 and ⱖ 6.1 7.8
ⱖ 6.1 and 7.0 7.8
Source: Reprinted with permission from Coronary Heart Disease Epidemiology from Aetiology to Public Health, M Marmot, P Elliott eds, DG Johnston, KGMM Alberti, IF Godsland, M Pierece, S Sheppard, p 715. © 2005 Oxford University Press.
46894_CH13_361_394.qxd
2/25/10
3:41 PM
Page 367
DETERMINANTS 367
tified because of lack of overall benefit of early detection and treatment.
DETERMINANTS Pathways of Insulin Action T2DM and insulin resistance together are the most relevant aspects of diabetes for risks of atherosclerosis and hypertension in the population at large. It is therefore especially important to appreciate the current understanding of the determinants of these conditions and the mechanisms by which they may contribute to these risks. The pathogenesis of insulin resistance is outlined schematically in Figure 13-5.19 Decrease in either glucose-induced insulin secretion from the pancreas, or in tissue response to insulin and uptake of glucose from the circulation, leads to
a chain of events that results in or intensifies insulin resistance. This is true, according to this scheme, even if the initiating process is impairment of beta-cell function in the pancreas, wherein which insulin is produced, although this is considered the less common sequence. The Thrifty Gene Hypothesis Neel introduced a genetic concept to explain the paradoxical observation that diabetes is detrimental to reproduction yet has also been a common condition in human populations, presumably over many generations.20 He postulated that if there was a selective disadvantage of diabetes, there must be a genetically determined selective advantage to counterbalance it. Under conditions of most of human evolution, this would operate by facilitating energy conservation against periods of acute starvation. Neel attributed
Insulin Deficiency
Glucose-Induced Insulin Secretion
Tissue Response to Insulin
Hepatic Glucose Production
Cellular Glucose Uptake
Impaired B-Cell Function Hyperglycemia Basal Hyperinsulinemia Glucose Transport
Postreceptor Defect
Insulin Binding Insulin Resistance
Note: Whether the primary defect initiating the glucose intolerance resides in the B cell or in per ipheral tissues, development of insulin resistance will eventually ensue or become aggravated, respectively. By the time that overt fasting hyperglycemia (>140 mg/dl) develops, both impaired insulin secretion and severe insulin resistance are present. Broken arrows represent positive feedback loops, which result in self-perpetuation of primary defect. Figure 13-5 Pathogenesis of Insulin Resistance in Non-Insulin Dependent Diabetes Mellitus (NIDDM). Source: RA DeFronzo, RC Bonadonna, E Ferrannini, Pathogenesis of NIDDM: A Precarious Balance Between Insulin Action and Insulin Secretion, in International Textbook of Diabetes Mellitus, KGGM Alberti et al., eds. Vol 1, p 617, Copyright © 1992, John Wiley & Sons Limited. Reproduced with permission.
46894_CH13_361_394.qxd
368
2/25/10
3:41 PM
Page 368
CHAPTER 13 DIABETES AND THE METABOLIC SYNDROME
this advantage to what he termed a “thrifty” genotype, once advantageous but becoming detrimental under the dietary conditions of modern societies. The same argument, in principle, has been advanced to explain epidemic obesity. An extensive review of current knowledge concerning specific genetic mechanisms in type 1 diabetes and T2DM was given in Goldbourt and others’ Genetic Factors in Coronary Heart Disease, as well as by Vadheim and Rotter.21,22 Congruence of type 2 between identical twins has been reported to be approximately 90%, indicating strong genetic determination of susceptibility. In type 2, the gene NIDDM1 was identified and reported in a Mexican American population in 1996, but establishing its correspondence to the population-level phenomena of diabetes has been challenging.23 A recent review of genomewide association studies in connection with T2DM noted that some 20 loci had been identified and replicated, with allele frequencies considered as common though with only “modest to small” effects on disease risk.24 Studies of nonobese healthy subjects with a strong family history of T2DM have revealed subclinical metabolic disorders with tissue insulin resistance and pancreatic beta-cell susceptibility to toxic effects of free fatty acids.25 Genetic studies of the metabolic syndrome have also been reported, for example, from a multicenter study under the aegis of the ADA that identified linkages for three distinct metabolic syndrome components—obesity, blood pressure, and blood lipids—at different gene locations between Mexican American and non-Hispanic White participants.26 Predisposing Factors Population studies have suggested a number of predisposing factors for diabetes, such as those reported for men in the British Regional Heart Study.27 Men age 40–59 years were followed for more than 12 years on average, and 178 of 7097 participants developed T2DM. Factors studied included BMI, prevalent coronary heart disease, physical activity, alcohol intake, current smoking, systolic blood pressure, HDLcholesterol, triglycerides, heart rate, and uric acid concentration. Each characteristic was evaluated for its contribution to relative risk of T2DM, with adjustment for all of the others. BMI was clearly the strongest positive predictor of later T2DM, whereas physical activity was strongly inverse or negative as a predictor. Serum triglyceride concentration was second only to BMI in its estimated relative risk. An epidemiologic follow-up study after the first NHANES found incident diabetes to be significantly related to race (Blacks Whites), sex (women
men), age, BMI, subscapular-to-triceps skinfold ratio, systolic blood pressure, and having fewer than 9 years of education.28 Other studies of T2DM, IGT, or insulin resistance in diverse populations, including women, support the finding of BMI, central adiposity, and other anthropometric indices of obesity as consistent predictors. There is conflicting evidence of whether fetal or neonatal influences bear on risk of T2DM. This question is addressed further in Chapter 16, “Social and Physical Environment.” The ADA has defined prediabetes in terms of IFG or IGT by levels of fasting or postload plasma glucose considered most predictive of progression to diabetes—crossing the threshold to levels above 126 or 200 mg/dl, respectively. These are considered the strongest predictors of diabetes, although consideration of other factors such as those indicated above add to reliability of prediction. The ADA includes family history of diabetes and waist–hip ratio as factors independently related to risk of diabetes.29 Other predictors identified in various population studies include fasting insulin levels, fasting and 2-hour postload proinsulin (an insulin precursor also produced in the pancreas), and diminished insulin response to oral glucose.18 Studies of determinants of the metabolic syndrome have identified family history as a predictor, without distinction between genetic and environmental contributions. Racial/ethnic group comparisons have noted variation in prevalence of specific components of the syndrome, but little insight is offered as to whether there are aspects of race/ethnicity that predispose to the syndrome itself. Review of lifestyle behaviors similarly recognizes well-known relations between television watching and other forms of “screen time,” physical inactivity, and dietary patterns on the component risk factors.17 Analysis of follow-up data in the Framingham Heart Study showed consumption of soft drinks to be modestly associated (odds ratios ranging from 1.18 to 1.44) with new-onset metabolic syndrome (defined as three or more of five components: high waist circumference, fasting blood glucose, serum triglycerides, blood pressure, or low HDL-cholesterol) and each of the five components alone.30
MECHANISMS How diabetes and states now recognized as insulin resistance contribute to exacerbation of atherogenesis has been a long-standing question.1 Stout advanced the proposition, later incorporated in the concept of the insulin resistance syndrome, that insulin has a direct
46894_CH13_361_394.qxd
2/25/10
3:41 PM
Page 369
DISTRIBUTION 369
atherogenic effect in addition to potential secondary effects on other risk factors.31 Isolating the postulated effects of insulin from those of concomitant factors affecting atherosclerosis has been a major challenge, addressed, for example, by the Insulin Resistance Atherosclerosis Study (IRAS).32 In that study, insulin sensitivity, rather than insulin resistance, was assessed. Measures of atherosclerosis in IRAS were internal and common carotid arterial thickness, measured by B-mode ultrasound examination. In a cross-sectional analysis of these measurements in several hundred Hispanic and non-Hispanic White men and women with an average age of about 55 years, an inverse relation was found between insulin sensitivity and arterial wall thickness, especially of the internal carotid artery. This association was attenuated by adjustment for other risk factors. No such association was found for Blacks. With this latter exception, and with the limitation of the cross-sectional design, the results suggested a direct role of insulin in atherogenesis. Several mechanisms at the molecular level support the concept of a direct effect of insulin on cardiovascular pathology, as well as indirect effects on inflammation and a prothrombotic state.17,33–35 Mechanisms linking the metabolic syndrome with cardiovascular death, beyond the well-established links with its defining components, have been investigated to include several factors in adipose tissue regulation (adiponectin, leptin, and ghrelin) and the proinflammatory factors interleukin-6 (IL-6) and C-reactive protein (CRP).36 Association of the metabolic syndrome with coronary mortality appeared to account for its modest association with CRP. IL-6 appeared to be independent of the metabolic syndrome.
RELATION TO OTHER FACTORS It seems clear that the presence of diabetes is a marker for multiple influences on development and progression of atherosclerosis. Roots of these disturbances in dietary imbalance and physical inactivity, with their many known adverse effects, would predict these relationships. Diabetes, insulin resistance, and all of the conditions proposed to define the metabolic syndrome are closely related. Additional indirect roles of insulin operating through dyslipidemia and hypertension are supported by evidence of several mechanisms linking diabetes with these conditions.37 Other factors may participate in the increased atherogenesis of insulin resistance and diabetes, such as elevated blood concentrations of glucose, triglyceride, and fibrinogen.
Association of elevated insulin concentrations with obesity, hypertension, and adverse blood lipid profiles is evident in early adulthood and even in childhood.38,39 For example, already at ages 9–10 years, Mexican American children were found to have, in comparison with non-Hispanic White children, significantly greater clustering of elevated blood sugar and insulin concentrations with high triglyceride and low HDL-cholesterol concentrations, high systolic blood pressure, and high BMI. Features of the insulin resistance syndrome or metabolic syndrome appeared early in this population, whose risk of diabetes in adulthood is also especially high. Factors by which diabetes and hyperinsulinemia accelerate atherosclerosis can clearly be present well before adulthood. The Bogalusa Heart Study demonstrated clustering of metabolic syndrome components both in childhood and in early adulthood, with higher levels predicting greater changes from childhood into adulthood.40 Increases in BMI accounted for approximately 50% of the clustering of risk factors in this population. In Quebec, children and adolescents were examined at age 9, 13, or 16 years to investigate patterns of factors implicated in the insulin resistance syndrome.41 Prevalence of the syndrome was 11.5% overall without variation by age or sex. Adiposity was a stronger determinant of clustering of factors than was insulin concentration. Factor analysis revealed three clusters of risk factors: BMI/insulin/ lipids, BMI/insulin/glucose, and diastolic/systolic blood pressure. This finding suggested that characteristics identified with the syndrome were influenced by multiple underlying factors rather than reflecting a single pathophysiologic process.
DISTRIBUTION The United States Prevalence of diabetes among US adults is estimated on the basis of self-reported physician diagnosis of diabetes, with or without addition of laboratory measurement of blood glucose. Data from the NHANES 2003–2006 are shown in Table 13-2, with comparisons from 1988–1994 and 1999–2002.42 For 2003–2006, total prevalence at ages 20 years and older was 10.2%, with 7.7% and 2.5% diagnosed and undiagnosed, respectively. Prevalence was somewhat greater among males, nearly twice as great for non-Hispanic Blacks or African Americans and for Mexicans as for non-Hispanic Whites, and increased sharply with age. Total prevalence increased across
3:41 PM
Sources: CDC/NCHS, National Health and Nutrition Examination Survey. Data from Health, United States, 2008. National Center for Health Statistics, p 276.
370
2/25/10
*Estimates are considered unreliable. Data preceded by an asterisk have a relative standard error (RSE) of 20%–30%. Data not shown have an RSE of greater than 30%. 1 Physician-diagnosed diabetes was obtained by self-report and excludes women who reported having diabetes only during pregnancy. 2 Undiagnosed diabetes is defined as a fasting blood glucose (FBG) of at least 126 mg/dl and no reported physician diagnosis. Respondents had fasted for at least 8 hours and less than 24 hours. Estimates in some prior editions of Health, United States included data from respondents who had fasted for at least 9 hours and less than 24 hours. In 2005–2006, FBG testing was performed at a different laboratory and using a different instrument than testing in earlier years. NHANES conducted a crossover study to evaluate the impact of these changes on FBG measurements. As a result of that study, NHANES recommended that 2005–2006 data on FBG measurements be adjusted to be compatible with earlier years. Undiagnosed diabetes estimates in Health, United States were produced after adjusting the 2005–2006 FGC data as recommended. For more information, see http://www.cdc.gov/nchs/data/nhanes/nhanes_05_06/glu_d.pdf. 3 Persons of Mexican origin may be of any race. Starting with 1999 data, race-specific estimates are tabulated according to the 1997 Revisions to the Standards for the Classification of Federal Data on Race and Ethnicity and are not strictly comparable with estimates for earlier years. The two non-Hispanic race categories shown in the table conform to the 1997 Standards. Starting with 1999 data, race-specific estimates are for persons who reported only one racial group. Prior to data year 1999, estimates were tabulated according to the 1977 Standards. Estimates for single-race categories prior to 1999 included persons who reported one race or, if they reported more than one race, identified one race as best representing their race. See Appendix II, Hispanic origin; Race. 4 Estimates are age-adjusted to the year 2000 standard population using three age groups: 20–39 years, 40–59 years, and 60 years and over. Age-adjusted estimates in this table may differ from other age-adjusted estimates based on the same data and presented elsewhere if different age groups are used in the adjustment procedure. See Appendix II, Age adjustment. 5 Includes all other races and Hispanic origins not shown separately. Notes: Standard errors are available in the spreadsheet version of this table. Available from: http://www.cdc.gov/nchs/hus.htm. Starting with Health, United States, 2007, data use a revised weighting scheme. Data have been revised and differ from previous editions of Health, United States. Data for additional years are available. See Appendix III.
Diabetes Among Adults 20 Years of Age and over, by Sex, Age, and Race and Hispanic Origin: United States, 1988–1994, 1999–2002, and 2003–2006 (Data are based on interviews and physical examinations of a sample of the civilian noninstitutionalized population) Sex, Age, and Race Physician-Diagnosed and Physician-Diagnosed Undiagnosed and Hispanic Origin3 Undiagnosed Diabetes1,2 Diabetes1 Diabetes2 1988–1994 1999–2002 2003–2006 1988–1994 1999–2002 2003–2006 1988–1994 1999–2002 2003–2006 20 years and over, Percent of population age-adjusted4 All persons5 8.3 9.4 10.2 5.4 6.6 7.7 2.9 2.8 2.5 Male 8.8 10.7 11.2 5.4 7.0 7.6 3.4 3.6 3.6 Female 7.9 8.3 9.4 5.4 6.2 7.8 2.5 2.1 *1.6 Not Hispanic or Latino: White only 7.5 7.9 8.8 5.0 5.2 6.4 2.5 2.7 2.4 Black or African American 12.6 14.9 16.0 8.6 11.3 13.2 4.0 3.6 2.8 only Mexican 14.2 13.7 15.7 9.7 10.5 12.4 4.5 3.1 *3.3 20 years and over, crude All persons5 7.8 9.3 10.3 5.1 6.5 7.7 2.7 2.8 2.5 Male 7.9 10.2 10.9 4.8 6.7 7.4 3.0 3.5 3.5 Female 7.8 8.5 9.7 5.4 6.3 8.1 2.4 2.2 1.7 Not Hispanic or Latino White only 7.5 8.4 9.5 5.0 5.5 6.9 2.5 2.9 2.6 Black or African American 10.4 13.4 14.4 6.9 10.1 11.8 3.4 *3.3 2.5 only Mexican 9.0 8.3 10.9 5.6 6.5 7.9 3.4 1.8 *3.0 Age 20–39 years 1.6 *2.3 2.5 1.1 1.7 1.7 *0.6 * * 40–59 years 8.8 9.8 10.6 5.5 6.6 8.3 3.3 3.3 *2.3 60 years and over 18.9 20.9 22.9 12.8 15.1 16.9 6.1 5.8 6.0
Table 13-2
46894_CH13_361_394.qxd Page 370
CHAPTER 13 DIABETES AND THE METABOLIC SYNDROME
46894_CH13_361_394.qxd
2/25/10
3:41 PM
Page 371
DISTRIBUTION
the three survey periods in every sex–race category, as did diagnosed diabetes; undiagnosed diabetes declined slightly or remained constant. The National Diabetes Fact Sheet, a Web-based information service from the Centers for Disease Control and Prevention (CDC), provides updated estimates of prevalence, incidence, and general information about diabetes in the United States.43 Projecting 2003–2006 NHANES data to the 2007 US population estimates, 23.6 million Americans or 7.8% of the population have diabetes and nearly 200,000 people younger than age 20 have diabetes, type 1 or 2. Prediabetes was estimated to affect 57 million adults aged 20 or older, or 25.9% of the adult population, in 2006.44 Among adults, more than 1.5 million new cases are estimated to occur each year. State-specific incidence was estimated for 33 states from self-reported data in the Behavioral Risk Factor Surveillance System.45 Persons with diabetes reported as diagnosed within the year preceding the survey were in-
Table 13-3
cluded in the calculation of incident cases, over two time periods—1995–1997 and 2005–2007. Overall incidence increased from 4.8/1000 per year in the first period to 9.1/1000 in the second. Rates increased in every state, ranging from an increase of 15% in Wyoming to 216% in Idaho. Reasons for these increases may include improved case detection, but this has been deemed unlikely as a full explanation. Rates are among the highest in American Indians and Alaska Natives, according to data from the Indian Health Service.46 Increasing prevalence among women and men younger than age 35 from 1994 to 2004 based on health services data may represent a mix of increasing incidence and increasing detection, but reached 17.1% by 2004. T2DM is now known to occur in childhood and adolescence but has not yet been tabulated in Health US as has been the case for adults. Difficulty in comparing estimates in children is readily apparent in Table 13-3, drawn from varied population- and clinic-based sources in North America.47 Depending on data type,
Estimates of the Magnitude of Type 2 Diabetes in North American Children Age Years Race/Ethnicity (Years)
Study types Population-based studies Arizona
1992–1996
Pima Indians
Manitoba NHANES III, all US
1996–1997 1988–1994
First Nations Whites, African Americans, Mexican Americans
Clinic-based studies Indian Health Services (all US)
10–14 15–19 10–19 12–19
1996
American Indians
1998
First Nations
Cincinnati, OH Case series
1994
Whites, African-Americans
10–19
Cincinnati, OH
1994
Whites, African-Americans
Charleston, SC San Diego, CA
1997 1993–1994
0–19 10–19 0–19 0–16
San Antonio, TX Ventura, CA
1990–1997 1990–1994
Manitoba
371
0–14 15–19 5–14 15–19
Clinic-based studies
Blacks Whites, African-Americans, Hispanics, Asian Americans Hispanics, Whites Hispanics
0–17
Estimates Prevalence per 1,000 22.3 50.9 36.0 in girls 4.1*
1.3* 4.5* 1.0 2.3 Incidence per 100,000/year 7.2 Percentage of type 2 diabetes among new cases of diabetes 16 33 46† 8 18 45
*Estimates include type 1 and 2 diabetes; †percentage of type 2 diabetes among nonincident cases of diabetes. Adapted from A. Fagot Campagna, D.J. Pettitt, MM Engelgau, NR Burrows, LS Geiss, R Valdez, G Beckles, J Saaddine, EW Gregg, DF Williamson, KM Venkat Narayan, J Pediatrics. In press. Source: Reprinted with permission from Diabetes Care, Vol 23, No 3, March 2000. American Diabetes Association, Consensus Statement, Type 2 Diabetes in Children and Adolescents, p 383.
46894_CH13_361_394.qxd
372
2/25/10
3:41 PM
Page 372
CHAPTER 13 DIABETES AND THE METABOLIC SYNDROME
race/ethnicity, and age, prevalence is reported from 1.0 to 50.9%. One report of incidence indicates a rate of 7.2/100,000 per year among mixed Whites and African Americans aged 10–19 years. Clinical diabetic populations with newly diagnosed diabetes are described as 8–45% T2DM, whereas historically virtually all diagnoses at this age were type 1. Healthy People 2010 identified 17 objectives for improvement in diabetes.48 The objective of preventing diabetes was represented by a baseline (1997–1999 data) incidence of 5.5/1000/year and a target level of 3.8/1000/year. At the Midcourse Review evaluation (2000–2002 data) incidence had increased rather than decreased, as data above indicated. Similarly, the total prevalence of diagnosed diabetes had increased—baseline 40/1000, target 25/1000, midcourse evaluation, and prevalence increased. However, large gains were made in reducing diabetes-related deaths and cardiovascular deaths in persons with diabetes, as well as performance of annual examinations for microalbuminuria. Regarding the metabolic syndrome, prevalence among adults aged 20 years or older was estimated from NHANES III, 1988–1994, by Ford and others, using the ATPIII criteria (three of five measures at critical levels of: waist circumference, serum triglycerides, HDL-cholesterol, blood pressure, and serum
Table 13-4
glucose).49 Overall prevalence was 23.7%, with a steep age gradient—6.7% at age 20–29 years, and greater than 40% over age 60 years. The column “ⱖ 3” in Table 13-4 refers to all who qualified as having the metabolic syndrome. For both men and women, prevalence was highest among Hispanics. Striking variation appeared by sex and race/ethnicity in prevalence of the individual components of the syndrome. For each of the five components, shown in Table 13-5, the range from least to greatest prevalence (%) across the three major race/ethnic groups was abdominal obesity, 23.3 for African American men to 62.7 for Mexican American women; hypertriglyceridemia, 14.4 for African American women to 39.7 for Mexican American men; low HDL-cholesterol, 22.6 in African American men to 46.3 in Mexican American women; high blood pressure or medication use, 27.8 in White women to 46.3 in Mexican American women; and high fasting glucose or medication use, 8.5% in White women to 21.4% in Mexican American men. The metabolic syndrome, as defined, appears to be dominated by different components across sex and race/ethnic groups, although this analysis did not present the distribution of characteristics restricted to those who met criteria for the syndrome.
Age-Adjusted Prevalence of 1 or More Abnormalities of the Metabolic Syndrome Among 8814 US Adults ⱖ 20 Years, National Health and Nutrition Survey III, 1988–1994 No. of Metabolic Abnormalities, % (SE) ⱖ1 ⱖ2 ⱖ3 ⱖ4 5 71.2 (1.0) 43.9 (1.1) 23.7 (0.8) 10.4 (0.5) 2.7 (0.3) 71.5 (1.2) 44.9 (1.3) 24.0 (1.1) 11.1 (0.9) 2.4 (0.4) 70.9 (1.2) 42.7 (1.3) 23.4 (0.9) 9.6 (0.5) 2.9 (0.3)
Total Men Women Race or ethnicity White African American Mexican American Other Men White African American Mexican American Other Women White African American Mexican American Other
70.1 (1.2) 75.6 (0.8) 78.9 (1.0) 71.2 (3.1)
43.2 (1.2) 45.1 (1.0) 54.4 (1.1) 41.8 (4.0)
23.8 (1.0) 21.6 (0.8) 31.9 (1.3) 20.3 (3.3)
10.8 (0.6) 8.4 (0.7) 12.0 (1.0) 7.1 (1.5)
2.9 (0.3) 1.8 (0.4) 2.3 (0.4) 1.5 (0.6)
71.5 (1.4) 70.3 (1.3) 74.8 (1.5) 70.2 (4.6)
45.5 (1.5) 37.3 (1.5) 51.5 (1.5) 42.9 (4.2)
24.8 (1.4) 16.4 (1.1) 28.3 (1.8) 20.9 (4.7)
12.4 (1.1) 6.3 (0.8) 9.4 (1.0) 3.6 (1.2)
2.8 (0.5) 1.2 (0.3) 1.6 (0.4) 0.9 (0.5)
68.4 (1.5) 80.0 (1.0) 84.0 (0.9) 71.3 (4.6)
40.7 (1.5) 51.3 (1.3) 57.7 (1.4) 40.0 (4.6)
22.8 (1.1) 25.7 (1.3) 35.6 (1.5) 19.9 (3.1)
9.2 (0.6) 10.0 (0.9) 14.7 (1.3) 10.5 (2.5)
3.0 (0.3) 2.3 (0.5) 3.1 (0.6) 2.1 (1.2)
*See the “Methods” section for a description of the 5 criteria of the metabolic syndrome. Source: Reprinted with permission from Journal of the American Medical Association, Vol 287, No 3, ES Ford, WH Giles, WH Dietz, p 358, © 2002 American Medical Association.
30.5 (1.2) 23.3 (1.3) 30.6 (1.7) 26.4 (7.5) 43.5 (1.4) 62.1 (1.5) 62.7 (1.7) 40.0 (4.8)
1712 1116 1277 160 1887 1296 1172 194
25.0 (1.1) 14.4 (1.0) 35.2 (1.3) 26.0 (4.4)
36.9 (2.0) 21.4 (1.2) 39.7 (1.5) 29.4 (4.0)
31.1 (1.3) 17.7 (0.8) 37.7 (1.0) 27.3 (3.3)
39.3 (1.9) 34.0 (1.7) 46.3 (1.7) 39.6 (4.6)
36.8 (1.6) 22.6 (1.7) 33.7 (2.0) 33.2 (5.2)
37.9 (1.5) 28.8 (1.3) 39.6 (1.5) 37.1 (4.5)
27.8 (0.9) 43.3 (1.3) 32.4 (1.7) 23.7 (2.3)
37.2 (1.8) 49.6 (1.5) 40.2 (1.7) 34.4 (4.0)
32.8 (1.0) 46.3 (0.9) 36.6 (1.2) 29.6 (2.9)
Source: Reprinted with permission from Journal of the American Medical Association, Vol 287, No 3, ES Ford, WH Giles, WH Dietz, p 357, © 2002 American Medical Association.
*HDL indicates high-density lipoprotein.
37.2 (0.9) 44.6 (1.2) 45.7 (1.3) 33.6 (5.2)
3599 2412 2449 354
8.5 (0.6) 15.5 (1.3) 18.5 (1.2) 14.4 (2.9)
15.6 (1.0) 14.5 (1.1) 21.4 (1.5) 15.1 (3.3)
11.9 (0.6) 15.1 (0.9) 20.0 (1.0) 14.3 (2.0)
3:41 PM
Total Men Women Race or ethnicity White African American Mexican American Other Men White African American Mexican American Other Women White African American Mexican American Other
Age-Adjusted Prevalence of Individual Metabolic Abnormalities of the Metabolic Syndrome Among 8814 US Adults ⱖ 20 Years, National Health and Nutrition Survey III, 1988–1994* % (SE) High Blood High Fasting No. of Abdominal Low HDL Pressure or Glucose or Participants Obesity Hypertriglyceridemia Cholesterol Medication Use Medication Use 8814 38.6 (0.8) 30.0 (1.1) 37.1 (1.2) 34.0 (0.8) 12.6 (0.5) 4265 29.8 (1.2) 35.1 (1.7) 35.2 (1.5) 38.2 (1.4) 15.6 (0.8) 4549 46.3 (1.2) 24.7 (0.9) 39.3 (1.4) 29.3 (0.8) 10.0 (0.6)
2/25/10
Table 13-5
46894_CH13_361_394.qxd Page 373
DISTRIBUTION 373
46894_CH13_361_394.qxd
374
2/25/10
3:41 PM
Page 374
CHAPTER 13 DIABETES AND THE METABOLIC SYNDROME
Global Distribution Figure 13-6 indicates the prevalence of diabetes (type 1 and type 2) and glucose intolerance according to WHO criteria in 33 population groups.51 The range is from less than 5% to more than 60% of affected persons in the group aged 30–64 years. These results combine sex groups and are age standardized to the world population. In most populations, the combined frequency was below 25% but in two it exceeded 50%. With the exception of Nauru and the Pima Indian population of the United States, prevalence varied along a continuous gradient. This suggests that common influences varying only in degree may account for differences in prevalence in most popula-
Another analysis from NHANES III presented data on prevalence of metabolic syndrome in participants aged 12–19 years, by use of the same characteristics as in the ATPIII definition but with cut-points defined for adolescents.50 Overall, 6.1% of males and 2.1% of females were affected, with a concentration of 30% among those who were overweight (partly, of course, because of abdominal circumference as one of the defining criteria). Prevalence varied by sex, race/ethnicity, region, and BMI status. As described above for adults, prevalence of the component conditions varied markedly by sex and race/ethnicity, giving the syndrome different meanings across the groups.
Mapuche Indian, Chile urban Chinese, Da Qing rural Melanesian, PNG Polish rural Polynesian, W. Samoa rural Bantu, Tanzania rural Indian, India Russian Brazilian rural Melanesian, Fiji Italian, Sanza Maltese urban Bantu, Tanzania Italian, Laurino White, USA Tunisian urban Hispanic, USA* rural Micronesian, Kiribati urban Polynesian, W. Samoa urban Indian, India urban Melanesian, Fiji rural Hispanic, USA Black, USA urban Indian, S. Africa Puerto Rican, USA urban Hispanic, USA# Chinese, Mauritius urban Hispanic, USA• rural Indian, Fiji urban Indian, Fiji urban Micronesian, Kiribati Micronesian, Nauru Pima Indian, USA
25%
0
10
20
30
50%
40
50
60
70
Prevalence (%) * upper income # middle income low income
Diabetes mellitus Impaired glucose tolerance
Figure 13-6 Prevalence of Total Glucose Intolerance (Diabetes and Impaired Glucose Tolerance) in Selected Populations Aged 30–64 Years, Age Standardized to the World Population of Segi, Sexes Combined. Source: Reprinted with permission from H King and M Rewers, WHO Ad Hoc Diabetes Reporting Group, Diabetes Care, Vol 16, p 170, © 1993, American Diabetes Association.
46894_CH13_361_394.qxd
2/25/10
3:41 PM
Page 375
RATES AND RISKS
tions, whereas determinants of prevalence in the two extreme cases may be qualitatively distinct. The impression from the figure is that the prevalence of diabetes contributed more than did IGT to these population differences. More detailed information from several populations demonstrates ethnic differences in prevalence of T2DM within the same country.52 In some instances, ethnic variation was marked, with a fourfold or greater difference among groups such as White and Aborigine groups in Australia. The gradient in the United States was from 6.1 to 9.9, 12.6, and 34.1% prevalence among Whites, Blacks, Mexican Americans, and Pima Indians, respectively. Similarly, the contrasting situation of Asians and Europeans in the United Kingdom has been of special interest, the former having prevalence of diagnosed diabetes 3.8 or more times that of Europeans in the same geographic area.53 Evidence suggests strongly that diabetes is a significant problem in developing countries and especially in their urban components. The global frequency of diabetes has been estimated by the Disease Control Priorities in Developing Countries Project, at a prevalence of 5.1% in 2003 and projected to 6.3%—more than a 20% increase—by 2025 (Table 13-6).8 Diabetes would then affect more than 330 million people worldwide. Current estimates are that 959,000 deaths and nearly 20 million DALYs (life years lost to disability) worldwide are attributable to diabetes annually. Developing countries have the greatest burden to date, although this is nearly equaled in the region of Europe and Central Asia. Overall, developing countries dominate by far in diabetes burden and are far short in medical expenditures to address it, relative to developed countries.
RATES AND RISKS The Seven Countries Study, so informative regarding population differences in coronary or all-cause death rates attributable to other factors, did not examine diabetes and related conditions. Keys ascribed this decision to the view that reliable assessment of blood glucose required both fasting and postglucose load blood samples, for which funding was insufficient. In addition, risk of loss to participation was considered to be high.54 At the level of individual differences in cardiovascular risks and outcomes due to diabetes and related conditions, many questions have been addressed through population surveys, case-comparison studies, cohort studies, and community surveillance programs.
375
On this basis, several topics regarding the relation of diabetes to cardiovascular disease can be reviewed. Asymptomatic Hyperglycemia and Coronary Heart Disease A landmark international collaboration regarding blood glucose and coronary heart disease, reported in 1979, led to assembly of data from 14 population studies in Australia, Europe, Japan, and the United States.55 The purpose was to determine whether elevated blood glucose concentration, in the absence of diabetes, was associated with coronary heart disease. Because methods of glucose determination differed among studies, the focus was on the relation between blood glucose concentration and coronary disease as within-population associations. In four of the seven studies in which postload glucose determinations were available, prevalent coronary heart disease defined as definite post-myocardial infarction was associated with baseline blood glucose concentration. In most of the studies coronary heart disease, however defined, was significantly more frequent in the highest than in lower quintile groups of the blood glucose distribution. Mortality from coronary heart disease, cardiovascular diseases, and all causes at 5 years was examined in 11 of the studies. The independent contribution of blood glucose levels to risk was assessed after adjustment for age, body mass index, systolic blood pressure, serum cholesterol, and cigarette smoking. No significant relation was found for any of the three outcomes in most of the studies. A striking exception was the analysis based on postload glucose, unique to the Chicago Peoples Gas Study, which showed a highly significant association for each cause of death category. Casual—nonfasting—glucose in the same study was not associated with coronary or all-cause mortality. In the same 11 studies, the ratio of all-cause mortality in the highest quintile of baseline blood glucose concentration to that in the lowest quintile ranged from 0.34 to 6.07; five results were below 1.00, or inverse. Overall, the investigators concluded that “At this juncture, therefore, asymptomatic hyperglycemia cannot be designated an established risk factor for coronary heart disease and the major adult cardiovascular diseases.”55, p 837 The indication that results might differ between fasting and postload blood glucose levels as predictors of cardiovascular outcomes continued to stimulate interest that gained new attention when the ADA approved diagnostic criteria for diabetes based on fasting blood glucose (FBG). Comparison of the predictive role of FBG versus 2hBG (2 hour postload blood glucose) for death from all causes, cardiovascular disease,
Source: Number of persons with diabetes, prevalence of diabetes, and direct medical costs of diabetes, International Diabetes Federation 2003b; all other information WHO 2004. From Disease Control Priorities in Developing Countries, Second Edition, edited by DT Jamison et al., Copyright © 2006. Courtesy of The International Bank for Reconstruction and Development/ The World Bank.
3:41 PM
Estimated Numbers of People Aged 20 to 79 with Diabetes, Mortality, DALYs, and Direct Medical Costs Attributable to Diabetes, by Regions Number of People Prevalence Direct Medical Costs, 2003 Disability (Thousands) (Percent) (US$ million) Deaths Adjusted Life Low High 2001 Years, 2001 Region 2003 2025 2003 2025 Estimate Estimate (Thousands) (Thousands) Developing countries 140,849 264,405 4.5 5.9 12,304 23,127 757 15,804 East Asia and the Pacific 31,363 60,762 2.6 3.9 1368 2656 234 4930 Europe and Central Asia 25,764 33,141 7.6 9.0 2884 5336 51 1375 Latin America and the Caribbean 19,026 36,064 6.0 7.8 4592 8676 163 2775 Middle East and North Africa 10,792 23,391 6.4 7.9 2347 4340 31 843 South Asia 46,309 94,848 5.9 7.7 840 1589 196 4433 Sub-Saharan Africa 7595 16,199 2.4 2.8 273 530 82 1448 Developed countries 53,337 68,345 7.8 9.2 116,365 217,760 202 4192 World 194,186 332,750 5.1 6.3 128,669 240,887 959 19,996
376
2/25/10
Table 13-6
46894_CH13_361_394.qxd Page 376
CHAPTER 13 DIABETES AND THE METABOLIC SYNDROME
46894_CH13_361_394.qxd
2/25/10
3:41 PM
Page 377
RATES AND RISKS 377
coronary heart disease, and stroke was undertaken by the DECODE (Diabetes Epidemiology: Collaborative analysis Of Diagnostic criteria in Europe) Study Group.56 Ten prospective studies in Europe, all based on 75 g glucose load for baseline blood glucose, included more than 22,000 participants followed for an average period of 8.8 years. It was found that 2hBG was significantly predictive of all-cause, cardiovascular, and coronary mortality (but not stroke), whereas FBG was significant only for all-cause mortality (hazard ratio 1.21, confidence interval 1.01–1.44, versus 1.73, confidence interval 1.45–2.06, for 2hBG). When the associations with 2hBG were adjusted by addition of FBG, no significant improvement in prediction was observed, but when the reverse analysis was done, 2hBG significantly improved the predictions found for FBG. This led the DECODE Group to conclude that 2hBG was the better predictor. The long-term cohort study of British civil service workers, the Whitehall Study, reported on the relation between 2 hour postload blood glucose levels determined at baseline examinations in 1967–1969 and deaths through 2002.57 They found a significant linear gradient of risk related to glucose levels upward from 4.6 mmol/L (83 mg/dl) as a threshold, with no increase in risk at lower levels, which ranged from 3.0 to 4.6 mmol/L. Adjustment for other factors (existing coronary disease and several risk factors assessed at baseline) attenuated the graded relation by 45%. Direct comparison of the observed glucose values with other studies was limited, because the glucose load in this study was 50 g rather than the more widely used 75 g, which would result in higher glucose concentrations. The report concluded that “our evidence is consistent with no association between 2hBG and CHD risk in the lower and central parts of the population distribution of postload glucose but a relatively low level at which CHD risk associated with postload glycemia begins to rise.”57, pp 29–30 A separate question concerns the relation between hyperglycemia, or elevated blood glucose, on the outcome of acute coronary events. This was the topic of an AHA Scientific Statement summarizing evidence that both 30-day and 1-year mortality after hospitalization for acute coronary syndromes (ACS) increased with the blood glucose level determined at hospital admission.58 Higher blood glucose was associated with 30-day mortality, which increased fourfold in patients without diabetes but only slightly in the presence of diabetes. Whether blood glucose is directly related to this increase or is only a marker of an unidentified mechanism was considered to be an important question for further research.
Insulin and Coronary Heart Disease Whether blood insulin concentration is itself a risk factor for coronary heart disease has been studied with respect to both endogenous insulin (hyperinsulinemia or insulin resistance) and exogenous insulin (insulin therapy). Studies of both aspects have been inconsistent. Possible reasons for differences in findings were discussed in a review in Diabetes in America (1995), by Wingard and Barrett-Connor, who concluded that “the role of insulin as a heart disease risk factor remains controversial.”59, p 444 McKeigue and Keen, evaluating the evidence available in 1991, expressed doubt that insulin caused the relation of the insulin resistance syndrome to coronary risk. They pointed to the possibility that proinsulin or split proinsulins, which are separately identifiable insulin-like molecules, may be more directly involved.60 Further population studies, they noted, may help to resolve this question. Diabetes and Coronary Heart Disease Wingard and Barrett-Connor also summarized the extensive literature on coronary heart disease and diabetes.59 The National Hospital Discharge Surveys between 1989 and 1991 reported the frequencies of cardiovascular diagnoses, including several categories of coronary heart disease, among persons whose discharge diagnoses did or did not include diabetes. In every category, the diagnosis of diabetes was associated with a higher percentage of cardiovascular diagnoses, for both women and men. Studies of prevalence of coronary heart disease in relation to concurrent findings of NIDDM, IGT, or normal blood glucose indicated higher frequency of coronary heart disease among women in each of four categories of coronary heart disease and among men in three of four categories. Excess prevalence was greater for women than for men, counter to the usually observed advantage of women over men in age-specific coronary heart disease rates; this suggested that diabetes may offset factors accounting for the usual advantage of women. Incident coronary heart disease was compared between categories of diabetes in four cohort studies. Three of these studies included women. The estimated risk ratio for new coronary events was greater in women than in men, whether adjusted for age alone or for multiple (unspecified) factors. Similarly, risks of reinfarction after acute myocardial infarction were reviewed and were found elevated for diabetics, variously defined. A subsequent meta-analysis of 8 studies with 10 race/ethnic group-specific analyses further established the absence of a female advantage
46894_CH13_361_394.qxd
378
2/25/10
3:41 PM
Page 378
CHAPTER 13 DIABETES AND THE METABOLIC SYNDROME
in coronary mortality relative to men, in the presence of diabetes (Figure 13-7).61 A further report from the DECODE Study Group, cited previously, addressed a gender difference in all-cause and coronary mortality in the presence of hyperglycemia and newly diagnosed diabetes.62 These conditions had greater adverse effects on all-cause and cardiovascular mortality among women than among men. Mortality from ischemic heart disease in four US sex–race groups, at ages 45–64 years, was studied on the basis of national death certificate data, the National Health Interview Survey, and the 1986 National Mortality Follow-Back Survey.63 These data sources were used for improving estimates of the true relation of diabetes to coronary heart disease deaths. This report emphasized limitations of death certificate data as a basis for ascertaining the presence of diabetes: the follow-back data showed diabetes was present in from two to three times as many decedents as was indicated on their death certificates. For women and men, Blacks and Whites, there was a marked excess of ischemic heart disease mortality among persons with diabetes, after adjustment for underreporting of diabetes. The relation of diabetes to cardiovascular death has evidently been underestimated substantially in many studies. Mexican Americans were once thought to experience lower mortality from coronary heart disease than non-Hispanic Whites, on the basis of state-level statistical reports. However, in the one community, Corpus Christi, Texas, they have been found in direct community surveillance to have greater hospitalization rates and shorter survival after acute myocardial infarction than their non-Hispanic White counterparts.64 Because diabetes is more prevalent in the Mexican American population, analysis of the relation of a self-reported history of diabetes to coronary events was undertaken (Figure 13-8). The pattern of survival after myocardial infarction indicates that diabetes conferred similar losses in survival for both Mexican Americans and non-Hispanic Whites. Much, though not all, of the mortality differential between ethnic groups can be accounted for by the more frequent presence of diabetes among Mexican Americans. The contribution of diabetes to risk of subsequent coronary heart disease was the focus of a study by Stern and others in a Finnish cohort.65 They compared 7-year incidence of fatal and nonfatal coronary events between nondiabetic persons with or without prior myocardial infarction at baseline (18.8 and 3.5%, respectively) and diabetic persons in the corresponding subgroups (45.0 and 20.2%, respec-
tively). There was no significant difference in risk between persons with prior myocardial infarction but not diabetes and persons with diabetes but no prior myocardial infarction. It was concluded that treatment of cardiovascular risk factors in persons with diabetes should therefore be fully as aggressive as for persons with prior myocardial infarction. (This became the basis for regarding diabetes as a “CHD equivalent” in risk stratification in ATPIII—see Chapter 11, “Adverse Blood Lipid Profile.”) Several further studies of this question have supported this finding and also noted that greater duration of diabetes is associated with increased coronary heart disease mortality.66–69 For example, a 25-year follow-up of the men screened for the Multiple Risk Factor Intervention Trial (MRFIT) permitted comparison of cause-specific mortality among those who reported a previous hospitalization for heart attack, use of medications for diabetes, both conditions, or neither condition at screening (Table 13-7). Ageadjusted death rates for all causes and cardiovascular causes showed nearly identical rates for those with past history of MI only or diabetes only; differences between these groups in non-CVD causes were due largely to deaths from diabetes in the group with diabetes at entry. Metabolic Syndrome and Coronary Heart Disease Studies of the relation between metabolic syndrome and coronary heart disease can be illustrated by examples of cross-sectional analysis of data from NHANES III and the prospective British Regional Heart Study.70,71 The former analysis indicated that prevalence of coronary heart disease was greatest among persons with both diabetes and metabolic syndrome defined by the ATPIII criteria. Presence of diabetes in the absence of the metabolic syndrome was not associated with increased prevalence of coronary heart disease. The second of these reports compared the metabolic syndrome and the Framingham Risk Score regarding their predictive power for coronary heart disease, stroke, and T2DM mellitus. Metabolic syndrome predicted diabetes better than coronary heart disease, but Framingham risk score was the better predictor of coronary heart disease. Stroke Association between diabetes and stroke has been demonstrated in many studies, including autopsy examination of cerebral circulation and carotid arteries, clinical series of diabetics, and population studies in several countries. A review by Pyörälä and
46894_CH13_361_394.qxd
2/25/10
3:41 PM
Page 379
RATES AND RISKS 379
Summary Estimates for Coronary Heart Disease Mortality in Men Slievers et al77 (Pima Indian) Pan et al56 (White) Keil et al53 (Black) Keil et al53 (White) Kleinman et al54 (White) Vilbergsson et al13 (White) Jousilahti et al22 (White) Collins et al52 (Melanesian) Collins et al52 (Indian) Barrett-Connor et al51 (White) Random Effects 0.10
10.00
1.00
100.00
(A) Summary Estimates for Coronary Heart Disease Mortality in Women Slievers et al77 (Pima Indian) Pan et al56 (White) Keil et al53 (Black) Keil et al53 (White) Kleinman et al54 (White) Vilbergsson et al13 (White) Jousilahti et al22 (White) Collins et al52 (Melanesian) Collins et al52 (Indian) Barrett-Connor et al51 (White) Random Effects 0.10
1.00
10.00 Odds Ratio (B)
100.00
1000.00
Figure 13-7 Odds Ratios and 95% Confidence Intervals for the Risk of Coronary Heart Disease Mortality for Diabetic vs Nondiabetic Men (A) and Women (B). Source: Reprinted with permission from Archives of Internal Medicine, Vol 162, AM Kanaya et al., p 1741. © 2002 American Medical Association.
Source: Reprinted with permission from Archives of Internal Medicine, Vol 164, O Vaccaro et al., p 1440. © 2004 American Medical Association.
Abbreviations: CHD, coronary heart disease; CVD, cardiovascular disease; MI, myocardial infarction; MRFIT, Multiple Risk Factor Intervention Trial. *A total of 688 deaths (646, 22, 19, and 1 in the Neither MI nor Diabetes, MI Only, Diabetes Only, and Both Diabetes and MI groups, respectively) had unknown causes. † Rates are given as age-adjusted deaths per 10,000 person-years.
Cause of Death All causes CVD CHD Stroke Other CVD Non-CVD Cancer Renal disease Diabetes Other
3:41 PM
All-Cause and Cause-Specific Mortality According to History of MI and Diabetes at Initial Screening for the MRFIT Neither MI nor Diabetes MI Only Diabetes Only Both Diabetes and MI (n 322,775) (n 4625) (n 4809) (n 338) No. of Age-Adjusted No. of Age-Adjusted No. of Age-Adjusted No. of Age-Adjusted Deaths* Rate† Deaths* Rate† Deaths* Rate† Deaths* Rate† 76,419 107.3 2715 278.0 2855 277.3 250 421.8 30,620 43.1 1852 193.7 1502 144.2 159 263.6 20,795 29.2 1498 159.4 1087 104.0 126 207.9 3168 4.5 90 8.1 154 14.2 10 11.7 6657 9.4 264 26.2 261 26.0 23 44.1 45,153 63.3 841 81.6 1334 131.5 90 156.1 28,244 39.6 470 45.5 431 40.2 32 52.4 665 0.9 24 2.0 55 5.8 3 2.7 1101 1.5 35 3.2 445 46.3 31 52.8 15,143 21.2 342 30.9 403 39.2 24 48.3
380
2/25/10
Table 13-7
46894_CH13_361_394.qxd Page 380
CHAPTER 13 DIABETES AND THE METABOLIC SYNDROME
46894_CH13_361_394.qxd
2/25/10
3:41 PM
Page 381
RATES AND RISKS 381
Figure 13-8 Proportion of Patients Hospitalized with Acute Myocardial Infarction (MI) Surviving 0–44 Months, by Diabetic Status and Ethnicity, Corpus Christi Heart Project, 1988–1990. Source: Reprinted with permission from PRF Orlander et al., The Relation of Diabetes to the Severity of Acute Myocardial Infarction and Post-Myocardial Infarction Survival in Mexican Americans and Non-Hispanic Whites: The Corpus Christi Heart Project, Diabetes, Vol 43, p 900, © 1994, American Diabetes Association.
colleagues demonstrated increased risk of stroke by a factor of two to four times in diabetics when compared with nondiabetics.72 The 20-year follow-up of women and men in the Framingham Heart Study reported that fatal and nonfatal occlusive stroke, together, occurred 2.6 times more frequently in diabetic men and 3.8 times more frequently in diabetic women than in their nondiabetic counterparts. Again, as was the case for coronary heart disease, diabetes had a greater impact on risk of stroke for women than for men. The American Heart Association report on risk factors for stroke emphasized diabetes among the “potentially modifiable” factors in relation to occlusive stroke, implicating impaired glucose tolerance, hyperinsulinemia, and insulin resistance syndrome in the increased risk.73 Associations of diabetes with cardiovascular conditions extend beyond coronary heart disease and stroke. The review by Wingard and Barrett-Connor linked diabetes with congestive heart failure and cardiomyopathy, and Pyörälä and colleagues described associations of diabetes with atherosclerosis of the aorta and peripheral arteries as well.59,72
Large- and Small-Vessel Disease A leading source of information on factors associated with development of vascular disease among diabetics is the WHO Multinational Study of Vascular Disease in Diabetics.74 In this study, the prevalence of large-vessel disease (affecting the coronary, cerebral, and peripheral arteries) and small-vessel disease of the eyes and kidneys was determined among 6695 diabetic men and women in three age groups from 35 to 54 years in local centers in 14 countries. The results indicated an overall prevalence of large-vessel disease in nearly 30% of males and 40% of females. For all ages together, prevalence of this condition ranged from 21.9 to 52.9% for women and from 19.1 to 38.4% for men across 14 countries. Cultural differences in interpretation of questionnaire items for chest pain and other symptoms limit the interpretation of these population differences, however. In multivariate regression analysis, age, duration of diabetes, systolic blood pressure, body mass index, and total cholesterol were related to large-vessel disease in men; only age, blood pressure, and body mass index were related in women. In both men and
46894_CH13_361_394.qxd
382
2/25/10
3:41 PM
Page 382
CHAPTER 13 DIABETES AND THE METABOLIC SYNDROME
women, small-vessel disease of the eye was associated with duration of diabetes, systolic blood pressure, body mass index, and treatment of diabetes, whereas vascular disease of the kidney was associated with duration of diabetes and cholesterol concentration alone. Microalbuminuria, another manifestation of renal involvement with diabetes, has been found to be significantly associated with all-cause mortality in several European studies. It has often, though not always, correlated with manifestations of coronary heart disease in cross-sectional surveys.59 It is unclear to what extent these associations are independent of other risk factors and whether microalbuminuria may be an expression of the underlying atherosclerotic process. Public Health Impact Diabetes has been estimated by the Global Burden of Disease and Risk Factors Study as the cause of 960,000 deaths at all ages worldwide in 2001.75 It ranked among the top 10 causes of death in Latin America and the Caribbean, but not other Regions. Yach and others estimated that 175 million persons worldwide had diabetes in 2000 and projected 353 million in 2030.76 Among the most populous countries, of which 12 have more than 100 million population, percentage prevalence ranged from 2.4 (China) to 8.8 (US) % in 2000 and from 3.7 (China) to 11.2 (US) % in 2030. Prevalence was projected to increase in each of these countries, as well as all developing countries (4.1 to 6.0%), developed countries (6.3 to 8.4%), and the world (4.6 to 6.4%). Given the evidence reviewed above regarding rates and risks of cardiovascular diseases in consequence of diabetes, these projected trends imply a significantly growing impact of diabetes in contributing to cardiovascular morbidity and mortality in the coming years.
PREVENTION AND CONTROL Recent demonstration, through controlled clinical trials, that T2DM mellitus could be prevented by lifestyle intervention has greatly broadened the scope of prevention and control of diabetes. Formerly the focus was on prevention or treatment of complications in people with diabetes. Much of the research addressed questions of improving control of blood glucose levels, and the benefits of doing so, in both type 1 and T2DM. The new emphasis on prevention of T2DM remains largely a high-risk oriented strategy, based on identification and selection of persons with above-average risk of developing diabetes- those with IGT, IFG, or prediabetes. Individual-level measures still predominate in discussion of diabetes prevention.
However, the conditions immediately predisposing to the early pathogenesis of diabetes—dietary imbalance, physical inactivity, and obesity—and the additional factors that lead to its macrovascular complications—adverse lipid profile, high blood pressure, and smoking—all offer community or population-wide dimensions to prevention and control. Prevention and control of diabetes therefore incorporates the population-wide strategy, as well as the global measures addressed in other chapters. Individual Measures Prevention of T2DM New knowledge established the efficacy of lifestyle modification (diet and physical activity) in preventing progression from IGT or IFG to reach the critical level for diagnosis of diabetes. On the basis of earlier epidemiologic studies of factors leading to progression of IGT or IFG to T2DM and efforts to modify this progression by drug or behavioral intervention, the World Bank and the Ministry of Health of The People’s Republic of China sponsored the Da Qing IGT and Diabetes Study to investigate this concept further.77 Screening in 33 healthcare clinics in 1986 identified persons with IGT, and mean BMI 25.8 kg/m2, who were randomly assigned by clinic to one of four groups: control (patient education and information only); diet only; physical activity only; or diet plus physical activity. Monitoring over a period of 6 years demonstrated incidence of T2DM to be 67.7% in the control group and from 41.1 to 46.0% in the intervention groups. Both lean and overweight participants showed benefit of intervention. Overall, the interventions resulted in a 54 to 69% decrease in diabetes incidence, as reported in 1997. A trial in Finland, with recruitment from 1993–1998, also implemented lifestyle interventions aimed at prevention of progression from IGT to diabetes, in middle-aged adults with mean BMI 31 kg/m2.7 Intervention targets were improvements in diet and physical activity that would bring about weight loss. Cumulative incidence of diabetes was 11% in the intervention group and 23% in the control group, a 58% reduction. Notable were statistically significant improvements in weight, fasting, and 2hPG blood glucose, HDL-cholesterol and triglycerides, and systolic and diastolic blood pressure attributable to the lifestyle intervention. Both lifestyle and pharmacologic intervention were tested in the Diabetes Prevention Program (DPP), a US trial of prevention in adults with mean BMI 34 kg/m2.78 After an average of 2.8 years of follow-up, incidence of diabetes was 11.0% in the placebo control group, 7.8%
46894_CH13_361_394.qxd
2/25/10
3:41 PM
Page 383
PREVENTION AND CONTROL 383
in the metformin-treated intervention group, and 4.8% in the lifestyle intervention group. Lifestyle intervention was more effective than medication and reduced diabetes incidence by 58%. A subsequent trial in India enrolled participants who were younger, leaner, and more insulin resistant, overall at greater risk of progression to diabetes, than the three preceding trial populations.79 Four groups studied were controls, lifestyle intervention alone, metformin alone, and lifestyle intervention plus metformin. Incidence of diabetes at 30 months was 55% in controls and ranged from 39.3 to 40.5% in the three other groups, with no advantage to combined intervention. Translation of the DPP intervention into practical management programs has received attention in a number of studies. A cost-effectiveness evaluation by use of the Archimedes model, a simulation model for assessment of clinical interventions, concluded that the lifestyle intervention of DPP:80, p 251 “. . . is likely to have important effects on the morbidity and mortality of diabetes and should be recommended to all high-risk people. The program used in the DPP study may be too expensive for health plans or a national program to implement. Less expensive methods are needed to achieve the degree of weight loss seen in the DPP.” A review of epidemiologic studies and clinical trials relevant to primary prevention of diabetes concluded:81, p 445 “Overall, a healthy diet, together with regular physical activity, maintenance of a healthy weight, moderate alcohol consumption, and avoidance of sedentary behaviors and smoking, could nearly eliminate T2DM. However, there is still a wide gap between what we know and what we practice in the field of public health; how to narrow that gap remains a major public health challenge.” Part of the gap is frequent lack of awareness of having prediabetes, a necessary step in the link to preventive measures.82 Data from the US National Health Interview Survey conducted in 2006 demonstrated that, although the estimated prevalence of prediabetes includes at least one-fourth of the adult population, only 4% of respondents reported awareness of having it. Even among those who were aware, taking action to prevent progression to diabetes was limited, according to self-report, to 68% for attempted weight loss, 60% for reduced dietary fat or calories, 55% for increased physical activity, and 42% for all three interventions. Diabetes Control Control of diabetes, once present, requires detection, evaluation, and long-term management. These are in principle clinical activities, although public health
support may be critical for development and adoption of relevant policies and practices, including assurance of guideline implementation. In addition, community resources can be instrumental in facilitating favorable long-term lifestyle patterns, self-management, and interaction with healthcare settings for recommended periodic follow-up and evaluation. Central to control of diabetes and the other factors contributing to cardiovascular complications, morbidity, and mortality is availability of recognized practice guidelines. Examples in the arena of diabetes are guidelines of the ADA and counterparts in Europe— the Task Force on Diabetes and Cardiovascular Diseases of the European Society of Cardiology (ESC) and the European Association for the Study of Diabetes (EASD) (Task Force).83,84 Highlights of the ADA Standards for Medical Care in Diabetes–2009 include aspects of screening, diabetes management, risk-factor assessment and control, monitoring for complications, and special considerations.83 Screening for prediabetes or T2DM in asymptomatic people is limited to people who are overweight or obese and have one or more additional risk factors or who are more than 45 years of age. Screening in children depends on family history, race/ethnicity, detection of associated conditions, and gestational history and should begin at the earlier of age 10 years or onset of puberty. IGT or IFG requires intervention to achieve 5–10% weight loss. Treatment goals for diabetes control are based on measurement of hemoglobin A1C (HgbA1C), a stable indicator of usual blood glucose levels; values 7% are considered a reasonable target for prevention of macrovascular complications. Medical Nutrition Therapy (MNT) is recommended in both prediabetes and diabetes and includes dietary change and physical activity. Diabetes Self-Management Education (DSME) is advised. Blood pressure, blood lipids, platelet function, and smoking are to be addressed. Monitoring for such complications as kidney, eye, and peripheral nerve complications, and attention to all of these issues in children with diabetes is also recommended. Special issues may arise in management of diabetes at school, at work, and in the event of hospitalization. The Task Force report, in addition to its recommendations, provides a valuable current overview of epidemiology and prevention of diabetes.84 Recommendations are presented in the context of applicable evidence, in more general form than those of the ADA. For example, assessment of risk of T2DM is recommended as part of routine health care, without qualification as to a target group or exclusions from screening. The link with cardiovascular risk is
46894_CH13_361_394.qxd
384
2/25/10
3:41 PM
Page 384
CHAPTER 13 DIABETES AND THE METABOLIC SYNDROME
indicated in the recommendation that persons with CVD and without diabetes should be screened with an oral glucose tolerance test. High risk of diabetes should trigger lifestyle counseling, with pharmacologic therapy if needed, and drugs can be used to delay onset of diabetes in those with IGT. Treatment recommendations to reduce cardiovascular risk include lifestyle and comprehensive management, glycemic control, and attention to specific guidelines for addressing dyslipidemia, high blood pressure, and management of coexisting cardiovascular disease. Discussion of health economics is included, with the observation that diabetes is associated with a considerable share of healthcare costs in Europe, due mainly to its cardiovascular complications. The US Preventive Services Task Force recommendations either are silent or find evidence insufficient regarding screening for diabetes except in adults with hypertension or hyperlipidemia.85 Recent trials have provided evidence regarding glucose control among persons with diabetes and its role in preventing cardiovascular complications.86 This remains an area of uncertainty as to benefits and risks due to increased frequency of adverse decreases in blood glucose (hypoglycemia) when treatment is directed to strictly preventing above-target levels (“tight control”). The main protocol of the Action to Control Cardiovascular Risk in Diabetes (ACCORD) Trial, for example, was terminated early because of excess deaths in the intensive therapy group. The UK Prospective Diabetes Study (UKPDS) observed a decrease in microvascular but not macrovascular complications of diabetes, with increased frequency of hypoglycemia.87 (These results differ importantly from those in intensive treatment of type 1 diabetes, where the Diabetes Control and Complications Trial (DCCT) found at the close of the trial and after extended follow-up that tight glucose control reduced incidence of major cardiovascular events by 57%.88) Metabolic Syndrome A 2005 AHA Scientific Statement addressed diagnosis and management of the metabolic syndrome, as defined by ATPIII, as a strategy for prevention of both cardiovascular disease and T2DM:89, p e289 It is recognized that the metabolic syndrome is a complex disorder, with no single factor as the cause. . . . The presence of the syndrome is associated with increased long-term risk for both ASCVD [atherosclerotic cardiovascular disease] and type 2 diabetes mellitus, and thus requires attention in clinical practice. Lifestyle interven-
tions deserve prime consideration for risk reduction across a lifetime; these interventions include weight control, increased physical activity, and a diet designed to reduce the risk of ASCVD. . . . Drug therapies should be used according to current recommendations for individual risk factors. At the present time, drug therapy is not recommended specifically to reduce risk for type 2 diabetes mellitus independent of treatments to prevent ASCVD. Community or Population-Wide Measures On the basis of available intervention studies either in communities or in healthcare systems, the Community Preventive Services Task Force found some populationlevel interventions sufficiently supported to warrant recommendation.90 These were disease management and case management in the healthcare system and diabetes self-management education in community settings (for adults with T2DM) and in the home (for children and adolescents with type 1 diabetes). Intervention on disease management was characterized as using “organized, proactive, multicomponent approaches to healthcare delivery for people with diabetes” with care that is “focused on, and integrated across, the spectrum of the disease and its complications, the prevention of comorbid conditions, and the relevant aspects of the delivery system, with the goal of improving both short- and long-term health or economic outcomes.”90, p 193 The Task Force added to its definition inclusion of all members of a population with diabetes and enumerated essential components of disease management to be considered. Criteria for effective programs were achievement of glycemic control as measured by glycosylated hemoglobin, and the percentage of providers who perform annual monitoring for control and potential complications of diabetes. Case management, by contrast, “identifies people at risk for excessive use of healthcare resources, poor coordination of healthcare services, or poor health outcomes and addresses their needs through improved planning, coordination, and provision of care.”90, p 197 Criteria for effectiveness and essential components were elaborated for case management as for disease management. Diabetes self-management education (DMSE) has longstanding recognition in diabetes care and refers to the process of teaching people to manage their diabetes. The Task Force required evidence of quantitative improvement in community levels of glycosylated hemoglobin as a criterion of effectiveness and found this to be met only in the specific settings
46894_CH13_361_394.qxd
2/25/10
3:41 PM
Page 385
PREVENTION AND CONTROL 385
that are recommended. This was not the case for DSME in summer camps or at worksites. The focus of these measures is clearly on persons with diabetes, as individuals or as the affected group within a community. However, the interventions have a community or population-wide impact in establishing the policies or systems by which disease management or case management become established and are supported in a community. Other interventions at the community or population level that would contribute to diabetes prevention and control are those addressing dietary imbalance, physical inactivity, obesity, adverse blood lipid profile, high blood pressure, and smoking. These are discussed in other chapters as interventions for cardiovascular disease prevention. Diabetes prevention and control is an important additional benefit of effective interventions in these several areas. Global Strategies Prevention and control of diabetes mellitus were first addressed by the World Health Assembly in 1989 in a resolution calling on Member States to assess the problem, implement population-based measures for prevention and control, share training opportunities, and establish model community programs.91 The Director-General was requested to support these activities, develop collaborative arrangements with the IDF and other agencies, and engage the WHO collaborating centers on diabetes in these efforts. A Study Group convened in 1992 was a part of the response to this resolution.11 The concepts of prevention outlined in the Study Group report addressed both population-wide and high-risk strategies. On the basis that risks of morbid events in diabetes are low except for those with blood glucose values near the upper extreme, it was judged that the population strategy may be inappropriate in many populations and that high-risk approaches for those with familial risks or clusters of other risk characteristics may be more cost-effective. Assessment of the distribution of diabetes and its determinants in a given target population was recommended in establishing the best approach or combination of approaches for that population. Against this background, the Study Group report noted the importance of IGT as a target for intervention, given its intermediate position in the distribution between normal and diabetic categories. This is analogous, as noted earlier, to the identification of borderline high blood pressure or cholesterol concentration as especially warranting preventive measures. The report suggests, however, that IGT is
an entity in itself and not simply part of the continuum of the blood glucose distribution. A focus on the “putative risk factors”—physical inactivity, nutritional factors, and obesity—was proposed, with the conclusion that “there is general agreement that dietary modification and exercise should serve as the cornerstones in the prevention of diabetes and the treatment of people with the disease.”11, p 29 It was noted that malnutrition in utero leading to low birth weight may also predispose to diabetes in adult life, suggesting preventive measures in maternal health before and during pregnancy. High-risk individuals were characterized as those with strong family history of NIDDM; persons changing through migration or otherwise to Westernized, urban, or sedentary lifestyles; women with histories of gestational diabetes or IGT; and persons with other components of the metabolic (or insulin resistance) syndrome. Measures to be taken for such high-risk persons were control of obesity; maintenance of lowfat, high-fiber dietary habits; increased physical activity; and avoidance of specific drugs that may impair glucose metabolism. Guidelines and programs to implement these recommendations were addressed in publications from WHO.91,92 These reports provide references to a number of detailed manuals and materials valuable in the support of such programs. It is noteworthy that recommendations generally disfavor screening for purposes of case detection, even among family members of known cases, except in high-risk populations.93 The perspective of low- and middle-income countries is fundamental to the work of the Disease Control Priorities Project in addressing the broad range of major global health burdens. Accordingly, discussion of diabetes centers on cost-effectiveness in developing country settings of interventions for preventing and treating diabetes and its complications (Table 13-8).55 Three levels of intervention are distinguished: level 1, cost saving and feasible; level 2, cost saving but not entirely feasible or costing less than US$1500 per QALY (quality of life-adjusted years saved) and “at least moderately feasible”; and level 3, estimated cost from US$1640 to US$8550 per QALY and judged less feasible than level 1 or 2 interventions. As described in the table, the three levels of intervention are alike in identifying personal interventions that are heavily reliant on clinical care. Distinct from these are the “Essential background intervention,” diabetes education, and the “Other promising intervention,” a “polypill.” Diabetes education refers once again to the concept of self-management discussed previously. It was considered that this intervention may be cost-effective, and diabetes education
46894_CH13_361_394.qxd
386
2/25/10
3:41 PM
Page 386
CHAPTER 13 DIABETES AND THE METABOLIC SYNDROME
Table 13-8 Intervention
Key Cost-Effective Interventions for Preventing and Treating Diabetes and Its Complications Description Applicable Population Major Effect
Level 1a • Glycemic control in people with poor control
Insulin, oral glucose-lowering agents, diet and exercise
People with diabetes, all ages, HbA1c greater than 9 percent
Reduction in microvascular disease
• Blood pressure control
Blood pressure control medications
People with diabetes, hypertensive, all ages
Reduction in macrovascular disease, microvascular disease, and mortality
• Foot care
Patient and provider education, foot examination, foot hygiene, and appropriate footwear
People with diabetes, middleaged or older
Reduction in serious foot diseases and amputations
Patient self-management
Women with diabetes who plan to become pregnant
Reduction in HbA1c level and hospital expenses of the mother and baby
• Lifestyle intervention to prevent diabetes
Behavioral change, including diet and physical activity, to reduce bodyweight
People who are at high risk (for example, prediabetes for type 2 diabetes)
Reduction in type 2 diabetes incidence by 58 percent
• Influenza vaccination
Vaccination
Elderly people with diabetes
Reduction in hospitalizations, respiratory conditions, and mortality
• Detection and treatment of eye diseases
Eye examination to screen for and treat eye diseases
People with diabetes, middleaged or older
Reduction in serious vision loss
• ACE inhibitors
Angiotensin-converting enzyme medication
People with diabetes
Reduction in nephropathy, cardiovascular disease, and death
• Smoking cessation
Physician counseling and nicotine replacement therapy
People with diabetes, all ages, smokers
Increase in quitting rate and reduction in cardiovascular disease
Metformin medication
People who are at high risk (for example, prediabetes for type 2 diabetes)
Reduction in type 2 diabetes incidence by 33 percent
• Intensive glucose control
Insulin, oral glucose-lowering agents, or both
Diabetes, all ages, with HbA1c less than 9 percent
Reduction in microvascular disease
• Lipid control
Cholesterol-lowering medication
Diabetes, all ages, with high cholesterol
Reduction in cardiovascular disease events and mortality
• Screening for microalbuminuria
Screening for microalbuminuria and treating those who test positive
Diabetes, all ages
Reduction in kidney diseases
• Screening for undiagnosed diabetes
Screening for undiagnosed diabetes and treating those who test positive
People who are at high risk for type 2 diabetes
Reduction in microvascular disease
Essential background interventiond Diabetes education Patient self-management
Diabetes, all ages
Reduction in HbA1c level and better compliance with lifestyle changes
Other promising interventione Polypill
Diabetes, all ages
Reduction in cardiovascular disease
Level 2b
• Preconception care for women of reproductive age
Level 3c • Metformin therapy for preventing diabetes
Hypothetical pill combining low doses of antihypertensive medication, aspirin, statin, and folate
a
Level 1 interventions are cost saving and highly feasible. Level 2 interventions are cost saving or cost less than US$1,500 per quality-adjusted life year but pose feasibility challenges. Level 3 interventions cost between US$1,640 and US$8,550 per quality-adjusted life year and pose significant feasibility challenges. d Diabetes education is the backbone on which many diabetes interventions depend, but empirical data on the effectiveness of diabetes education on outcomes and on the precise components of diabetes education are still lacking. e An intervention that appears promising but needs further research to document its effectiveness and/or safety. The polypill is only a theoretical concept at this time and is not available for implementation. b c
Source: Authors. From Disease Control Priorities in Developing Countries, Second Edition, edited by DT Jamison et al., Copyright © 2006. Courtesy of The International Bank for Reconstruction and Development/ The World Bank.
46894_CH13_361_394.qxd
2/25/10
3:41 PM
Page 387
CURRENT ISSUES 387
was recommended as “a high-priority intervention for all developing regions.”8, p 598 A combination tablet comprising three half-dose antihypertensive medications with aspirin, a statin, and folic acid remained hypothetical, not available for use, and not yet evaluated as to its benefits and side effects; it was thought to be cost-effective if available at US$1.28 or less per tablet. Analysis of effectiveness and cost effectiveness of interventions for developing countries was also presented, including preventing diabetes, screening for diabetes in the general population, and treating diabetes and its complications (Table 13-9). These interventions ranged from cost saving, for certain measures to prevent complications, to US$73,500/ QALY for general population screening. Prevention of diabetes through use of metformin was found to cost US$31,200/QALY, whereas lifestyle intervention would be considerably more cost-effective at US$1100 per QALY. In order for these interventions, singly or in combination, to become global strategies, their advocacy and adoption as policy would be required by many national governments or by regional or global health authorities. A research agenda was proposed that would be expected to clarify and support these interventions.
CURRENT ISSUES Natural History How early in the development of diabetes the risk of cardiovascular disease begins to increase is an important question. The emergence of cardiovascular risk factors long in advance of the diagnosis of diabetes was reported in a study of Mexican Americans in San Antonio, Texas;94 the clustering of risk factors in children by age 9–10 years in Mexican American versus non-Hispanic White children was noted previously;39 and detection of a genetic marker for insulin regulation and its relation to risk factors in 5-year-old children was reported from the Bogalusa Heart Study.95 These observations suggest the value of further investigation of early precursors of frank diabetes or IGT and their relation to cardiovascular risk. Better information is needed in many populations about blood glucose distributions, related risk factors for cardiovascular diseases, and needs for intervention at both population and high-risk levels, beginning in childhood and adolescence. Screening for Prediabetes and T2DM Methodologic issues remain, at the basic level of determining prevalence of diabetes and IGT. Various
criteria and sampling strategies have been advocated for prevalence surveys, and detailed protocols have been proposed.11,91,92,96,97 These proposals are relevant to population screening as well. Population screening, generally considered unwarranted on the basis of expected yield and benefit, may warrant reassessment in view of recognition of prediabetes and demonstrated prevention of T2DM by lifestyle intervention. Both developments add importance to recognizing the much greater number of individuals at increased risk of diabetes or cardiovascular disease than those detected as having frank diabetes. Diabetes, Insulin Resistance, Metabolic Syndrome, and Cardiovascular Disease Numerous factors link atherosclerosis with T2DM, and the insulin resistance syndrome or the metabolic syndrome. a more integrated view of these relations in terms of both natural history and intervention strategies is perhaps emerging. Grundy’s essay, “Metabolic Syndrome: Connecting and Reconciling Cardiovascular and Diabetes Worlds,” provides a promising perspective.4 Prevention and Control The case for prevention and control of diabetes depends on its prevalence and actual contribution to disability and mortality and, in turn, costs and costbenefit balances.98,99 The new concept of diabetes prevention that includes its primary prevention through lifestyle interventions changes the argument fundamentally. How best to develop comprehensive, long-range strategies with maximum public health impact has been addressed recently through system dynamics modeling, which offers valuable insights into policy alternatives. For example, it has been possible through this approach to simulate100, p 493 “. . . strategies that represent a mix of increased diabetes management and reduced obesity prevalence. Comparing a mixed strategy to one that focuses entirely on diabetes management, the [policy] experiments suggest that the focused diabetes management scenario may quickly reduce diabetes-related complications and deaths but is less effective in the long term than the mixed strategy.” From this perspective, prevention and control of diabetes calls for incorporating and reinforcing policies already well established and advocated in other areas––improving nutrition, increasing physical activity, achieving and maintaining healthy weight, and preventing or discontinuing tobacco use, in order to prevent chronic diseases broadly, including diabetes and cardiovascular disease.
46894_CH13_361_394.qxd
388
2/25/10
3:41 PM
Page 388
CHAPTER 13 DIABETES AND THE METABOLIC SYNDROME
Table 13-9
Strategy
Effectiveness and Cost-Effectiveness of Interventions for Preventing and Treating Diabetes in Developed Countries Quality of Cost-Effectiveness Ratio Benefit Evidencea (US$/QALY)b
Preventing diabetes Lifestyle interventions for preventing type 2 diabetes
Reduction of 35–58 percent in incidence among people at high risk
I
1,100 (Diabetes Prevention Program Research Group forthcoming)
Metformin for preventing type 2 diabetes
Reduction of 25–31 percent in incidence among people at high risk
I
31,200 (Diabetes Prevention Program Research Group forthcoming)
Screening for diabetes Screening for type 2 diabetes in general population
Reduction of 25 percent in microvascular disease
III
73,500 (CDC Diabetes CostEffectiveness Study Group 1998)
Reduction of 30 percent in microvascular disease per 1 percent drop in HbA1c
I
Cost saving (CDC Diabetes CostEffectiveness Study Group 1998)
Glycemic control in people with HbA1c greater than 8 percent
Reduction of 30 percent in microvascular disease per 1 percent drop in HbA1c
I
34,400 (CDC Diabetes CostEffectiveness Study Group 1998; Klonoff and Schwartz 2000)
Blood pressure control in people whose pressure is higher than 160/95 mmHg
Reduction of 35 percent in macrovascular and microvascular disease per 10 mmHg drop in blood pressure
I
Cost saving (CDC Diabetes CostEffectiveness Study Group 1998)
Cholesterol control in people with total cholesterol greater than 200 milligrams/deciliter
Reduction of 25–55 percent in coronary heart disease events; 43 percent fall in death rate
II-1
63,200 (CDC Diabetes CostEffectiveness Study Group 1998)
Smoking cessation with recommended guidelines
16 percent quitting rate
I
12,500 (CDC Diabetes CostEffectiveness Study Group 1998)
Annual screening for microalbuminuria
Reduction of 50 percent in nephropathy using ACE inhibitors for identified cases
III
47,400 (Klonoff and Schwartz 2000)
Annual eye examinations
Reduction of 60–70 percent in serious vision loss
I
6,000 (Klonoff and Schwartz 2000; Vijan, Hofer, and Hayward 2000)
Foot care in people with high risk of ulcers
Reduction of 50–60 percent in serious foot disease
I
Cost saving (Ragnarson and Apelqvist 2001)
Aspirin use
Reduction of 28 percent in myocardial infarctions, reduction of 18 percent in cardiovascular disease
I
Not available
ACE inhibitor use in all people with diabetes
Reduction of 42 percent in nephropathy; 22 percent drop in cardiovascular disease
I
8,800 (Golan, Birkmeyer, and Welch 1999)
Influenza vaccinations among the elderly for type 2 diabetes
Reduction of 32 percent in hospitalizations; 64 percent drop in respiratory conditions and death
II-2
3,100 (Sorensen and others 2004)
Preconception care for women of reproductive age
Reduction of 30 percent in hospital charges and 25 percent in hospital days
II-2
Cost saving (Klonoff and Schwartz 2000)
Treating diabetes and its complications Glycemic control in people with HbA1c greater than 9 percent
a
I indicates evidence from at least one randomized, controlled trial; II-1 indicates evidence from a well-designed, controlled trial without randomization; II-2 indicates evidence from cohort or case control studies; and III indicates opinions of respected authorities (US Preventive Services Task Force 1996). b We adjusted cost-effectiveness ratios to 2002 US dollars using the consumer price index for medical care. In cases in which multiple studies evaluated the cost-effectiveness of an intervention, we report the median cost-effectiveness ratio. Note: mm Hg millimeters of mercury; QALY quality-adjusted life year. Source: Authors. From Disease Control Priorities in Developing Countries, Second Edition, edited by DT Jamison et al., Copyright © 2006. Courtesy of The International Bank for Reconstruction and Development/The World Bank.
46894_CH13_361_394.qxd
2/25/10
3:41 PM
Page 389
REFERENCES 389
REFERENCES 1. West KM. Epidemiologic approaches in the study and control of diabetes. In: West KM, ed. Epidemiology of Diabetes and Its Vascular Lesions. New York: Elsevier North-Holland Inc.; 1978;1:1–13. 2. Grant PJ. Diabetes and cardiology: The common soil comes of age. Diabetes Vasc Dis Res. 2004;1:9. 3. National Cholesterol Education Program Expert Panel on Detection, Evaluation and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). NIH Publication No. 02-5215. Bethesda, MD: National Cholesterol Education Program, National Heart, Lung, and Blood Institute, National Institutes of Health; September 2002 4. Grundy SM. Metabolic syndrome: Connecting and reconciling cardiovascular and diabetes worlds. J Am Coll Cardiol. 2006;47: 1093–1100. 5. Ford ES, Ajani UA, Croft JB, et al. Explaining the decrease in U.S. deaths from coronary disease, 1980-2000. N Engl J Med. 2007;356: 33–43. 6. American Diabetes Association. T2DM in children and adolescents. Diabetes Care. 2000;23: 381–389. 7. Tuomilehto J, Lindström J, Eriksson JG, et al. for the Finnish Diabetes Prevention Study Group. Prevention of T2DM mellitus by changes in lifestyle among subjects with impaired glucose tolerance. N Engl J Med. 2001;344:1343–1350. 8. Venkat Narayan KM, Zhang P, Kanaya AM, et al. Diabetes: The pandemic and potential solutions. In: Jamison DT, Breman JG, Measham AR, et al., eds. Disease Control Priorities in Developing Countries. 2nd ed. Washington, DC: International Bank for Reconstruction and Development/The World Bank; 2006:591–603.
9. Harris MI, Zimmet P. Classification of diabetes mellitus and other categories of glucose intolerance. In: Alberti KGMM, Defronzo RA, Keen H, Zimmet P, eds. International Textbook of Diabetes Mellitus. Chichester, England: John Wiley & Sons; 1992:3–18. 10. Unwin N, Shaw J, Zimmet P, Alberti KGMM. Impaired glucose tolerance and impaired fasting glycaemia: The current status on definition and intervention. Diabetes. 2002;19:708–723. 11. Report of a WHO Study Group. Prevention of Diabetes Mellitus. WHO Technical Report Series 844. Geneva (Switzerland): World Health Organization; 1994. 12. World Health Organization. International Statistical Classification of Diseases and Related Health Problems, Tenth Revision. Geneva, Switzerland: World Health Organization; 1992:l. 13. Reaven GM. Banting Lecture 1988. Role of insulin resistance in human disease. Diabetes. 1988;37:1595–1607. 14. Hu G, Qiao Q, Tuomilehto J. Metabolic syndrome, diabetes, and coronary heart disease. In: Marmot M, Elliott P, eds. Coronary Heart Disease Epidemiology: From Aetiology to Public Health. 2nd ed. Oxford (England): Oxford University Press; 2005:311–330. 15. Kahn R, Buse J, Ferrannini E, Stern M. The metabolic syndrome: Time for a critical appraisal. Joint statement from the American Diabetes Association and the European Association for the Study of Diabetes. Diabetes Care. 2005;28:2289–2304. 16. Keen H. Introduction. In: Alberti KGMM, DeFronzo RA, Keen H, Zimmet P, eds. International Textbook of Diabetes Mellitus. Chichester, England: John Wiley & Sons; 1992;1:xxvii–xxix. 17. Steinberger J, Daniels SR, Eckel RH, et al. Progress and challenges in metabolic syndrome in children and adolescents. A Scientific Statement from the American Heart Association Atherosclerosis, Hypertension, and Obesity in the Young Committee of the
46894_CH13_361_394.qxd
390
2/25/10
3:41 PM
Page 390
CHAPTER 13 DIABETES AND THE METABOLIC SYNDROME
Council on Cardiovascular Disease in the Young; Council on Cardiovascular Nursing; and Council on Nutrition, Physical Activity, and Metabolism. Circulation. 2009;119: 628–647. 18. Johnston DG, Alberti KGMM, Godsland IF, Pierce M, Shepperd S. Screening for T2DM. In: Marmot M, Elliott P, eds. Coronary Heart Disease Epidemiology: From Aetiology to Public Health. 2nd ed. Oxford (England): Oxford University Press; 2005:714–750. 19. DeFronzo RA, Bonadonna RC, Ferrannini E. Pathogenesis of NIDDM: a precarious balance between insulin action and insulin secretion. In: Alberti KGMM, DeFronzo RA, Keen H, Zimmet P, eds. International Textbook of Diabetes Mellitus. Chichester, England: John Wiley & Sons; 1992;1:569–633. 20. Neel JV. Diabetes mellitus: a “thrifty” genotype rendered detrimental by “progress”? Am J Hum Genet. 1962;14:353–362. 21. Raffel LJ, Shohat T, Rotter JI. Diabetes and insulin resistance. In: Goldbourt U, de Faire U, Berg K, eds. Genetic Factors in Coronary Heart Disease. Dordrecht, the Netherlands: Kluwer Academic Publishers; 1994:203–215. 22. Vadheim CM, Rotter JI. Genetics of diabetes mellitus. In: Alberti KGMM, DeFronzo RA, Keen H, Zimmet P, eds. International Textbook of Diabetes Mellitus. Chichester, England: John Wiley & Sons; 1992;1:31–98. 23. Hanis CL, Boerqinkle E, Chakraborty R. A genome-wide search for human non-insulindependent (type 2) diabetes genes reveals a major susceptibility locus on chromosome 2. Nat Genet. 1996;13:161–166. 24. McCarthy MI, Zeggini E. Genome-wide association studies in T2DM. Curr Diab Rep. 2009;9:164–171. 25. Cusi K. Lessons learned from studying families genetically predisposed to T2DM mellitus. Curr Diab Rep. 2009;9:200–207. 26. Edwards KL, Hutter CM, Wan JY, Kim H, Monks SA. Genome-wide linkage scan for the
metabolic syndrome: the GENNID study. Obesity (Silver Spring) 2008;16:1596–1601. 27. Perry IJ, Wannamethee SG, Walker MK, et al. Prospective study of risk factors for development of non-insulin dependent diabetes in middle-aged British men. Br Med J. 1995;310: 560–564. 28. Lipton RB, Liao Y, Cao G, et al. Determinants of incident non-insulin-dependent diabetes mellitus among Blacks and Whites in a national sample: the NHANES I Epidemiologic Follow-Up Study. Am J Epidemiol. 1993; 138:826–839. 29. American Diabetes Association, National Institute of Diabetes and Digestive and Kidney Diseases. Prevention or delay of T2DM. Diabetes Care. 2004;27(suppl 1): S47–S54. 30. Dhingra R, Sullivan L, Jacques PF, et al. Soft drink consumption and risk of developing cardiometabolic risk factors and the metabolic syndrome in middle-aged adults in the community. Circulation. 2007;116:480–488. 31. Stout RW. Insulin and atheroma: 20-yr perspective. Diabetes Care. 1990;13:631–654. 32. Howard G, O’Leary DH, Zaccaro D, Haffner S, et al. for the IRAS Investigators. Insulin sensitivity and atherosclerosis. Circulation. 1996; 93:1809–1817. 33. Nigro J, Osman N, Dart AM, Little PJ. Insulin resistance and atherosclerosis. Endocrine Rev. 2006;27:242–259. 34. Grundy SM, Howard B, Smith Jr S, Eckel R, Redberg R, Bonow RO. Prevention Conference VI. Diabetes and Cardiovascular Disease. Executive Summary. Conference proceeding for healthcare professionals from a special writing group of the American Heart Association. Circulation. 2002;105:2231–2239. 35. Eckel RH, Wassef M, Sobel B, et al. Prevention Conference VI. Diabetes and Cardiovascular Disease. Writing Group II: Pathogenesis of atherosclerosis in diabetes. Circulation. 2002; 105:e138–e143.
46894_CH13_361_394.qxd
2/25/10
3:41 PM
Page 391
REFERENCES 391
36. Langenberg C, Bergstrom J, Scheidt-Nave C, Pfeilschifter J, Barrett-Connor E. Cardiovascular death and the metabolic syndrome. Role of adiposity-signaling hormones and inflammatory markers. Diabetes Care 2006;29: 1363–1369.
44. Lloyd-Jones D, Adams R, Carnethon M et al. Heart disease and stroke statistics 2009 update. A report from the American Heart Association Statistics Committee and Stroke Statistics Committee. doi: 10.1161/CIRCULATIONAHA. 108.191261. http://circ.ahajournals.org. Accessed December 18, 2008.
37. Orchard TJ. Diabetes. In: Pearson TA, Criqui MH, Luepker RV, Oberman A, Winston M, eds. Primer in Preventive Cardiology. Dallas, TX: American Heart Association; 1994: 159–171.
45. Centers for Disease Control and Prevention. State-specific incidence of diabetes among adults—participating states, 1995–1997 and 2005–2007. MMWR. 2008;57:1169–1173.
38. Bao W, Srinivasan SR, Berenson GS. Persistent elevation of plasma insulin levels is associated with increased cardiovascular risk in children and young adults: the Bogalusa Heart Study. Circulation. 1996;93:54–59.
46. Centers for Disease Control and Prevention. Diagnosed diabetes among American Indians and Alaska Natives aged 35 years—United States, 1994–2004. MMWR. 2006;55: 1201–1203.
39. Steffen-Batey L, Goff DC Jr, Tortolero SR, et al. Summary measures of the insulin resistance syndrome are adverse among MexicanAmericans versus non-Hispanic White children: the Corpus Christi Heart Study. Circulation. 1997;96:4319–4325.
47. American Diabetes Association. T2DM in children and adolescents. Diabetes Care. 2000;23: 381–389.
40. Chen W, Srinivasan SR, Li S, Xu J, Berenson GS. Clustering of long-term trends in metabolic syndrome variables from childhood to adulthood in blacks and whites. The Bogalusa Heart Study. Am J Epidemiol. 2007;166: 527–533. 41. Lambert M, Paradis G, O’Loughlin J, Delvin EE, Hanley JA, Levy E. Insulin resistance syndrome in a representative sample of children and adolescents from Quebec, Canada. Int J Obesity. 2004;28:833–841. 42. US Department of Health and Human Services. Health, United States, 2008 with Special Feature on the Health of Young Adults. Washington DC: US Department of Health and Human Services. Centers for Disease Control and Prevention. National Center for Health Statistics; 2008. 43. Centers for Disease Control and Prevention. National diabetes fact sheet: general information and national estimates on diabetes in the United States, 2007. Atlanta, GA: US Department of Health and Human Services, Centers for Disease Control and Prevention, 2008.
48. US Department of Health and Human Services. Healthy People 2010 Midcourse Review: Washington, DC: US Government Printing Office; December 2006. 49. Ford ES, Giles WH, Dietz WH. Prevalence of the metabolic syndrome among US adults. Findings from the Third National Health and Nutrition Examination Survey. JAMA. 2002; 287:356–359. 50. Cook S, Weitzman M, Auinger P, Nguyen M, Dietz WH. Prevalence of a metabolic syndrome phenotype in adolescents. Findings from the Third National Health and Nutrition Examination Survey, 1988–1994. Arch Pediatr Adolesc Med. 2003;157:821–827. 51. King H. Global estimates for prevalence of diabetes mellitus and impaired glucose tolerance in adults. Diabetes Care. 1993;16: 157–177. 52. Bennett PH, Bogardus C, Tuomilehto J, Zimmet P. Epidemiology and natural history of NIDDM: non-obese and obese. In: Alberti KGMM, DeFronzo RA, Keen H, Zimmet P, eds. International Textbook of Diabetes Mellitus. Chichester, England: John Wiley & Sons; 1992;1:147–176.
46894_CH13_361_394.qxd
392
2/25/10
3:41 PM
Page 392
CHAPTER 13 DIABETES AND THE METABOLIC SYNDROME
53. Mather HM, Keen H. The Southall Diabetes Survey: prevalence of known diabetes in Asians and Europeans. Br Med J. 1985;291: 1081–1084. 54. Keys A. Seven Countries: A Multivariate Analysis of Death and Coronary Heart Disease. Cambridge, MA: Harvard University Press; 1980. 55. The International Collaborative Group. Joint Discussion. Glycemia and prevalence of ECG abnormalities. J Chronic Dis. 1979;32: 829–837. 56. The DECODE Study Group, on behalf of the European Diabetes Epidemiology Group. Glucose tolerance and cardiovascular mortality. Comparison of fasting and 2-hour diagnostic criteria. Arch Intern Med. 2001;161: 397–404. 57. Brunner EJ, Shipley MJ, Witte DR, Fuller JH, Marmot MG. Relation between blood glucose and coronary mortality over 33 years in the Whitehall Study. Diabetes Care. 2006;29: 26–31.
disease mortality among patients with T2DM mellitus. Arch Intern Med. 2002;162: 1737–1745. 62. The DECODE Study Group. Gender difference in all-cause and cardiovascular mortality related to hyperglycaemia and newly-diagnosed diabetes. Diabetologia. 2003;46:608–617. 63. Will JC, Casper M. The contribution of diabetes to early deaths from ischemic heart disease: US gender and racial comparisons. Am J Public Health. 1996; 86:576–579. 64. Orlander PR, Goff DC, Morrissey M, et al. The relation of diabetes to the severity of acute myocardial infarction and post-myocardial infarction survival in Mexican-Americans and non-Hispanic Whites: the Corpus Christi Heart Project. Diabetes. 1994;43:897–902. 65. Haffner SM, Lehto S, Rönnemaa, Pyörälä K, Laakso M. Mortality from coronary heart disease in subjects with T2DM and in nondiabetic subjects with and without prior myocardial infarction. N Engl J Med. 1998;339:229–234.
58. Deedwania P, Kosiborod M, Barrett E, et al. Hyperglycemia and acute coronary syndrome. A Scientific Statement from the American Heart Association Diabetes Committee of the Council on Nutrition, Physical Activity, and Metabolism. Circulation. 2008;117: 1610–1619.
66. Vaccaro O, Eberly LE, Neaton JD, Yang L, Riccardi G, Stamler J for the Multiple Risk Factor Intervention Trial (MRFIT) Research Group. Impact of diabetes and previous myocardial infarction on long-term survival. 25year mortality follow-up of primary screenees of the Multiple Risk Factor Intervention Trial. Arch Intern Med. 2004;164:1438–1443.
59. Wingard DL, Barrett-Connor E. Heart disease and diabetes. In: National Diabetes Data Group. Diabetes in America. 2nd ed. NIH publication 95-1468. Bethesda, MD: National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health; 1995: 429–448.
67. Mukamal KJ, Nesto RW, Cohen MC, et al. Impact of diabetes on long-term survival after acute myocardial infarction. Comparability of risk with prior myocardial infarction. Diabetes Care. 2001;24:1422–1427.
60. McKeigue PM, Keen H. Diabetes, insulin, ethnicity, and coronary heart disease. In: Marmot M, Elliott P, eds. Coronary Heart Disease Epidemiology: From Aetiology to Public Health. Oxford (England): Oxford University Press; 1992:217–232. 61. Kanaya AM, Grady D, Barrett-Connor E. Explaining the sex difference in coronary heart
68. Whiteley L, Padmanbhan S, Hole D, Isles C. Should diabetes be considered a coronary heart disease equivalent? Results from 25 years of follow-up in the Renfrew and Paisley Survey. Diabetes Care. 2005;38:1588–1593. 69. Fox CS, Sullivan L, D’Agostino RB, Wilson PWF. The significant effect of diabetes duration on coronary heart disease mortality. Diabetes Care. 2004;27:704–708.
46894_CH13_361_394.qxd
2/25/10
3:41 PM
Page 393
REFERENCES 393
70. Alexander CM, Landsman PB, Teutsch SM, Haffner SM. NCEP-defined metabolic syndrome, diabetes, and prevalence of coronary heart disease among NHANES III participants age 50 years and older. Diabetes. 2003;52: 1210–1214. 71. Wannamethee SG, Shaper GA, Lennon L, Morris RW. Metabolic syndrome vs Framingham risk score for prediction of coronary heart disease, stroke, and T2DM mellitus. Arch Intern Med. 2005;165:2644–2650. 72. Pyörälä K, Laakso M, Uusitupa M. Diabetes and atherosclerosis: an epidemiologic view. Diabetes Meta Rev. 1987;3:463–524. 73. Sacco RL, Benjamin EJ, Broderick JP, Dyken M, et al. Risk factors. Stroke. 1997;28: 1507–1517. 74. Diabetes Drafting Group. Prevalence of small vessel and large vessel disease in diabetic patients from 14 centres: the World Health Organization Multinational Study of Vascular Disease in Diabetics. Diabetologia. 1985; 28:615–640. 75. Mathers CD, Lopez AD, Murray CJL. The burden of disease and mortality by condition: Data, methods, and results for 2001. In: Lopez AD, et al., eds. Global Burden of Disease and Risk Factors. The International Bank for Reconstruction and Development/The World Bank, Washington, DC; 2006:45–240. 76. Yach D, Stuckler D, Brownell KD. Epidemiologic and economic consequences of the global epidemics of obesity and diabetes. Nature Med. 2006;1:62–66. 77. Pan X-R, Li G-W, Hu J-X, et al. Effects of diet and exercise in preventing NIDDM in people with imparied glucose tolerance. Diabetes Care. 1997;20:537–544. 78. Diabetes Prevention Program Research Group. Reduction in the incidence of T2DM with lifestyle intervention or metformin. N Engl J Med. 2002;346:393–403. 79. Ramachandran A, Snehalatha C, Mary S, Mukesh B, Bhaskar AD, Vijay V, Indian
Diabetes Prevention Programme (IDPP). The Indian Diabetes Preventino Programme shows that lifestyle modification and metformin prevent T2DM in Asian Indian subjects with impaired glucose tolerance (IDDP-1). Diabetologia. 2006;49:289–297. 80. Eddy DM, Schlessinger L, Kahn R. Clinical outcomes and cost-effectiveness of strategies for managing people at high risk for diabetes. Ann Intern Med. 2005;143:251–264. 81. Schulze MB, Hu FB. Primary prevention of diabetes: What can be done and how much can be prevented? Annu Rev Public Health. 2005;26: 445–467. 82. Centers for Disease Control and Prevention. Self-reported prediabetes and risk-reduction activities – United States, 2006. MMWR. 2008;57:1203–1205. 83. American Diabetes Association. Executive summary: Standards of medical care in diabetes – 2009. Diabetes Care. 2009;32 (suppl 1):S6–S12. 84. The Task Force on Diabetes and Cardiovascular Diseases of the European Society of Cardiology (ESC) and of the European Association for the Study of Diabetes (EASD). Guidelines on diabetes, pre-diabetes, and cardiovascular diseases: Full text. J Europ Soc Cardiol. 2007;9(Suppl C):C1–C74. 85. US Preventive Services Task Force. The Guide to Clinical Preventive Services 2006. Recommendations of the US Preventive Services Task Force. Washington, DC: Agency for Healthcare Research and Quality; 2006. www.ahrq.gov/clinic/uspstf/uspstbac.htm. Accessed October 14, 2007. 86. Liebson PR. Diabetes control and cardiovascular risk: ACCORD, ADVACNE, AVOID, and SANDS. Prev Cardiol. 2008;11:230–236. 87. UK Prospective Diabetes Study (UKPDS) Group. Intensive blood-glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with T2DM (UKPDS 33). Lancet. 1998;352:837–853.
46894_CH13_361_394.qxd
394
2/25/10
3:41 PM
Page 394
CHAPTER 13 DIABETES AND THE METABOLIC SYNDROME
88. The Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions and Complications (DCCT/EDIC) Study Research Group. Intensive diabetes treatment and cardiovascular disease in patients with type 1 diabetes. N Engl J Med. 2005;353: 2643–2653. 89. Grundy SM, Cleeman JI, Daniels SR, et al. Diagnosis and management of the metabolic syndrome. An American Heart Association/ National Heart, Lung, and Blood Institute Scientific Statement. Executive Summary. Circulation. 2005;112:e285–e290. 90. Zaza S, Briss PA, Harrris K eds. The Guide to Community Preventive Services. What Works to Promote Health? Oxford, England: Oxford University Press; 2005. 91. Reiber GE, King H. Guidelines for the Development of a National Programme for Diabetes Mellitus. Geneva (Switzerland): Division of Noncommunicable Diseases and Health Technology, World Health Organization; 1991. 92. King H, Gruber W, Lander T, eds. Implementing National Diabetes Programmes: Report of a WHO Meeting. Geneva Switzerland: Division of Noncommunicable Diseases, World Health Organization; 1995. 93. Tuomilehto J, Tuomilehto-Wolf E, Zimmet P, et al. Primary prevention of diabetes mellitus. In: Alberti KGMM, DeFronzo RA, Keen H, Zimmet P, eds. International Textbook of Diabetes Mellitus. Chichester, England: John Wiley & Sons; 1992;2:1655–1673.
94. Haffner ST, Stern MP, Hazuda HP, et al. Cardiovascular risk factors in confirmed prediabetic individuals: does the clock for coronary heart disease start ticking before the onset of clinical diabetes? JAMA. 1990;263: 2893–2898. 95. Amos CI, Cohen JC, Srinivasan SR, et al. Polymorphism in the 5-flanking region of the insulin gene and its potential relation to cardiovascular disease risk: observations in a biracial community: the Bogalusa Heart Study. Atherosclerosis. 1989;79:51–57. 96. Finch CF, Zimmet PZ, Alberti KGMM. Determining diabetes prevalence: a rational basis for the use of fasting plasma glucose concentrations? Diabetic Med. 1990;7: 603–610. 97. LaPorte RE, McCarty D, Bruno G. Counting diabetes in the next millennium. Diabetes Care. 1993;16:528–534. 98. Huse DM, Oster G, Killen AR, et al. The economic costs of non-insulin-dependent diabetes mellitus. JAMA. 1989;262:2708–2713. 99. Vaughan JP, Gilson L, Mills A. In: Jamison DT, Mosley WH, Measham AR, Bobadilla JL, eds. Disease Control Priorities in Developing Countries. Oxford (England): Oxford University Press; 1993:561–576. 100. Jones AP, Homer JB, Murphy DL, Essien JDK, Milstein B, Seville DA. Understanding diabetes population dynamics through simulation modeling and experimentation. Am J Public Health. 2006;96:488–494.
46894_CH14_395_430.qxd
2/25/10
4:10 PM
Page 395
C H A P T E R
14 Smoking and Other Tobacco Use prevalent. Decreasing prevalence of smoking in the United States and other developed countries has led to intensified marketing of tobacco products in lowand middle-income countries, where the great majority of tobacco-related deaths are occurring. Called “the risk factor of the [20th] century” as of the mid-1990s, smoking has become an increasingly common target of intervention not only by health professionals but also by legislative and regulatory bodies. Both smoking cessation and enforcement of clean indoor air standards are associated with marked reduction in morbidity and mortality caused by tobacco. The public health goals are to reduce both demand and supply of tobacco products by measures to reduce incidence and promote cessation of their use and to reduce their production, marketing, and illicit distribution. A wide array of available interventions ranges from individual-level assistance in quitting tobacco use to population-level regulatory actions, from local clean-air regulations to the WHO Framework Convention on Tobacco Control (FCTC), a global treaty to strengthen the ability of countries to resist the forces behind the tobacco pandemic. Among current issues concerning tobacco are the concept of “harm reduction” through development of questionably less toxic products, the continued efforts of the tobacco industry to discredit evidence regarding secondhand smoke, and the real prospects of the FCTC to enable low- and middleincome countries to reverse this pandemic.
SUMMARY Cigarette smoking, as well as use of tobacco in other forms, is associated with a resulting addiction to nicotine. The process of addiction begins with recruitment of the school-age population through marketing and distribution of tobacco products, which in many countries reach youth in violation of existing laws or regulations. The cigarette as a source of nicotine carries with it all the combustion products of tobacco, additives, and paper that together cause subclinical atherosclerosis, coronary heart disease, stroke, peripheral arterial disease, and abdominal aortic aneurysm, as well as other major chronic diseases. Passive smoking or secondhand smoke exposure, as well as active smoking, is causally related to the tobacco pandemic. Epidemiologic studies of smoking habits include population surveys to determine patterns of use and trends, cohort studies to measure the effects of smoking on individual risks and population rates of cardiovascular and other diseases, and trials of strategies for prevention or cessation of the smoking habit. Assessment of individual-level tobacco use and exposure depends on self-report or on measurement of biomarkers such as concentration of cotinine (a product of nicotine metabolism) in body fluids. Populationlevel exposures are estimated as per capita consumption based on sales or tax revenues or as environmental exposures based on air quality measurement in worksites, restaurants, and other public places. Despite widespread knowledge of the hazards of smoking, marketing of tobacco products and other influences account for more than 20% of adults in the United States and 1.1 billion people worldwide being smokers. Secondhand smoke exposure is even more
INTRODUCTION More than a decade ago, cigarette smoking was dubbed “the risk factor of the [20th] century” (attribution to
395
46894_CH14_395_430.qxd
396
2/25/10
4:10 PM
Page 396
CHAPTER 14 SMOKING AND OTHER TOBACCO USE
Richard Peto unconfirmed). The WHO Report on the Global Tobacco Epidemic, 2008: The MPOWER package, begins with a page reading, “In the 20th century, the tobacco epidemic killed 100 million people worldwide.” The next page reads, “During the 21st century, it could kill 1 billion.”1, frontispiece Mortality, morbidity, and consumption of healthcare resources on a vast scale are attributable to the rising epidemic of cigarette smoking during the 20th century. Projections into the new century portend a stillmounting public health burden as a result of intensified marketing and use of tobacco products especially in low- and middle-income countries.2 The hazards of cigarette smoking for increased risks of lung cancer and other respiratory conditions are widely recognized. In fact, the resulting public health burden due to vascular diseases is much greater.3 Other forms of tobacco use have adverse health consequences as well, including nicotine addiction, an inherent aspect of habitual cigarette smoking.4 Additionally, passive exposure to tobacco smoke—“environmental tobacco smoke (ETS)” or “secondhand smoke (SHS)”—by nonsmokers has gained prominence as a major public health issue.5 Public health, or at least public, opposition to tobacco use especially in the form of cigarette smoking has a long history. Two notable recent accounts include Cigarette Wars: The Triumph of “The Little White Slaver” and Ashes to Ashes: America’s Hundred-Year Cigarette War, the Public Health, and the Unabashed Triumph of Philip Morris.6,7 These chronicles convey the flavor of controversy surrounding tobacco in a journalistic mode, complementing the extensive documentation of health research as provided in more than four decades of US Surgeon General’s Reports beginning in 1964.8 A concise history of the first 25 years of Surgeon General’s Reports was published in 1989.9 Global dimensions of the tobacco epidemic and control efforts are demonstrated in reports from the Global Burden of Disease Study, Disease Control Priorities Project, and the landmark Framework Convention on Tobacco Control, the world’s first global public health treaty, in force as of February 2005.10–13 Together, these sources provide valuable insight into social, political, and commercial as well as scientific perspectives on tobacco. These aspects are fundamental to understanding current discussion of policies and practices to address this major public health issue, on both national and global levels.
CONCEPTS AND DEFINITIONS Concepts and definitions of smoking and other tobacco use concern types of tobacco exposure, smok-
ing status, and categories of exposure. These aspects are encountered throughout the epidemiologic literature on smoking and tobacco use. Types of Tobacco Exposure The personal habit of cigarette smoking is important for the occurrence of cardiovascular diseases because of its attendant risks and very high prevalence in many populations. Accordingly, tobacco use has been studied extensively. For smokers, current or past, detailed information is sometimes sought about the particular brand or type of cigarette smoked, extent of smoke inhalation, or proportion of the cigarette usually smoked. Inquiry about pipe and cigar smoking is often included, but these practices are generally much less prevalent. Exposure to smoked tobacco occurs passively when nonsmoking persons share the environment of smokers. This form of exposure underlies studies of the effects of secondhand smoke exposure. In such studies, mainstream smoke (drawn through the cigarette and exhaled into the environment) may be distinguished from sidestream smoke (entering the air directly from the burning cigarette, whether or not it is actively being smoked). The 2006 Report of the Surgeon General, The Health Consequences of Involuntary Exposure to Tobacco Smoke, reviews the history of concern about this form of exposure, from at least as early as 1972.5 Smokeless tobacco is commonly used in the forms of a plug, dipping or chewing tobacco, or snuff. Among smokeless tobacco products in the United States, snuff is most common. This is powdered tobacco now usually in moistened form, fermented, and used orally by placing a quantity of it between the gum and cheek or lower lip.4 These forms of smokeless tobacco exposure cause local pathology, including cancer in the mouth and upper airway, but their immediate relevance is due to their yield of nicotine by solution into the saliva. This results in their potential for causing nicotine addiction, which may lead to cigarette smoking. Many forms of tobacco products, both smoked and smokeless, are used throughout the world. A report on tobacco control in India, for example, identifies as types of smoking tobacco beedies (the most common form, a loosely hand-rolled leaf containing tobacco), cigarettes, cigars, cheroots, chuttas, dhumti, and a variety of pipes; smokeless forms of tobacco for oral use are described as compounded with various other ingredients such as betel leaf, areca nut, and slaked lime. Several dental hygiene preparations also contain tobacco, are used by children as well as adults (chiefly women), and are potentially addicting because of their nicotine content.14
46894_CH14_395_430.qxd
2/25/10
4:10 PM
Page 397
MEASUREMENT 397
Smoking Status and Categories of Exposure With respect to personal smoking behavior, a common classification system distinguishes three categories: “never smoked,” “former smoker,” and “current smoker.” Explicit criteria are necessary to define each class and may vary among studies. For both former and current smokers, quantitation of exposure may be approximated by determining the age at the start of smoking, present age (or the age at stopping if not a current smoker), and the average number of cigarettes smoked per day over the smoking lifetime, in units of cigarettes, packs, or portions of packs. This approach provides a rough estimate of exposure in units of pack-years (e.g., 36 years of smoking 1⁄2 pack of cigarettes per day 18 pack-years of exposure) for either former or current smokers. With growing recognition of the importance of exposure to secondhand smoke, classification becomes more complex. For nonsmokers this is the only source, whereas for smokers, there is often additional exposure to smoke generated by others. Beyond the determination of whether a nonsmoker shares a residence or workplace with a smoker, and if so over what interval of time, it is difficult to gauge the extent of previous exposure.5 Another aspect of smoking status concerns the time course of starting, decreasing, and stopping tobacco use especially for those who may have quit on multiple occasions. This is important for observational studies estimating pack-years of exposure and even more so for studies assessing behavioral responses to intervention for smoking cessation.
MEASUREMENT Measurement of exposure to tobacco smoke is of interest at both individual and population levels, and for both active and passive smokers. Active Smoking For individuals, self-reports of smoking and tobacco use are commonly elicited by means of interviews or questionnaires designed to permit classifying individuals as described previously. Where cultural factors make smoking undesirable or even illegal, response bias in the direction of underreporting is to be expected. Similarly, participants in smoking prevention or cessation programs might be expected to exaggerate their success and thereby to minimize their reported tobacco use. For such reasons, self-report alone is of uncertain reliability. Different issues arise when the respondent is a relative or an acquaintance of the index subject (a
surrogate respondent) or when existing records are searched for data on an individual’s smoking history. Even with the likely limitations of such indirect assessment, introduction of smoking history data to the state of Washington’s standard death certificate was considered a potentially valuable addition to data on smoking and causes of death.15 Biochemical markers of tobacco smoke or smokeless tobacco products can be used to supplement or replace self-report methods.16 Nicotine, cotinine, thiocyanate, and carbon monoxide can be tested for this purpose. Depending on the choice of marker, samples of saliva, urine, blood, or expired air may be used. In a strategy designed to improve the reliability of selfreported smoking behavior in school populations (the “bogus pipeline” method), such samples are obtained from all participants but are processed for only a small proportion of them. All participants are informed that the samples can be tested, and evidence suggests that more reliable reporting occurs as a result. Secondhand Smoke The methods just described, principally cotinine concentration in biological samples, are useful in estimating immediate secondhand smoke exposure of individuals. But because the half-life of cotinine in such samples is on the order of roughly 1⁄2 to 2 days, exposures over several days or weeks, much less cumulative lifetime exposures, cannot be assessed by these means. A marker such as cotinine is also limited in reflecting only one substance, nicotine, among the great many components of tobacco smoke—including not only chemical substances but also respirable suspended particles. For these reasons, environmental measures are important. These include assessing atmospheric concentrations of nicotine and particulate air pollutants, as well as effects of heating, ventilating, and air conditioning systems in clearing or distributing contaminated air. Exposure models can be used for estimating secondhand smoke exposure by accounting for multiple factors operating in homes, worksites, and other indoor locations where exposure typically occurs.5,17 Population Exposures As a convenient indirect indicator of tobacco use for the population of a geopolitical area, data on tobacco sales or tax revenues may be available. Knowledge of the tax rate and revenues permits estimating the quantity of tobacco sold in a given time period. Changes in tax rates over time must be taken into account. Potential error may result from use of imported tobacco products or those acquired through untaxed, black-market sources. The estimates that result are per capita cigarette consumption rates.
46894_CH14_395_430.qxd
398
2/25/10
4:10 PM
Page 398
CHAPTER 14 SMOKING AND OTHER TOBACCO USE
DETERMINANTS Factors that influence the likelihood of smoking include marketing of cigarettes and other tobacco products, societal factors that may favor or oppose smoking, and individual behavioral characteristics. Marketing Manufacture of tobacco products in great quantity and their effective marketing by the tobacco industry through multiple media channels are powerful determinants of exposure to tobacco. Commercial profitability of cigarette manufacture is both a cause and a consequence of this phenomenon. The relation between government and this industry is often complex, especially in countries where the national government may be the producer of tobacco as well as the beneficiary of substantial tax revenues. In the United States, as of 1985, taxes on tobacco were estimated to bring $4.4 billion, $4.3 billion, and $0.2 billion to federal, state, and local government, respectively; corporate income tax of $1.7 billion was also received by the federal government. Of a total consumer expenditure of $30.2 billion for the purchase of cigarettes, the net profit to industry was $3.4 billion, or more than 10%. Industry expenditures for advertising and promotion of cigarettes in that year were estimated to be $2.5 billion.11 That figure, adjusted to 2006 dollars, amounted to $4.6 billion; by 2005, it had grown to $13.5 billion.18 An extensive study of the role of media examines their multifaceted influences on the tobacco epidemic through tobacco marketing, coverage of tobacco-related issues in news and entertainment media, communicating tobacco prevention and control messages, and addressing industry efforts to counter these messages.18 A key lesson from this study, in the words of its editors, is that “Most critical from a policy standpoint is the conclusion, supported by strong evidence, that both exposure to tobacco marketing and depictions of tobacco in movies promote smoking initiation. . . . The tobacco industry continues to succeed in overcoming partial restrictions on tobacco marketing in the United States, and tobacco marketing remains pervasive and effective in promoting tobacco use”18, p xvii–xviii Societal Influences To return to the historical perspective on society’s attitudes toward the cigarette, Tate’s Cigarette Wars depicts “The cigarette today” as “the most vilified product available legally in the United States, blamed for causing the premature deaths of more than
400,000 Americans a year, banned from most public buildings, besieged in the courts, and subject to increasing restrictions on advertising, promotion, and sales.”6, p 147 She notes that, regardless of this characterization of cigarettes, prevalence of smoking had not decreased substantially among US adults through the 1990s. By her account, the first wave of societal opposition to smoking, in the United States, began in the late 1880s, peaked in 1917, and waned over the next decade. The second wave, beginning in the 1960s and continuing to the present, has a scientific foundation not present in the earlier period and incorporates a strong focus on passive smoking. Today it appears that a change in culture has occurred making nonsmoking, rather than smoking, the norm and enabling rules, regulations, and legislation to be adopted that would have been implausible before the 1960s. (Related experience will be discussed later—see Prevention and Control.) Individual Behavior Initiation of smoking and other forms of tobacco use most often occurs in childhood and adolescence. This period of life was therefore the special focus of the 1994 report of the US Surgeon General, Preventing Tobacco Use Among Young People.19 Figure 14-1 outlines the stages of smoking initiation, described as the preparatory stage, trying stage, experimental stage, regular use, and addiction/dependent behavior. Factors that influence the progression from one stage to the next are described on the left of the figure, and particular patterns of behavior that define each stage are described on the right. These determinants of smoking behavior provide points of intervention in programs aimed at preventing progression from the preparatory stage to addiction or dependency. The same report presents a summary of studies indicating conditions predictive of progression in smoking behavior—socioeconomic, environmental, behavioral, and personal factors, as well as the influences of the tobacco industry.
MECHANISMS Active and Passive Smoking Cigarette smoke contains thousands of organic and inorganic compounds, components of tobacco itself or products of the combustion of both tobacco and cigarette additives. The toxic compounds include nicotine, carbon monoxide, hydrogen cyanide, acrolein, and mutagens and carcinogens (e.g., polycyclic aromatic hydrocarbons), in addition to respirable sus-
46894_CH14_395_430.qxd
2/25/10
4:10 PM
Page 399
MECHANISMS 399
Preparatory Stage Adolescent forms attitudes and beliefs about the utility of smoking.
Psychosocial risk factors include advertising and adult/sibling role models who smoke cigarettes.
Never smokes
Trying Stage Adolescent smokes first few cigarettes.
Psychosocial risk factors include peer influences to smoke, the perception that smoking is normative, and the availability of cigarettes.
No longer smokes Experimental Stage Adolescent smokes repeatedly, but irregularly.
Psychosocial risk factors include social situations and peers who support smoking, low self-efficacy in ability to refuse offers to smoke, and the availability of cigarettes.
No longer smokes Regular Use Adolescent smokes at least weekly across a variety of situations and personal interactions.
Psychosocial risk factors include peers who smoke, the perception that smoking has personal utility, and few restrictions on smoking in school, home, and community settings.
Quits smoking Addiction/Dependent Smoker
Adolescent has developed the physiological need for nicotine.
Figure 14-1 Stages of Smoking Initiation Among Children and Adolescents. Source: Reprinted from Preventing Tobacco Use Among Young People, a Report of the Surgeon General, p 126, 1994, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health.
pended particulates. These products are classified as components of either the tar (particulate) phase or the gas phase of cigarette smoke. Mainstream and sidestream smoke differ in the relative composition regarding these two phases, sidestream smoke having somewhat higher gas phase content. Secondhand smoke is about 85% sidestream and 15% exhaled mainstream smoke.20
Clinical and laboratory research provide extensive evidence of links between cigarette smoking and atherosclerosis through mechanisms including vasomotor dysfunction, inflammation, and modification of the lipid profile with lipid abnormalities such as increased low-density lipoprotein cholesterol, very-lowdensity lipoprotein cholesterol, and triglycerides and decreased high-density lipoprotein cholesterol. Several
46894_CH14_395_430.qxd
400
2/25/10
4:10 PM
Page 400
CHAPTER 14 SMOKING AND OTHER TOBACCO USE
pathways are thought to converge in increased oxidative stress. Thrombosis is also related to cigarette smoking, through platelet dysfunction and alteration of anti- and prothrombotic factors and fibrinolysis.20 Some of these relationships are also addressed in epidemiologic data, such as those from the Third National Health and Nutrition Examination Survey (NHANES III). Inflammatory markers and other cardiovascular risk factors were found to follow a gradient from most adverse levels among the heaviest smokers to successively less adverse among persons who had stopped smoking from 1 to 9 or more years prior to examination.21 Consistent findings from a later analysis from NHANES III based on cotinine levels among never-smokers showed increased levels of both homocysteine and fibrinogen when cotinine was detectable, even at low levels in contrast with being nondetectable.22 The heightened interest in secondhand smoke exposure, reflected in the Report of the Surgeon General for 2006, occasioned a detailed review of evidence on toxicology including heart disease as well as carcinogenesis and respiratory injury.5 That review concluded that evidence was sufficient to infer that exposure to secondhand smoke has a prothrombotic effect, causes endothelial cell dysfunction, and causes atherosclerosis in animal models. Further, “the immediate effects of even short exposures to secondhand smoke appear to be as large as those seen in association with active smoking of one pack of cigarettes a day.”5, p 64 Smokeless Tobacco Smokeless tobacco is toxic because of constituents of tobacco juices, even without tobacco’s combustion products. Its use may result in higher nicotine absorption than from smoked tobacco because of the acidity of saliva and prolonged contact between oral tissues and the tobacco. Studies of its toxicity for the cardiovascular system have not provided consistent evidence of sustained effects as with smoking. It is, however, addicting and is associated with oral malignancies. Whether snuff, as a popular form of smokeless tobacco, should be considered as a replacement therapy for nicotine addiction due to smoking has been discussed.4
DISTRIBUTION Prevalence—United States The prevalence of cigarette smoking in the US adult population aged 18 years and older, based on the
2006 National Health Interview Survey, is shown in Table 14-1 for men, women, and the total population by race/ethnicity, education, age group, and poverty status.23 Smokers were defined as persons who reported smoking at least 100 cigarettes during their lifetimes and who, at the time of interview, reported smoking every day or some days. The overall frequency of smoking was 20.8%, slightly greater for men and less for women. The prevalence was notably less for Asians/Pacific Islanders (10.4%) and greater for American Indians/Alaska Natives (32.4%). The strong gradient of less frequent smoking among successively more highly educated groups is also striking. The lower prevalence among the older age group, 65 years or older, is also noteworthy and could reflect reduced survival of smokers to these ages, greater frequency of smoking cessation, or less incidence of smoking at earlier ages among these groups, born in 1940 or earlier. Poverty (below versus at or above federal poverty level) was related to greater prevalence of smoking by 50% (30.6% versus 20.4%). For younger persons, in school grades 9–12, the Youth Risk Behavior Survey of 2007 distinguished among “lifetime smoking” (ever tried cigarette smoking, even one or two puffs), “current smoking” (smoked cigarettes on at least 1 day during the 30 days before the survey), and “current frequent smoking” (smoked cigarettes on 20 or more days during the 30 days before the survey (Table 14-2).24 Comparable data for alternate years, beginning in 1991, are also shown. The prevalence of the three levels of smoking in 2007 was 50.3%, 20.0%, and 8.1%, respectively. In each case, prevalence was less than the level in 1991 by about one-third, although trends in the three categories were not consistent. A companion table (not shown) provides subclassification for current smokers by sex, race/ethnicity (White, non-Hispanic; Black, non-Hispanic; and Hispanic) and school grade (9–12). Males were current smokers slightly more often than females, in most but not all years. Current smoking was less frequent among Black boys and girls than among Whites and intermediate in frequency among Hispanics. By grade 11, the same proportion of students as adults were classified as current smokers (although not by the same definition—see the preceding). It is a widely held view that adopting the cigarette smoking habit is infrequent after school age and that progression to daily smoking begins early and is completed before the end of high school for most eventual regular smokers. Table 14-3 illustrates this progression. Persons interviewed at ages 30–39 years were asked to recall the ages at which they had first tried a cigarette and at which they began smoking daily.
46894_CH14_395_430.qxd
2/25/10
4:10 PM
Page 401
DISTRIBUTION 401
Estimated Percentage of Persons Aged 18 Years Who Were Current Smokers,* by Sex and Selected Characteristics—National Health Interview Survey, United States, 2006 Men Women Total (n 10,715) (n 13,560) (n 24,275) % (95% CI†) % (95% CI) % (95% CI)
Table 14-1
Characteristic Race/Ethnicity§ White, non-Hispanic Black, non-Hispanic Hispanic American Indian/Alaska Native, non-Hispanic¶ Asian, non-Hispanic**
24.3 27.6 20.1 35.6
(23.0–25.6) (24.2–31.0) (17.8–22.4) (18.7–52.5)
19.7 19.2 10.1 29.0
(18.6–20.8) (17.3–21.1) (8.5–11.7) (15.7–42.3)
21.9 23.0 15.2 32.4
(21.0–22.8) (21.1–24.9) (13.7–16.7) (19.7–45.1)
16.8
(13.1–20.5)
4.6
(3.0–6.2)
10.4
(8.4–12.4)
Education (yrs) 0–12 (no diploma) 8 9–11 12 GED§§ diploma High school diploma Associate degree Some college Undergraduate degree Graduate degree
30.6 22.3 40.1 27.9 51.3 27.6 25.4 26.1 10.8 7.3
(27.9–33.3) (18.5–26.1) (35.7–44.5) (21.5–34.3) (43.4–59.2) (25.3–29.9) (22.1–28.7) (24.2–28.0) (9.0–12.6) (5.4–9.2)
23.0 12.3 31.4 23.3 40.2 20.4 17.8 20.0 8.4 5.8
(20.7–25.3) (9.7–14.9 (27.7–35.1) (17.5–29.1) (33.2–47.2) (18.7–22.1) (15.2–20.4) (18.3–21.7) (7.0–9.8) (4.1–7.5)
26.7 17.4 35.4 25.6 46.0 23.8 21.2 22.7 9.6 6.6
(25.0–28.4) (15.1–19.7) (32.5–38.3) (21.2–30.0) (40.5–51.5) (22.3–25.3) (19.1–23.3) (21.4–24.0) (8.5–10.7) (5.3–7.9)
Age group (yrs) 18–24 25–44 45–64 65
28.5 26.0 24.5 12.6
(24.7–32.3) (24.3–27.7) (22.7–26.3) (10.6–14.6)
19.3 21.0 19.3 8.3
(16.7–21.9) (19.7–22.3) (17.9–20.7) (7.0–9.6)
23.9 23.5 21.8 10.2
(21.7–26.1) (22.4–24.6) (20.6–23.0) (9.2–11.2)
Poverty status¶¶ At or above federal poverty level Below federal poverty level Unknown
22.9 34.0 23.3
(21.6–24.2) (30.0–38.0) (21.0–25.6)
17.8 28.0 14.2
(16.8–18.8) (25.2–30.8) (12.6–15.8)
20.4 30.6 18.3
(19.6–21.2) (28.0–33.2) (16.9–19.7)
Total
23.9
(22.8–25.0)
18.0
(17.2–18.8)
20.8
(20.1–21.5)
††
*Persons who reported smoking at least 100 cigarettes during their lifetimes and who, at the time of interview, reported smoking every day or some days. Excludes 315 respondents whose smoking status was unknown. † Confidence interval. § Excludes 266 respondents of unknown race or multiple races. ¶ Wide variances in estimates reflect small sample sizes. **Does not include Native Hawaiians or Other Pacific Islanders. †† Among persons aged ≥ 25 years. Excludes 305 persons whose educational level was unknown. §§ General Educational Development. ¶¶ Based on family income reported by respondents and 2005 poverty thresholds published by the US Census Bureau. Source: Reprinted from MMWR, Vol 54, November 11, 2005, p 1123.
The prevalence of both experimentation and daily smoking increased from those younger than age 12 years to each successively older age category, most sharply over the groups from younger than age 12 years to younger than age 18 years. By the latter age, more than one-third of all persons had begun smoking daily. More than 70% of those who became daily smokers had done so by this age.19 Exposure to secondhand smoke among nonsmokers was estimated from NHANES 1999–2002 in
age groups from 3–11 to 60 or more years.5 By the criterion of serum cotinine levels 0.05 ng (nanograms)/ml, 47.0% of persons were exposed—nearly 40 million younger than age 20 years and more than 125 million overall. These proportions were declining at the time, however, as comparison between findings of NHANES 1988–1994 and 1999–2004 later showed.25 Exposure at home decreased between periods by approximately 40–60% for nearly all sex, age, race/ethnicity, and income groups. Prevalence of
Source: Reprinted from MMWR, Vol 55, July 7, 2006, p 1123.
*Ever tried cigarette smoking, even one or two puffs. † Smoked cigarettes on at least 1 day during the 30 days before the survey. § Smoked cigarettes on 20 or more days during the 30 days before the survey. ¶ Linear, quadratic, and cubic trend analyses were conducted using a logistic regression model controlling for sex, race/ethnicity, and grade. These prevalence estimates are not standardized by demographic variables. **Confidence interval. †† Significant linear and quadratic effects only (p 0.05). §§ Significant linear, quadratic, and cubic effects (p 0.05).
Current Frequent††
Current§§
Cigarette Use Lifetime¶†
4:10 PM
Percentage of High School Students Who Reported Lifetime Cigarette Use,* Current Cigarette Use,† and Current Frequent Cigarette Use§—Youth Risk Behavior Survey, United States, 1991–2007 1991 1993 1995 1997 1999 2001 2003 2005 2007 % (95% CI**) % (95% CI) % (95% CI) % (95% CI) % (95% CI) % (95% CI) % (95% CI) % (95% CI) % (95% CI) 70.1 69.5 71.3 70.2 70.4 63.9 58.4 54.3 50.3 (67.8–72.3) (68.1–70.8) (69.5–73.0) (68.2–72.1) (67.3–73.3) (61.6–66.0) (55.1–61.6) (51.2–57.3) (47.2–53.5) 27.5 30.5 34.8 36.4 34.8 28.5 21.9 23.0 20.0 (24.8–30.3) (28.6–32.4) (32.5–37.2) (34.1–38.7) (32.3–37.4) (26.4–30.6) (19.8–24.2) (20.7–25.5) (17.6–22.6) 12.7 13.8 16.1 16.7 16.8 13.8 9.7 9.4 8.1 (10.6–15.3) (12.1–15.5) (13.6–19.1) (14.8–18.7) (14.3–19.6) (12.3–15.5) (8.3–11.3) (7.9–11.0) (6.7–9.8)
402
2/25/10
Table 14-2
46894_CH14_395_430.qxd Page 402
CHAPTER 14 SMOKING AND OTHER TOBACCO USE
46894_CH14_395_430.qxd
2/25/10
4:10 PM
Page 403
DISTRIBUTION 403
Table 14-3
Age (yrs) 12 14 16 18 18 20 25 30 39 Never smoked Mean age
Cumulative Percentages of Recalled Age at Which a Respondent First Tried a Cigarette and Began Smoking Daily, Among Persons Aged 30–39, National Household Surveys on Drug Abuse, United States, 1991 Persons Who Had Ever Tried Persons Who Had Ever All Personsa a Cigarette Smoked Daily First Tried Began Smoking First Tried First Tried Began Smoking a Cigarette Daily a Cigarette a Cigarette Daily 14.1 0.9 18.0 15.6 1.9 29.7 3.9 38.0 36.7 8.0 48.2 12.2 61.9 62.2 24.9 63.7 26.0 81.6 81.9 53.0 68.8 34.9 88.2 89.0 71.2 71.0 37.8 91.0 91.3 77.0 76.6 46.5 98.2 98.4 94.6 77.4 48.1 99.3 99.4 98.1 78.0 49.0 100.0 100.0 100.0 100.0 100.0 NA NA NA NA NA 14.5 14.6 17.7
Note: NA, not applicable. a All persons (N 6388). Source: Reprinted from Preventing Tobacco Use Among Young People, a Report of the Surgeon General, p 65, 1994, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health.
detectable cotinine levels in personal samples also decreased, but by a lesser degree and less consistently across groups—least for children, non-Hispanic Blacks, and persons of low income. These differences reflect group differences in reduction in secondhand smoke exposure outside the home. Prevalence—Global From the Disease Control Priorities Project, Table 14-4 presents estimates for the year 2000 of prevalence of smoking (definition not provided) by World Bank region, by sex, and overall.12 For each of six discrete geographic regions and two economic regions,
both percent of population and numbers of smokers (in millions) are shown, as well as the proportionate contribution of each region to the world total of more than 1.1 billion smokers. Low- and middle-income economies account for more than 80% of all smokers, nearly 40% in East Asia and the Pacific (including China) alone. Overall prevalence ranges from 18–21% in sub-Saharan Africa, South Asia, the Middle East, and North Africa to 34–35% in East Asia and the Pacific, and Europe and Central Asia. Males predominate over females by about 6:1 in lowand middle-income countries but only 2:1 in highincome countries.
Table 14-4
Estimated Smoking Prevalence (by Gender) and by Number of Smokers, 15 Years of Age and Older, 2000 Smoking Prevalence (Percent) Total Smokers World Bank Region Males Females Overall Millions Percentages of All Smokers East Asia and the Pacific 63 5 34 429 38 Europe and Central Asia 56 17 35 122 11 Latin America and the Caribbean 40 24 32 98 9 Middle East and North Africa 36 5 21 37 3 South Asia 32 6 20 178 15 Sub-Saharan Africa 29 8 18 56 6 Low- and middle-income economies 49 8 29 920 82 High-income economies 37 21 29 202 18 Source: Authors. Reprinted with permission from Disease Control Priorities in Developing Countries, Second Edition, edited by DT Jamison et al., Copyright © 2006. Courtesy of the International Bank for Reconstruction and Development/The World Bank.
46894_CH14_395_430.qxd
404
2/25/10
4:10 PM
Page 404
CHAPTER 14 SMOKING AND OTHER TOBACCO USE
The Global Youth Tobacco Survey (GYTS) provides tobacco-control surveillance data for youth in all WHO Regions, with a total of 151 survey sites worldwide as of 2007.26 Grade levels corresponding to ages 13–15 years are targeted, and standardized methods of survey design and implementation are used throughout the network. Summary data on selected characteristics for each region and overall are shown in Table 14-5; site-specific data are available in the cited reference. Smoking ranged fourfold in prevalence from 4.9% in the Eastern Mediterranean Region to 19.2% in the European Region. The majority of current smokers in every region responded that they desired to quit—from 53.3% to 80.7%. From 20.2% to 61.7% reported purchasing their cigarettes from stores. Reported secondhand smoke exposure was common both at home and in public places—42.5% and 55.1% overall. These indicators of the global dimensions of smoking in youth have major implications for prevention of future morbidity and mortality from tobacco. Trends Trends in the prevalence of smoking among adults aged 18 years and older in the United States have indicated a marked reduction from 1965 to 2006, though more than 1 in 5 Americans still smokes (Table 14-6).27 The reduction for men was from 51.2% to 23.6%; for women it was from 33.7% to 18.1%. The proportionate declines have been very similar between Whites and Blacks, although prevalence remained higher among Black than White males. Trends of declining prevalence in US youth were noted previously. However, globally, and specifically in lowand middle-income countries, smoking prevalence is said to be increasing.12
RELATION TO RATES AND RISKS In folklore, smoking and other uses of tobacco have long been related to adverse health consequences and were cited by the late 19th century as the cause of 87 diseases, including heart disease, alongside baldness, insanity, and tooth decay, among other ills.28 An account presented by the late Sir Richard Doll leads from early anecdotes about tobacco to the history of more recent epidemiologic investigation, including case-control studies prior to the mid-20th century.29 By the time of the 1964 report of the US Surgeon General, evidence was considered sufficient to find that smoking causes lung cancer, but the only conclusion at that time regarding cardiovascular disease was: “Male cigarette smokers have a higher death
rate from coronary artery disease than non-smoking males, but it is not clear that the association has causal significance.”8, p 39 Two decades later, the 1983 Surgeon General’s Report focused on cardiovascular disease and concluded that: “Cigarette smoking is a major cause of coronary heart disease in the United States for both men and women. Because of the number of persons in the population who smoke and the increased risk that cigarette smoking represents, it should be considered the most important of the known modifiable risk factors for CHD.”30, pp 6–7 No less conclusive was the statement regarding atherosclerotic peripheral arterial disease. No such statement could be made at that time with respect to cerebrovascular disease. Through decades of study subsequent to the first report in 1964, sufficient evidence accumulated by 2004 to infer a causal relationship between smoking and subclinical atherosclerosis, coronary heart disease, stroke, and abdominal aortic aneurysm.31 The relation of cigarette smoking to the occurrence of coronary heart disease can be illustrated by comparing both rates between populations and risks among individuals within a population. In more recent studies, increased attention is found to secondhand smoke and, to a much lesser degree, other tobacco use. Population Differences The classic Seven Countries Study, described in Chapter 4, investigated the relation between cigarette smoking and population rates of coronary heart disease events in 13 of the 16 cohorts in the study. Use of cigarettes reported at the baseline examination was examined in relation to incidence of coronary heart disease over the next 10 years in these cohorts, grouped by geographic area (Figure 14-2).32 At the baseline examination, cigarette smoking was classified “never,” “stopped,” or, for current smokers, by amount smoked in intervals of less than 10, 10 to 19, or 20 or more cigarettes per day. Among nonsmokers, the number of “hard” coronary events per 100 men ranged from less than two for the Yugoslavian cohorts to four for the Northern European cohorts. With successively greater exposure, from those who had already stopped smoking to those smoking 20 or more cigarettes per day, this difference was amplified. The relation between level of smoking and increasing coronary heart disease event rates was strongest for the populations with the highest rates (Northern Europe) and notably less strong for the populations with lower rates (Yugoslavia, Italy, and Greece). This pattern suggests that smoking potenti-
Source: Data from Centers for Disease Control and Prevention, Global Youth Tobacco Surveillance, 2000–2007. Surveillance Summaries, January 25, 2008. MMWR, 2008;57 (No. SS-1).
4:10 PM
Selected Characteristics Regarding Tobacco Use, Secondhand Smoke Exposure, and Experience of Current Smokers by WHO Regions and Overall, Global Youth Tobacco Survey, 2000–2007 Desired to Stop Usually Bought Exposed to Smoke Exposed to Smoke % Currently Smoking Their Cigarettes at Home, in Public Places, Smoked (Current in a Store Preceding Preceding Cigarettes Smokers) (Current Smokers) Week Week WHO Region % (CI) % (CI) % (CI) % (CI) % (CI) African Region 8.0 (6.2–10.5) 74.5 (61.2–83.7) 34.2 (24.7–45.7) 27.6 (23.8–31.9) 43.7 (39.6–47.9) Region of the Americas 14.3 (12.4–16.6) 53.3 (47.3–59.0) 20.2 (16.3–24.4) 41.1 (38.2–44.1) 54.9 (52.1–57.8) Eastern Mediterranean Region 4.9 (3.5–6.9) 70.5 (58.1–80.6) 42.2 (31.9–53.4) 38.3 (35.2–41.6) 45.7 (41.7–49.8) European Region 19.2 (17.0–21.7) 62.5 (56.0–68.7) 61.7 (56.7–66.4) 77.8 (75.3–80.0) 86.1 (84.4–87.7) South-East Asia Region 5.9 (4.8–7.2) 72.5 (63.6–79.9) 53.2 (46.0–60.2) 34.3 (31.3–37.4) 48.5 (45.3–51.6) Western Pacific Region 13.4 (11.2–16.0) 80.7 (74.6–85.7) 46.1 (40.6–51.8) 50.6 (47.7–53.6) 64.1 (61.3–66.8) Total 9.5 (7.9–11.3) 68.7 (60.1–75.9) 46.7 (39.9–53.6) 42.5 (39.5–45.5) 55.1 (52.0–58.1)
2/25/10
Table 14-5
46894_CH14_395_430.qxd Page 405
RELATION TO RATES AND RISKS 405
18.8 18.5
Data from Health, United States, 2007, National Center for Health Statistics, pp 266–267.
Sources: Centers for Disease Control and Prevention, National Center for Health Statistics, National Health Interview Survey. Data are from the core questionnaire (1965) and the following questionnaire supplements: hypertension (1974), smoking (1979), alcohol and health practices (1983), health promotion and disease prevention (1985, 1990–1991), cancer control and cancer epidemiology (1992), and year 2000 objectives (1993–1995). Starting with 1997, data are from the family core and sample adult questionnaires.
4:10 PM
20.8 23.6 18.1 23.5 26.1
2006
2/25/10
*Estimates are considered unreliable. Data preceded by an asterisk have a relative standard error of 20%–30%. 1 Data prior to 1997 are not strictly comparable with data for later years due to the 1997 questionnaire redesign. See Appendix I, National Health Interview Survey. 2 Estimates are age-adjusted to the year 2000 standard population using five age groups: 18–24 years, 25–34 years, 35–44 years, 45–64 years, 65 years and over. Age-adjusted estimates in this table may differ from other age-adjusted estimates based on the same data and presented elsewhere if different age groups are used in the adjustment procedure. See Appendix II, Age adjustment. 3 Starting with 1993 data, current cigarette smokers were defined as ever smoking 100 cigarettes in their lifetime and smoking now on every day or some days. See Appendix II, Cigarette smoking. 4 The race groups, White and Black, include persons of Hispanic and non-Hispanic origin. Starting with 1999 data, race-specific estimates are tabulated according to the 1997 Revisions to the Standards for the Classification of Federal Data on Race and Ethnicity and are not strictly comparable with estimates for earlier years. The single-race categories shown in the table conform to the 1997 Standards. Starting with 1999 data, race-specific estimates are for persons who reported only one racial group. Prior to 1999, data were tabulated according to the 1977 Standards. Estimates for single-race categories prior to 1999 included persons who reported one race or, if they reported more than one race, identified one race as best representing their race. Starting with 2003 data, race responses of other race and unspecified multiple race were treated as missing, and then race was imputed if these were the only race responses. Almost all persons with a race response of other race were of Hispanic origin. See Appendix II, Hispanic origin; Race. For additional data on cigarette smoking by racial groups, see Table 65. Notes: Standard errors for selected years are available in the spreadsheet version of this table. Available from: www.cdc.gov/nchs/hus.htm. Data for additional years are available. See Appendix III.
Current Cigarette Smoking Among Adults 18 Years of Age and Over, by Sex, Race, and Age: United States, Selected Years 1965–2006 [Data are based on household interviews of a sample of the civilian noninstitutionalized population] Sex, Race, and Age 19651 19741 19791 19851 19901 19951 2000 2002 2003 2004 2005 18 years and over, Percent of Persons Who Are Current Cigarette Smokers3 age-adjusted2 All persons 41.9 37.0 33.3 29.9 25.3 24.6 23.1 22.3 21.5 20.8 20.8 Male 51.2 42.8 37.0 32.2 28.0 26.5 25.2 24.6 23.7 23.0 23.4 Female 33.7 32.2 30.1 27.9 22.9 22.7 21.1 20.0 19.4 18.7 18.3 White male4 50.4 41.7 36.4 31.3 27.6 26.2 25.4 24.9 23.8 23.0 23.3 Black or African 58.8 53.6 43.9 40.2 32.8 29.4 25.7 26.6 25.3 23.5 25.9 American male4 White female4 33.9 32.0 30.3 27.9 23.5 23.4 22.0 21.0 20.1 19.5 19.1 Black or African 31.8 35.6 30.5 30.9 20.8 23.5 20.7 18.3 17.9 16.9 17.1 American female4
406
Table 14-6
46894_CH14_395_430.qxd Page 406
CHAPTER 14 SMOKING AND OTHER TOBACCO USE
46894_CH14_395_430.qxd
2/25/10
4:10 PM
Page 407
RELATION TO RATES AND RISKS 407
Hard CHD Incidence per 100, Age-Standardized
18 16 14 12 10 N. EUROPE Y = 4.0 + 1.2x
8 6
ITALY and GREECE Y = 2.3 + 0.3x
4
Y = 1.7 + 0.6x YUGOSLAVIA
2 0 x=0
1
2
3
(NEVER)
(STOPPED)
(<10/d)
(10–19/d)
4 (≥20/d)
Cigarette Habit at Entry Figure 14-2 Regression of Age-Standardized 10-Year Incidence Rate of Hard Coronary Heart Disease (CHD) on Smoking Class of 8717 Men Free of Cardiovascular Disease at Entry in Northern Europe (East and West Finland, Zutphen), in Yugoslavia (Dalmatia, Slavonia, Velika Krsna, Zrenjanin, and Belgrade), and Italy and Greece (Crevalcore, Montegiorgio, Rome Railroad, Crete, and Corfu). Source: Reprinted by permission of the publisher from Seven Countries: A Multivariat Analysis of Death and Coronary Heart Disease by Ancel Keys, Cambridge, Mass: Harvard University Press, © 1980 by the President and Fellows of Harvard College.
ated the effect of other factors that were contributing to higher background rates in Northern Europe. Data for death from all causes (not shown) indicated a twofold stronger relation of smoking to total mortality than to coronary heart disease incidence. Individual Differences The effect of smoking on the risk of coronary heart disease risk within a population has been assessed in many studies. Among the forerunners of these was the US National Pooling Project, described in Chapter 4.33 By combining data from five independent cohort studies, the Pooling Project obtained more reliable estimates of risks due to smoking and other factors than from any of the studies separately. Men age 40–64 years and free of coronary heart disease at baseline examination were classified as nonsmokers (never, past, or less than one-half pack per day); smokers of cigars or pipes only; or cigarette smokers who
smoked about one-half pack, about one pack, or more than one pack per day. After 8.6 years of follow-up, the rate of first coronary events was 143.1 per 1000 men among nonsmokers and 343.3 per 1000 men among smokers of more than one pack of cigarettes per day. The corresponding risk ratio was 2.4, and a consistent gradient in rates was found in relation to the amount smoked. (The risk ratio for smokers of cigars and pipes only, relative to nonsmokers, was 1.2.) From a landmark study of the mortality of British doctors in relation to their smoking habits, 40 years of follow-up experience was reported in 1994.3 Table 14-7 presents data from that report, including 13 categories of vascular deaths. For each category of death, the annual mortality rate is given for nonsmokers, former smokers, and current smokers both overall and by number of cigarettes smoked at baseline. Other smokers (users of cigars or pipes) are also represented. The columns under “Standardized Test for Trend”
4.2 10.8 5.4 7.0 3.8 3.0 3.9 2.6 3.4 5.0 0.6 0.5 1.5 15.7
0/1–14/ 25b 15–24/ 3.7 14.2 5.6 1.4 2.1 14.2
Source: Reprinted with permission from R Doll et al., Mortality in Relation to Smoking: 40 Years’ Observations on Male British Doctors, British Medical Journal, Vol 309, pp 904–905, © 1994, BMJ Publishing Group.
4:10 PM
3.7 7.5 3.5 6.9 1.9 1.1 2.4 1.0 1.4 3.2 0.5 0.1 0.7 10.5
N/X/Sa 1.1 9.9 3.3 0.4 0.1 8.2
Standardized Test for Trend
2/25/10
Note: If smoking is unrelated to mortality from a particular disease, the standardized trend test has expectation 0 and a standard deviation of unity, so values above 1.96, 2.57, and 3.29 correspond to p values (two-tailed) of 0.05, 0.01, and 0.001. a N, nonsmokers; X, former smokers of any type of tobacco; S, current smokers of any type. b 0/1–14/15–24/ 25 nonsmokers, smokers of 1–14, 15–24, and 25 or more cigarettes only.
Mortality from Respiratory and Vascular Diseases, by Smoking Habits Annual Mortality per 100,000 Men Cigarette Smokers Other Smokers Non-Smokers Current No. of Cigarettes Type of Disease (Never Smoked 25 (No. of Deaths, 1951–1991) Regularly) Former Current 1–14 15–24 Former Current Pulmonary tuberculosis (66) 4 8 11 7 9 20 8 4 Chronic obstructive disease (542) 10 57 127 86 112 225 40 51 Pneumonia (864) 71 90 138 113 154 169 94 85 Asthma (70) 4 11 7 6 8 6 9 7 Other respiratory disease (216) 19 28 30 26 31 33 24 18 All respiratory disease 107 192 313 237 310 471 176 164 (No. of deaths—1758) (131) (455) (490) (161) (170) (159) (290) (392) Pulmonary heart disease (64) 0 7 10 5 10 21 3 10 Ischemic heart disease (6438) 572 678 892 802 892 1025 676 653 Myocardial degeneration (841) 61 88 125 122 109 173 96 85 Aortic aneurysm (331) 15 33 62 38 74 81 22 43 Arteriosclerosis (232) 22 18 40 31 38 72 28 23 Hypertension (330) 32 33 44 28 51 60 37 33 Cerebral thrombosis (956) 93 95 122 93 150 143 100 106 Cerebral hemorrhage (607) 59 63 81 74 81 92 69 58 Subarachnoid hemorrhage (82) 7 10 15 10 12 24 4 6 Other cerebrovascular disease (1025) 94 110 164 167 145 188 101 103 Venous thrombosis (103) 9 11 14 17 11 14 13 9 Rheumatic heart disease (125) 15 10 15 15 20 8 17 13 Other cardiovascular disease (575) 58 63 71 60 82 74 62 59 All vascular deaths 1037 1221 1643 1447 1671 1938 1226 1201 (No. of deaths—11,709) (1304) (2761) (2870) (1026) (1045) (799) (1878) (2986)
408
Table 14-7
46894_CH14_395_430.qxd Page 408
CHAPTER 14 SMOKING AND OTHER TOBACCO USE
46894_CH14_395_430.qxd
2/25/10
4:10 PM
Page 409
RELATION TO RATES AND RISKS 409
present the value of the t-statistic for each calculated trend in mortality both among nonsmokers, former smokers, and current smokers (N/X/S) and by quantity smoked for current smokers. Values of t greater than 1.96 indicate a statistically significant trend (see legend). For all vascular deaths, for example, annual mortality rates were 1037 per 100,000 men per year among nonsmokers, 1221 among former smokers, and 1643 among current smokers, giving a t-value of 10.5, which was highly significant. Other significant trends with respect to smoking status or amount smoked were found for every vascular disease category except venous thrombosis, rheumatic heart disease, and the residual category, “Other.” Extensive data on the relation between smoking history and occurrence of coronary heart disease and stroke, as well as cancer, in both men and women, are provided by the American Cancer Society Cancer Prevention Study II, reported in detail in a National Cancer Institute monograph.34 Additional data on these conditions among women were reported from the Nurses’ Health Study in the same publication.35 Together, these sources offer important information on the relation of smoking to these conditions. The Cancer Prevention Study II entailed mortality followup of 1,185,106 men and women representing all 50 states and the District of Columbia, Puerto Rico, and Guam when surveyed in 1982.34 In six years of follow-up, 70,802 deaths were identified. The basis for the present analyses was the subset of the total cohort for whom complete smoking information was available at baseline and who were classified as lifelong nonsmokers (482,681) and current smokers of cigarettes only (228,682). In Figure 14-3A and B, coronary heart disease death rates are shown by age at death for men and women, respectively, contrasting current smokers (at baseline) and those who had never smoked. Coronary deaths occurred at appreciable rates after age 35–39 years in men and 50–54 years in women, consistent with earlier observations of the difference in age at coronary death by gender. The curve of increasing coronary mortality, though beginning later for women, rose more steeply with age for women than for men, and in both women and men the gradient of increasing mortality with age was greater in smokers. This observation underlies the pattern in Figure 14-4A and B, in which the rate differences (death rate in smokers minus death rate in nonsmokers) increase continuously with age, as mortality among smokers continues to exceed by greater and greater degrees that of nonsmokers. The effect is greater in men, with higher mortality in both groups at all midadult and later ages. The rate ratio, by contrast, expresses the
relative excess in mortality between groups, a measure that is greatest at younger ages, when coronary death is less frequent and that decreases as coronary mortality increases in the population as a whole, smokers and nonsmokers alike. The patterns are essentially the same for women as for men. The Nurses’ Health Study provided more detailed information on exposure to cigarettes, for 121,700 female registered nurses first evaluated by mailed questionnaire in 1976, analogous to the data for British doctors.35 Smoking status was updated by questionnaire every 2 years, and incident cases of coronary heart disease and stroke were identified by standardized procedures through mid-1988. Over this period, 970 cases of definite or probable coronary heart disease and 448 cases of stroke were identified. Relative risks (RRs) were calculated first with adjustment for age alone and then for multivariate adjustment. All relative risks were greater than 1, and their confidence limits did not include 1. The risks were closely parallel for fatal and nonfatal events, increased consistently with increasing exposure, and also increased by adjustment for other risk factors in most of the dose-specific analyses. Numbers of cases in the extreme exposure groups were small. Regarding mortality from stroke for men and women in the Cancer Prevention Study II, the picture was much the same as for coronary heart disease, except that rate differences were smaller because of the less frequent occurrence of stroke death.34 Rate ratios for stroke were similar to those for coronary death. In addition, the onset of rising death rates with age was essentially the same for women as for men, with no lag as in coronary event rates. The rates for women did not reach those for men after age 74 years. The relation of smoking to stroke incidence among women also was examined in the Nurses’ Health Study.35 For total stroke, as well as subtypes (subarachnoid hemorrhage, ischemic stroke, and cerebral hemorrhage), relative risks were calculated as for coronary heart disease, as described previously. Risks were increased for former smokers (not beyond chance variation in all comparisons), for current smokers, and generally across successively higher degrees of exposure among current smokers. Numbers of cases were small in some categories of stroke type and exposure level, especially for cerebral hemorrhage. Smoking is also strongly related to risk of peripheral arterial disease, and Powell designated smoking as the primary risk factor for this cardiovascular condition.36 This form of advanced atherosclerosis is perhaps least dependent on other risk factors and most specifically related to smoking. When peripheral arterial disease in smokers has progressed to the point
46894_CH14_395_430.qxd
4:10 PM
Page 410
CHAPTER 14 SMOKING AND OTHER TOBACCO USE
Rate per 10,000 Person-Years
410
2/25/10
300 Current smoker Never-smoker 200
100
65–69
70–74
75–79
80–84
65–69
70–74
75–79
80–84
60–64
55–59
50–54
45–49
A
40–44
35–39
0
Rate per 10,000 Person-Years
Age, Years
300 Current smoker Never-smoker 200
100
60–64
55–59
50–54
45–49
40–44
B
35–39
0
Age, Years Figure 14-3 Coronary Heart Disease Death Rates in Current Smokers and Lifelong Nonsmokers, by Age, Men (Panel A) and Women (Panel B). Source: Reprinted with permission from National Cancer Institute, National Institutes of Health, Bethesda, Maryland.
of surgical treatment with vein grafts to bypass occluded arterial segments, continued smoking is associated with graft occlusion. This sequence of events is not unlike that for patients following acute myocardial infarction or coronary artery bypass procedures, in whom failure to stop smoking is associated with reinfarction. Whether these health consequences vary by type of cigarette or other tobacco use is important for prevention policy. Examples of studies of this question include large case-control studies, each with several thousand participants, one in the United Kingdom and one in 52 countries. The UK study of nonfatal myo-
cardial infarction compared cases and controls with respect to the tar yields of the reported types of cigarettes smoked.37 The difference in risk between smokers of any type of cigarette and nonsmokers greatly exceeded that between smokers of lower versus higher tar yield. The authors concluded that significant risk reduction required not smoking rather than choice of a lower-tar cigarette. The multinational study INTERHEART investigated several types of tobacco use, as illustrated in Figure 14-5.38 Compared with never smokers, smokers of filter or nonfilter cigarettes, beedies, or pipes and users of chewing tobacco with or without smok-
46894_CH14_395_430.qxd
2/25/10
4:10 PM
Page 411
RELATION TO RATES AND RISKS 411
800 Rate ratio 600
4 400 2
200
80–84
75–79
70–74
65–69
60–64
55–59
50–54
45–49
0
35–39
0
A
Rate difference
40–44
Rate Ratio
6
Rate Differences (per 100,000)
1,000
8
Age, Years
800
4 400 2
Rate difference
200
80–84
75–79
70–74
65–69
60–64
55–59
45–49
40–44
0
35–39
0
B
600
Rate ratio
50–54
Rate Ratio
6
Rate Differences (per 100,000)
1,000
8
Age, Years Figure 14-4 Coronary Heart Disease Rate Ratios and Rate Differences in Current Cigarette Smokers and Lifelong Nonsmokers, by Age, Men (Panel A), and Women (Panel B). Source: Reprinted with permission from National Cancer Institute, National Institutes of Health, Bethesda, Maryland.
ing exhibited odds ratios for acute myocardial infarction ranging from 2 to 4. Again, nonuse of tobacco was the only low-risk alternative. Regarding consequences of secondhand smoke exposure in nonsmokers, a meta-analysis of 10 cohort and 8 case-control studies reported in 1999 concluded that such exposure was related to occurrence of coronary heart disease with an overall relative risk of 1.25 (95% CI, 1.17–1.32) (Figure 14-6).39 Exposure in the home was more strongly related than workplace exposure, as was exposure to smokers of more rather than less than 1 pack per day of cigarettes. Subsequently, a cohort study was reported based on serum
cotinine concentration at baseline and 20 years’ followup.40 Adjusted relative hazards of major coronary heart disease events over 20 years were from 1.45 (95% CI 1.01–2.08) to 1.57 (1.08–2.28) in the second to fourth versus first quartiles of cotinine concentration. However, in the first 5 year period, the hazard ratios were greater—e.g., fully adjusted, 3.73 (95% CI 1.32–10.58). No consistent association was found for stroke. Cotinine levels were judged to be more sensitive than self-report as an indicator of exposure. The US Surgeon General’s report of 2006, focusing on secondhand smoke, found the evidence sufficient to infer a causal relationship between exposure
46894_CH14_395_430.qxd
412
2/25/10
4:10 PM
Page 412
CHAPTER 14 SMOKING AND OTHER TOBACCO USE
8
OR (95% CI)
4
2
1 0.75 Never
Filter
Non-Filter Beedies
Pipes
Chew
Chew and Smoke
Figure 14-5 Risk of AMI Associated with Type of Tobacco Use. OR for current smokers 2.95 (95% CI 2.77–3.14) indicated by broken horizontal line. Never never smokers. Filter = filter cigarettes. Non-filter non-filter cigarettes. Beedies smoking beedies alone. Pipes smoking pipes/cigars. Chew chewing tobacco alone. Chew and smoke both chewing and smoking tobacco. Source: Reprinted with permission from The Lancet, Vol 368, KK Teo, S Ounpuu, S Hawken, MR Pandey, et al., p 651.
to secondhand smoke and coronary heart disease morbidity and mortality among both men and women but insufficient for such conclusions for either stroke or subclinical vascular disease.5
STUDY (YEAR)
EXPOSURE
Regarding subclinical coronary artery disease, the Pathobiological Determinants of Atherosclerosis in Youth (PDAY) Research Group published further findings on smoking in 2005 (see also Chapter 3).41
NO EXPOSURE
No. of events/no. at risk Cohort Hirayama7,8 (1984) Garland et al.9 (1985) Svendsen et al.10 (1987) Butler11 (1988) Butler11 (1988) Sandler et al.12 (1989) Hole et al.13 (1989) Humble et al.14 (1990) Steenland et al.15 (1996) Kawachi et al.15 (1997)
376/69,645 17/492 5/286 4/430 50/2802 673/10,799 54/1538 49/296 571/67,369 135/25,959
118/21,895 2/203 8/959 60/6077 95/3630 685/8236 30/917 27/217 2574/164,831 17/6087
CASE PATIENTS
CONTROLS
No. with exposure/no. without exposure Case-control Lee et al.17 (1986) He et al.18 (1989) Jackson19 (1989) Dobson et al.20 (1991) La Vecchia et al.21 (1993) He et al.22 (1994) Muscar and Wynder23 (1995) Ciruzzi et al.24 (1998) Overall
70/48 25/9 18/21 65/278 24/66 48/11 63/51 131/205
269/182 30/38 87/148 133/692 37/157 76/50 70/88 117/329
0.1
0.5 1 Relative Risk
5
10
Figure 14-6 Relative Risks of Coronary Heart Disease Associated with Passive Smoking Among Nonsmokers in 18 Epidemiologic Studies. The horizontal bars represent the 95 percent confidence intervals. The relative risk in the study by Garland et al. was 14.9. Source: Reprinted with permission from The New England Journal of Medicine, Vol 340, J He, S Vupputuri, K Allen, MR Prerost, J Hughes, PK Whelton, p 923. Copyright © 1999 Massachusetts Medical Society.
46894_CH14_395_430.qxd
2/25/10
4:10 PM
Page 413
RELATION TO RATES AND RISKS 413
On the basis of postmortem examination of coronary artery specimens and assays of serum thiocyanate concentrations, they found evidence that smoking (defined as 90 µmol/L thiocyanate) was related to the presence of more advanced atherosclerotic lesions (odds ratio 9.61, 95% CI 2.34–39.57). They interpreted their findings to indicate a role of smoking in accelerating progression of advanced lesions and postulated a thrombotic mechanism underlying this association. Whether smokeless tobacco use influences individual risk of coronary heart disease events or risks was the subject of a review of 15 studies in 2004; 3 studies addressed coronary heart disease outcomes and 12 addressed possible associations with other risk factors.42 No consistent evidence was found, but the quality of evidence was limited. A study of stroke found smoking but not use of snuff to be associated.43 There was no evidence of an effect on stroke risk for use of smokeless tobacco. Public Health Impact The public health impact of smoking on cardiovascular diseases, as well as other major chronic diseases, is addressed in innumerable scientific reports and other documents. Several examples serve to illustrate the magnitude of this impact. The study of male British doctors, cited previously, completed 50 years of followup in 2001.3,44 This study estimated that from one-half to two-thirds of deaths among continuing smokers in this cohort were due to smoking. From the Disease Control Priorities Project, data in Table 1-8 present the individual contribution of
smoking to both ischemic heart disease and stroke, worldwide and by broad economic region.45 In highincome areas, the population-attributable fractions for these conditions are 23% and 21%, respectively, whereas in the low- and middle-income regions, the corresponding fractions are 15% and 12%; these would rise as prevalence of smoking increases in these regions. The Project estimated that more than 5 million tobacco deaths occurred in the year 2000, approximately three times as many in men as in women.12 The pace of the tobacco epidemic is increasing sharply if 1 20 th of the number of deaths in the entire 20th century has already occurred in a single year. In addition, more than 75 million DALYs were similarly distributed, three times as many in men as in women, a huge toll in combined morbidity and mortality. The global expanse of this impact is shown in Figure 14-7, presented in WHO’s World Health Report for 2002.46 The INTERHEART study, cited previously and presented in Figure 4-5 and Table 4-10, estimated age-sex- and region-adjusted population-attributable risks from smoking as 15.8% for women and 44.0% for men across the 52 countries represented.38 For men and women together, the region-specific estimates show smoking to be a major factor in every region. Again, should smoking prevalence increase, especially among women, these figures would increase correspondingly. Increases in prevalence of smoking, especially among youth, are a grave concern for the future. It is also clear that today’s smokers are already at risk.
Proportion of DALYs attributable to selected risk factor ⬍0.5% 0.5⫺0.9% 1⫺1.9% 2⫺3.9% 4⫺7.9% 8⫺15.9% 16%⫹
Figure 14-7 Burden of Disease Attributable to Tobacco (% DALYs in Each Subregion). Source: Reprinted with permission from The World Health Report 2002, Reducing Risks, Promoting Healthy Life, p 65. © World Health Organization 2002.
46894_CH14_395_430.qxd
414
2/25/10
4:10 PM
Page 414
CHAPTER 14 SMOKING AND OTHER TOBACCO USE
This means that cessation of smoking among current smokers is itself an urgent global priority, addressed later (see Prevention and Control). The public health benefit of preventing smoking in the first place is not less important, but its impact will be more distant.
able levels among those who had stopped smoking and appeared to be dose related among continuing smokers.21 These relationships become especially important in understanding benefits of smoking cessation or protection from secondhand smoke and the time course of reduction in cardiovascular risk that follows.
RELATION TO OTHER FACTORS With respect to between-population differences in the impact of smoking, it was suggested in discussion of Figure 14-2, in the preceding, that underlying rates attributable to other factors in the Northern European region of the Seven Countries Study were potentiated by smoking. A similar effect may underlie the finding in Japan of an interaction between smoking and levels of serum total cholesterol concentration.47 Cardiovascular mortality is most strongly associated with smoking in the presence of successively higher levels of cholesterol. The relation between smoking and two other major risk factors for coronary heart disease mortality is indicated by the experience of men screened as potential candidates for the Multiple Risk Factor Intervention Trial (MRFIT), described in Chapter 4.48 This large study population of more than 360,000 men permitted more-detailed cross-tabulation of resulting data than is usually possible. The men were grouped by quintiles of systolic blood pressure and serum total cholesterol concentration and by smoking status at the screening examination, as was shown in Table 4-9. Relative to nonsmokers, smokers experienced two to three times the rate of coronary heart disease death. The effect was greatest at the lowest levels of systolic blood pressure and serum total cholesterol concentration (10.37 versus 3.09 deaths per 10,000 person-years, risk ratio 3.4), to nearly two times the rate at the highest levels of these other factors (62.11 versus 33.40 deaths per 10,000 personyears, risk ratio 1.9). Smoking multiplied the risk at all levels of the other factors. In a study of British civil servants, a relative increase in risk among smokers versus nonsmokers was strongly related to successively higher levels of cholesterol, an interaction similar to that seen in Japan.49 Beyond the major risk factors, relations of smoking to other factors and mechanisms underlying cardiovascular disease have been reviewed extensively in connection with recent reports of the US Surgeon General. For example, the 2006 report concluded that both prothrombotic effects and endothelial cell dysfunction are caused by secondhand smoke.5 In NHANES III, C-reactive protein, white blood cell count, and fibrinogen concentrations all showed more favor-
PREVENTION AND CONTROL The US Surgeon General’s reports on health consequences of smoking from 1964 to the present have been cited extensively in the preceding discussion. Other sources of recommendations regarding prevention and control of tobacco use over the past halfcentury are also noteworthy. Early examples in the United States include a 1956 report and 1960 update from the American Heart Association Committee on Smoking and Cardiovascular Disease.50 The update proposed calling to the attention of both health professionals and the general public data linking heavy smoking with coronary heart disease. This advice predated the first Surgeon General’s report by 4 years— a report that hesitated on the causal significance of cigarette smoking for coronary deaths in males, but stated that a causative role of smoking was strongly enough suspected, along with other factors (high blood pressure, high cholesterol, and “excessive obesity”), to take countermeasures against them.8 Internationally, the World Health Organization first addressed tobacco in 1975.51 This first report presented a broad set of recommendations for action to prevent and control smoking. Subsequent reports in 1979 and 1983 became more emphatic, with the latter report focusing on the tobacco problem in developing countries:52,53, p 68 The Committee stressed that while its recommendations are again directed to WHO and, through WHO, to governments (and not only health ministries), and to official organizations, they are also intended for a wider public. The Committee hopes that this report will succeed in drawing attention, internationally and nationally, to the urgency of the need for action to control smoking in developing countries, if a preventable human disaster of proportions unprecedented in the modern world in time of peace is to be avoided. It is hoped that, for the sake of the developing countries, the recommendations will be implemented while there is still time to prevent the problem reaching in those countries the levels it has already assumed in developed countries.
46894_CH14_395_430.qxd
2/25/10
4:10 PM
Page 415
PREVENTION AND CONTROL 415
vertising, promotion and sponsorship, and Raise taxes on tobacco. It is self-evident that preventing tobacco use will prevent its health consequences. In addition, epidemiologic evidence establishes the health benefits of smoking cessation and lends impetus to strategies addressing this aspect of tobacco control. As noted previously, in reference to the studies of British doctors and US nurses, former smokers experienced mortality from vascular and other causes at rates intermediate between those who were never smokers and those who were current smokers.3,35 The excess risk can be reduced but not eliminated, even after an interval of many years. In the latter study, the risk of those classified as current smokers was taken as the reference value and the risks for those quitting for successively longer time intervals were then calculated (Table 14-8). Risks for both total and cardiovascular mortality decreased sharply and continuously after quitting smoking. Only for cancer death rates that included cancer of the lung was this benefit not observed. Cardiovascular mortality was reduced to an age-adjusted risk of 0.76 times that of current smokers, or a 24% reduction within 2 years, and an eventual reduction of 60–70% was observed after 15 years or more. In a clinical trial linking smoking cessation with health outcomes, changes in risk through individuallevel intervention were tested in the Whitehall Study
This advice of more than 25 years ago remains to be fully implemented today, with the predicted consequences already evident, as documented in detail for many countries in the 2008 report from WHO.1 The United States has seen a marked decline in tobacco consumption, as shown in the long-term secular trend from 1900 through the 1990s, punctuated by multiple critical events (Figure 14-8).54 Still, as noted––though understated––in the 2007 US National Institutes of Health state-of-the-science report, “Tobacco use remains a very serious public health problem.”55, p S312 Renewed and more aggressive approaches have been advocated on both national and global levels. One example is the 2007 report from the US Institute of Medicine, Ending the Tobacco Problem: A Blueprint for the Nation.56 Its 42 recommendations elaborate upon three broad strategies: strengthening traditional tobacco control measures, changing the regulatory landscape, and opening new frontiers—a new generation of modeling approaches for future policy innovations and potentially modifying cigarettes by reducing their nicotine content. A global counterpart is the “MPOWER package” featured in the WHO Report on the Global Tobacco Epidemic, 2008.1 The six policies central to the MPOWER package are to Monitor tobacco use and prevention policies, Protect people from tobacco smoke, Offer help to quit tobacco use, Warn about the dangers of tobacco, Enforce bans on tobacco ad-
1964 Surgeon General’s report Broadcast ad ban Coalescence of modern advocacy movement Synar Nonsmokers’ Amendment rights movement enacted
5000
U.S. entry into WWII
Number of Cigarettes
4000
begins
3000
Federal cigarette tax doubles
First modern reports linking smoking and cancer
2000
Cigarette price drop
Fairness Doctrine messages on broadcast media
U.S. entry into WWI 1000
FDA proposed rule
Great Depression 0 1900
1910
1920
1930
1940
1950 Year
1960
1970
1980
1990
Note: The 1999 data are preliminary.
Figure 14-8 Adult per Capita Cigarette Consumption and Major Smoking and Health Events, United States, 1900–1999. Source: Reprinted from Reducing Tobacco Use, a Report of the Surgeon General, Department of Health and Human Services, US Government Printing Office, 2000.
46894_CH14_395_430.qxd
416
2/25/10
4:10 PM
Page 416
CHAPTER 14 SMOKING AND OTHER TOBACCO USE
Table 14-8
Event Total Mortality Casesa RR Casesb RR
Total and Cause-Specific Mortality by Time Since Quitting: Comparison of Analyses with and Without 2-Year Exclusion of Disease at the Start of Each Follow-Up Period: Multivariate Relative Risks (RRs) Years Since Quitting Among Former Smokers NeverCurrent Smoker Smoker 2 2–4 5–9 10–14 15 933 0.56 632 0.49 (0.44–0.54)
1,115 1.00 884 1.00
127 1.19 51 0.76 (0.53–1.08)
106 1.00 58 0.73 (0.53–1.01)
131 0.79 84 0.70 (0.53–0.92)
66 0.53 46 0.47 (0.33–0.67)
231 0.61 137 0.49 (0.39–0.62)
Cardiovascular Disease Casesa 131 RR 0.30 Casesb 111 RR 0.29 (0.23–0.37)
284 1.00 254 1.00
20 0.76 11 0.63 (0.28–1.45)
24 0.90 11 0.53 (0.25–1.13)
32 0.75 23 0.67 (0.40–1.15)
9 0.29 7 0.27 (0.11–0.65)
39 0.42 33 0.46 (0.29–0.74)
Total Cancer, Including Lung Casesa 516 RR 0.99 Casesb 562 RR 0.54 (0.46–0.64)
502 1.00 339 1.00
75 1.37 13 0.42 (0.20–0.89)
48 0.97 19 0.66 (0.38–1.16)
69 1.12 33 0.75 (0.49–1.16)
37 0.91 20 0.56 (0.33–0.96)
134 1.10 53 0.51 (0.35–0.74)
Total Cancer, Excluding Lung Casesa 492 RR 0.60 Casesb 244 RR 0.85 (0.71–1.03)
351 1.00 201 1.00
49 1.22 9 0.44 (0.18–1.08)
33 0.99 11 0.71 (0.34–1.48)
57 0.63 25 1.03 (0.63–1.69)
34 0.70 18 0.85 (0.48–1.51)
127 0.72 50 0.81 (0.54–1.20)
a
Cases and multivariate RRs after baseline exclusion of coronary heart disease, stroke, and cancer except nonmelanoma skin cancer. Multivariate RRs were adjusted for age in five-year intervals, follow-up period (1976–1978, 1978–1980, 1980–1982, 1982–1984, 1984–1986, or 1986–1988), body mass index, history of hypertension, diabetes, high cholesterol levels, postmenopausal estrogen therapy, menopausal status, past use of oral contraceptives, parental history of myocardial infarction before age 60, and daily number of cigarettes smoked during the period before stopping smoking (95% confidence intervals in parentheses). b Cases and multivariate RRs after exclusion of coronary heart disease, stroke, and cancer (except nonmelanoma skin cancer) at the beginning of each two-year follow-up interval. Source: Reprinted from National Cancer Institute, National Institutes of Health, Bethesda, Maryland.
of London civil servants.57 Subjects were male smokers at the highest risks for coronary heart disease or chronic bronchitis as estimated from a multivariate risk score; 714 men were randomly allocated to a smoking cessation intervention group and 731 men served as controls, with no systematic advice concerning their smoking. Overall, after 10 years in the trial, the occurrence of coronary heart disease was reduced in the intervention group by 18%, on the basis of cumulative event rates of 7.3% versus 8.9%. At the first reported point in follow-up, at 2 years, there was already a reduction of nearly 50% in the observed rates, based on event rates of 1.1% and 0.6%. Although these were quite low rates in both groups,
it is notable that the benefit of smoking cessation was suggested at the earliest reported comparison, after only 2 years in the trial. (Overall mortality was not reduced in the intervention group despite reductions in both coronary heart disease and lung cancer, for reasons discussed in that report.) These findings are underscored by a current assessment based on meta-analysis of 20 observational studies of mortality reduction associated with smoking cessation among patients with coronary artery disease.58 Among those who quit smoking, relative risk of death was 36% less (RR 0.64, 95% CI 0.58–0.77) than among those who continued to smoke. This effect was noted to exceed that generally
46894_CH14_395_430.qxd
2/25/10
4:10 PM
Page 417
PREVENTION AND CONTROL 417
attributed to treatment of such patients with statins, aspirin, beta-blockers, or angiotensin-converting enzyme inhibitors. Given the broad strategic approaches that have been proposed and urged over several decades, what measures are being taken for prevention and control of tobacco use and its epidemic occurrence? A schematic figure offers an overview of Reducing Tobacco Use, a Report of the Surgeon General (2000) and identifies five broad types of intervention, with their respective targets, tools, study approaches, and outcome measurements (Table 14-9).54 Interventions are categorized as educational, clinical, regulatory, economic, and social/comprehensive. Approaches to prevent tobacco use among young people through educational intervention are discussed (updating the 1994 report that focused on this issue19), as is each of the other four categories. It is convenient to consider intervention at each of three levels—individual, community- or populationwide, and global. Individual measures principally concern smoking cessation; community- or populationwide measures address both prevention and cessation, as well as reduction of secondhand smoke exposure; and global measures include addressing determinants of the two key aspects of demand and supply of tobacco products, seeking reduction of both on a global scale. Individual Measures Jha and colleagues sum up 50 years of epidemiology regarding smoking-related diseases as leading to “three key messages for individual smokers worldwide”:12, p 870 • The eventual risk of death from smoking is high, with about one-half to two-thirds of longterm smokers eventually being killed by their addiction. • These deaths involve a substantial number of life years forgone. About half of all tobacco deaths occur at ages 35 to 69, resulting in the loss of about 20 to 25 years of life, compared with the life expectancy of nonsmokers. • Cessation works: Those adults who quit before middle age avoid almost all the excess hazards of continued smoking. A systematic review of smoking cessation approaches for adults updated previous meta-analyses and systematic reviews with studies published up to 2005.59 The new report addressed self-help ap-
proaches, counseling, pharmaceutical monotherapy, combined pharmacotherapies, and joint pharmacotherapy with psychological interventions. Self-help strategies alone were found only marginally beneficial, whereas the other approaches increased cessation success significantly. The finding regarding self-help seems counter to other observations, for example, that “the great majority of smokers (more than 90 percent) who successfully quit did so “on their own”––that is, without the assistance of formal cessation programs.”19, p 100 But estimating quit rates is complicated by high relapse rates (75–80% within 6 months) and cyclic periods of smoking and quitting over periods of years.19 Screening of all adults for tobacco use and offering interventions to support cessation are strongly recommended by the US Preventive Services Task Force; an algorithm to identify and treat patients who smoke is presented in Figure 14-9; and detailed approaches to making intervention efforts effective in clinical and hospital settings are presented in a 2007 review.60,61,62 In addition to such clinical interventions, an important adjunct to individual-level resources to support smoking cessation is the “quitline,” a method for connecting motivated smokers with a telephone-based support system.63 An evaluation of state-level programs in the United States as of 2004 indicated that the majority of smokers live in states with tobacco quitlines; per capita cost is very modest; and many features are common to these programs across the states. Federal support for quitline interventions is represented by the National Network of Tobacco Cessation Quitlines; the National Cancer Institute’s national portal number, 1-800-QUITNOW, which links callers with their own state quitline; and support to states from the Centers for Disease Control and Prevention to establish or enhance their quitline services. With respect to tobacco cessation programs for youths, individual-level intervention was found to be a component of the majority of them, in a survey of such programs in a sample of more than 400 US counties reported in 2007.64 The programs were considered quite homogeneous across the United States and were generally school based, minimally funded, prepackaged, multisession, and education and counseling oriented—they rarely used medications. Approximately 80% of programs indicated an evaluation component, but outcomes were not presented in this report. Total participation represented only a small fraction of estimated youth who were smokers in the studied counties, and programs were most often lacking in geographic areas where national data indicate smoking prevalence among youth to be increasing.
Local ordinance State regulation Federal regulation Federal law Nongovernment action (e.g., joint commission accreditation of hospital organization) Local ordinance State regulation Federal regulation Federal law International agreements Media advocacy Direct advocacy Community interventions Countermarketing Regulation Policy formation
Product manufacture Product sale Vendors and buyers Public venues Public transportation Worksites Health care sites Taxes Tariffs and trade Price supports
Legislators Media Communication networks Case-by-case strategy State/local programs
Regulatory
Economic
Social/Comprehensive
Source: Reprinted from Reducing Tobacco Use, a Report of the Surgeon General, p 9, Department of Health and Human Services, US Government Printing Office, 2000.
Observational Case study General epidemiologic methods Trend analysis Knowledge/attitude/practice studies
Econometric analysis Trend analysis Multivariate models
Linear trends Case study analysis Cross-sectional comparisons
Linear trend Parameter estimates (e.g., elasticities)
Linear trend Cross-sectional comparison of proportion Case analysis results
Relative risk Attributable risk Effect size (absolute or relative)
Epidemiologic and behavioral: • Usually a comparison of “treatment” and “no treatment” groups • Control of confounding by behavioral and demographic variables
4:10 PM
Observational Knowledge/attitude/practice studies Surveillance Case study
Outcome Measurements Relative risk Attributable risk Effect size (absolute or relative)
Study Approaches Epidemiologic and behavioral: • Usually a comparison of “treatment” and “no treatment” groups • Control of confounding by behavioral and social variables
2/25/10
Pharmacologic methods Behavioral modification Reinforcing environment
Persons who smoke, usually in a health care setting General population of smokers in a commercial or quasicommercial setting
Clinical
Tools School curricula Interactive training Targeted services Mass media
Children and adolescents, usually in school Administrative groups (e.g., members of health maintenance organizations) General population Health care providers
418
Educational
Characteristics of Interventions Type of Intervention Targets
Table 14-9
46894_CH14_395_430.qxd Page 418
CHAPTER 14 SMOKING AND OTHER TOBACCO USE
46894_CH14_395_430.qxd
2/25/10
4:10 PM
Page 419
PREVENTION AND CONTROL 419
Ask the patient about smoking (Table 1).
Never-smoker
Current smoker
Former smoker
Advise the patient to quit smoking (Table 1).
Congratulate and reinforce abstinence from smoking.
Assess the patient’s willingness to quit (Table 1).
Precontemplation (i.e., patient is not ready to quit within the next six months); address why the patient is not ready to quit and continue to address tobacco use at every visit.
Contemplation (i.e., patient is not ready to quit now but is considering attempting cessation within the next six months)
Motivate the patient to quit by discussing the risks of smoking and the benefits of quitting.
Preparation (i.e., patient is willing to attempt cessation within the next 30 days)
Has tried to quit before with pharmacotherapy
Never tried to quit before or never used pharmacotherapy to quit smoking
Assist (Table 1); set quit date; consider different medication or combination pharmacotherapy (e.g., nicotine patch plus another NRT, patch plus bupropion SR [Wellbutrin SR]).
Assist (Table 1); set quit date; consider the best treatment option for the patient (e.g., over-the-counter patch or gum versus prescription spray, inhaler, or bupropion).
Failed
Patient does not have a psychiatric or alcohol problem.
Succeeded
Set quit data; consider using combination pharmacotherapy.
Arrange for follow-up (Table 1).
Failed
Patient has a psychiatric or alcohol problem.
Refer patient to a smoking cessation clinic; offer selfhelp and additional counseling information (Table 3).
Figure 14-9 Algorithm for the Identification and Treatment of Patients Who Smoke. Source: Reprinted with permission from American Family Physician, Vol 74, No 2, KS Okuyemi, NL Nollen, JS Ahluwalia, p 269. Copyright © 2006 American Academy of Family Physicians.
Community- or Population-Wide Measures Consistent with the range of strategies identified in the Surgeon General’s report of 2000, several types of community-based or population-wide measures have been proposed and implemented.54 For example, measures addressed in the Guide to Community Preventive Services include strategies to reduce tobacco use initiation by children, adolescents, and young adults; to increase tobacco cessation; and to re-
duce exposure to environmental tobacco smoke.65 Interventions recommended on the basis of “strong evidence” included, for example, smoking bans and restrictions; increasing the unit price for tobacco products; and mass media campaigns when combined with other interventions. Interventions in healthcare settings were also strongly recommended and included support for action by both providers and patients, including quitlines. (A fuller discussion of the review of
46894_CH14_395_430.qxd
4:10 PM
Page 420
CHAPTER 14 SMOKING AND OTHER TOBACCO USE
evidence by the Task Force on Community Preventive Services as well as previous reviews was presented in a prior journal supplement and is examined further in Chapter 19, “Evidence and Decision Making.”)66 A Cochrane review of community interventions for reducing smoking among adults (again, see Chapter 19) assessed controlled trials of community interventions to reduce smoking prevalence in adult smokers.67 Among the 32 studies included, designs and methods varied considerably, with smoking prevalence in repeated cross-sectional surveys being the measure of outcome in 27 of the studies. The authors concluded that “The failure of the largest and best conducted studies to detect an effect on prevalence of smoking is disappointing. A community approach will remain an important part of health promotion activities, but designers of future programmes will need to take account of this limited effect in determining the scale of projects and the resources devoted to them.”67, p 1 One of the major studies included in the foregoing review was the Community Intervention Trial for Smoking Cessation (COMMIT), conducted in the United States.68,69 In COMMIT, one community from each of 11 pairs was randomly allocated to receive interventions aimed at media and community events, health professionals, worksites, and other organizations in the community and to facilitate (except through payment for services) awareness and use of existing resources for assistance in smoking cessation. Community residents identified through survey methods as either heavy or light-to-moderate smokers comprised two cohorts both followed for five years, 1988–1993, for their responses to these community-wide interventions. Two approaches to outcome assessment were used. In the cohort follow-up analysis, a modest but statistically significant effect was found for the lightto-moderate smokers, with an estimate of 30.6% quitting in the intervention communities and 27.5% quitting in the control communities. For heavy smokers, however, the quit rate was actually slightly higher in the control than in the intervention communities (18.7% versus 18.0%). In the cross-sectional comparison analysis, differences were not significant in either category of smokers. The cohort follow-up results are presented in Figure 14-10. In COMMIT, the quit rates in the control communities reflected influences already affecting smoking rates in these communities, which intervention amplified only modestly, and only among light-tomoderate smokers. Not only scale and resources, as noted in the Cochrane review, but also the force of secular trends in candidate study communities should
35 30 25
Percent Quit
420
2/25/10
20 15 10 5 0 1988
1989
1990
1991
1992
1993
Intervention-Heavy Cohort Comparison-Heavy Cohort Intervention-Light/Moderate Cohort Comparison-Light/Moderate Cohort Note: Different numbers of subjects contribute at each time point. MCAR = missing completely at random.
Figure 14-10 Observed Quit Rates (MCAR) over Time for Heavy and Light-to-Moderate Smoker Cohorts. Source: Reprinted with permission from The COMMIT Research Group (Sylvan B Green), p 187, American Journal of Public Health, Vol 85, No 2, © 1995, American Public Health Association.
be weighed in deciding on the timing, location, and design of such studies. This interpretation is reinforced by the experience of other studies—for example, a trial of community intervention within the Minnesota Heart Health Program, and a media campaign in California.70,71 It may also be the case that in some communities today those most motivated to quit smoking have done so, increasing the challenges of further reductions in rates. School health programs that target the school population as a whole, apart from the individualoriented cessation programs noted previously, represent a major public health strategy for prevention of smoking and other tobacco use.19 The focus on youth and young adults has been a feature of smoking prevention efforts since the mid-1960s because the smok-
46894_CH14_395_430.qxd
2/25/10
4:10 PM
Page 421
PREVENTION AND CONTROL 421
ing habit begins in youth and smoking cessation in adults is often unsuccessful or only temporary. Therefore, control of tobacco use is most effective through preventive efforts before the habit begins. The 1994 report reviewed programs to prevent tobacco use initiated over the previous three decades, including school-based programs, clinical and community programs, and the role of mass media in reducing tobacco use. Many specific studies were discussed. Policies were also reviewed relating to smoking restrictions for the general public, restriction on minors’ access to tobacco, warning labels on tobacco products, and tobacco taxation. Given the strong emphasis on prevention programs targeting youth, with the goal of averting the onset of tobacco use in the first place, school-based programs were identified as having a central role. These preventive strategies can be linked with the stages of adoption of tobacco use indicated in Figure 14-1. At the stage of “never smoker,” mass media programming, counter-advertising, and community-wide programs are applicable. Once “trying” of tobacco products has begun, programs to instill skills at resisting negative social influences, counterincentives through price increases, and restrictions on access become important. At the further stage of “experimentation,” broader social influences are considered to have added benefit along with continued pricing counterincentives. Once “regular use” has become established, further restrictive policies such as prohibition of smoking in the school setting and offering tobacco cessation programs have been attempted. The report concluded a review of programs to prevent tobacco use by noting the diversity of studies, complexity of design issues, and variable degrees of program implementation that have resulted in great heterogeneity in the available data. Overall, however, the programs to inoculate against adverse social influences such as peer pressure have had a favorable impact, resulting in reductions in prevalence of smoking by some 25–60% in groups receiving these programs relative to control groups, with differences lasting from 1 to 4 years. The duration of effect of programs targeting school-age groups appears to be longer when periodic “booster” programs are offered, prevention of tobacco use is incorporated in broader health curricula, and community-wide programs provide a reinforcing context for school-based programs. Guidelines for school health programs to prevent tobacco use were formulated by the Centers for Disease Control and Prevention in collaboration with representatives of a large number of organizations and agencies concerned with prevention of tobacco
use.72 The seven recommendations addressed: (1) development and enforcement of school policies on tobacco use; (2) instruction about health effects and social influences on smoking; (3) grades K–12 prevention education; (4) teacher training; (5) engaging parental and family support for school programs; (6) supporting cessation efforts for staff as well as students; and (7) periodic assessment of the program. The report concluded that such programs “could become one of the most effective national strategies to reduce the burden of physical, emotional, and monetary expense incurred by tobacco use.”72, p 359 It closed with the caveat that effectiveness of this strategy is conditional on commitment of school and community leaders both to implement and to sustain such programs. The need for long-term support was emphasized. In the United States, state health departments have become a major component of tobacco control efforts. With early state-funded programs in California and Massachusetts, sufficient experience was gained by 1999 for the Centers for Disease Control and Prevention to launch Best Practices for Comprehensive Tobacco Control Programs—an evidence-based guide for program development with state-by-state funding formulas that has now been updated to 2007.73 The guide identifies five components of a comprehensive program: state and community interventions, health communication interventions, cessation interventions, surveillance and evaluation, and administration and management. Best Practices presents goal funding levels for each component and proposes a tobacco control budget for each state, including base costs and per capita additions taking state population data into account. Recommended per capita funding is $12.34 for the United States overall and ranges from $9–$18 per state. The total annual cost is $3.6 billion for the entire United States, from $9–$442 million per state. For each state, allocation of recommended funding across the five program components is indicated. Cost of the program would be fully supported by tax revenues and payments from the Master Settlement Agreement. Under the Agreement, the US tobacco industry must make payments to states in compensation for their liability for medical care costs from tobaccorelated illnesses. Program costs range from 9–62% of these revenues from state to state. The Institute of Medicine report recommended foremost that each state should fund tobacco control efforts at the level presented in the Best Practices document.56 At the national level in the United States, regulation of tobacco products through the authority of federal agencies is a long-sought objective in the
46894_CH14_395_430.qxd
422
2/25/10
4:10 PM
Page 422
CHAPTER 14 SMOKING AND OTHER TOBACCO USE
tobacco control arena. How this might be accomplished has been a continuing point of contention involving public health, the tobacco industry, and the courts. Two companion publications in 2000, following a ruling by the Supreme Court that the Food and Drug Administration (FDA) did not have legal jurisdiction over tobacco, argued on the one hand for Congress to grant such authority to the FDA and on the other to create a Tobacco Control Agency with exclusive responsibility for this role.74,75 This issue has only recently been resolved by assignment of regulatory authority to the FDA. However, state and local regulatory authorities have taken action in a growing number of instances. By 2000, the Surgeon General’s report documented a sharp rise in the number of tobacco control ordinances for clean indoor air throughout the 1980s and 1990s, reaching 500–700 communities in each category of workplaces, restaurants, and public places.54 Subsequent reports gave evidence of decreased secondhand smoke exposure, validated by serial surveys of cotinine concentration in nonsmoking persons— both nationally, based on the National Health and Nutrition Examination Surveys, and in localities in New York State, based on the New York Adult Tobacco Survey.76,77 In 2004, reports began to appear that enactment of clean indoor regulations, whether at community, state, or national levels, was followed by abrupt declines in numbers of hospitalizations for acute myocardial infarction. The first such report came from
Helena, Montana, where a law requiring smoke-free workplaces and public places was enacted, enforced for 6 months, then suspended by court order after a legal challenge.78 Compared with the same calendar period the previous year, 40% fewer such admissions were recorded; after suspension of the law, the number increased toward prior levels. The Helena experience was soon followed by others, from Pueblo, Colorado; the Piedmont region of Italy; Bowling Green, Ohio; New York State; Ireland; Saskatoon, Canada; and Rome, Italy. These eight reports were summarized in a meta-analysis showing an overall estimate of a 19% (95% CI, 14–24%) immediate decrease in hospital admission rates for acute myocardial infarction (Figure 14-11).79 Among still more recent reports, one from Scotland further indicates that the reduction in admissions was accounted for mainly (67%) by nonsmokers, with the remainder by smokers.80 Evidence appears strong that such regulatory action, over a wide range from local to national coverage, can have a significant impact not only on secondhand smoke exposure but also on morbidity from acute coronary events. Global Strategies Global strategies to reduce both demand and supply of tobacco products have been advocated for several decades. WHO has issued several reports on the tobacco problem worldwide and specifically in relation to developing countries. The history of this activity through the early 1980s was reviewed by Roemer.81 She
Study
%
ID
ES (95% CI)
Weight
Helena Montana
0.60 (0.21, 0.99) 1.76
Pueblo Colorado
0.73 (0.63, 0.85) 10.13
Piedmont Italy
0.89 (0.81, 0.98) 12.14
Bowling Green Ohio
0.61 (0.55, 0.67) 14.24
New York State
0.80 (0.80, 0.80) 17.20
Ireland
0.89 (0.81, 0.97) 12.56
Saskatoon Canada
0.87 (0.84, 0.90) 16.35
Rome Italy
0.89 (0.85, 0.93) 15.61
Overall
0.81 (0.76, 0.86) 100.00
NOTE: Weights are from random effects analysis 0
1
Figure 14-11 Summary of Studies of the Effects of Smokefree Laws on Acute Myocardial Infarction. Source: Reprinted with permission from Preventive Medicine, Vol 47, Glantz S, pp 452–453. © 2008 Published by Elsevier Inc.
46894_CH14_395_430.qxd
2/25/10
4:10 PM
Page 423
PREVENTION AND CONTROL 423
cited national findings as well in the United Kingdom and Sweden that antedated the US Surgeon General’s Advisory Committee report of 1964, reviewed actions by WHO, and chronicled and analyzed legislative actions against tobacco throughout the world. In 1970, the World Health Assembly first adopted a resolution on this issue as the basis for the antismoking policy of WHO. The first WHO Expert Committee to consider this topic recommended that governments should assess the presence or potential for smoking-related health problems and, if findings were affirmative, institute programs for control and prevention of tobacco smoking.51 Programs should be planned as long-term activities. They should include educational efforts and target two groups in particular: women and health workers. These groups were selected because of the risks of smoking during pregnancy and the specific role models that nonsmoking women and healthcare workers could become. Legislative action should be taken to address advertising and promotion, labeling and warnings, taxation, protection of minors from distribution of or access to tobacco products, and securing the rights of nonsmokers. Measures to support smoking cessation should be adopted. Finally, research should be pursued addressing a range of topics from mechanisms of toxicity to determination of the country-specific economic burden of smoking. The Expert Committee Reports of 1979 and 1983, with their heightened urgency, were noted previously.52,53 Concern about smoking in youth was reflected in the later report Prevention in Childhood and Youth of Adult Cardiovascular Diseases: Time for Action, which recommended that the overall goal of national policies should be “the elimination of smoking and other forms of tobacco use.”82, p 83 Specific strategies were school-based prevention and control activities, prohibition of the sale of cigarettes to minors, prohibition of the advertising and promotion of tobacco products, disallowing tobacco industry sponsorship of sporting events, discouraging passive smoke exposure of children by their parents who smoke, legislation to prohibit the promotion of smokeless tobacco, and encouragement of parents to become nonsmoking role models. A landmark development in this long history of policy development and advocacy by WHO is the creation and implementation of the first global treaty on health, the WHO Framework Convention for Tobacco Control (FCTC).83 The FCTC is described in its Foreword as “a paradigm shift in developing a regulatory strategy to address addictive substances; in contrast to previous drug control treaties, the WHO
FCTC asserts the importance of demand reduction strategies as well as supply issues.” Further:83, p v The WHO FCTC was developed in response to the globalization of the tobacco epidemic. The spread of the tobacco epidemic is facilitated through a variety of complex factors with crossborder effects, including trade liberalization and direct foreign investment. Other factors such as global marketing, transnational tobacco advertising, promotion and sponsorship, and the international movement of contraband and counterfeit cigarettes have also contributed to the explosive increase in tobacco use. The FCTC addresses price and nonprice controls to reduce demand; illicit trade, sales to minors, and support for economically viable alternatives to tobacco production to reduce supply; protection of the environment; criminal and civil liability; scientific and technical cooperation and communication; institutional arrangements and financial resources; and other operational issues. The treaty entered into force February 27, 2005, and had, as of December 2008, 168 signatories and 161 parties (excludes countries which have signed but not ratified the treaty, of which there were seven, including the United States).84 The potential impact of the FCTC for developing countries is illustrated in the context of national tobacco control efforts in an extensive Report on Tobacco Control in India, the eighth country to ratify the treaty, in 2004.85 India had previously adopted its own national legislation, The Cigarettes and Other Tobacco Products Act, 2003, which is in some respects more stringent than the FCTC. Despite being a major tobacco producer, India became a significant force behind development and adoption of the FCTC. Among its benefits for India, in addition to implied cooperation in tobacco control by surrounding countries, are the further measures that must be implemented for full conformity, strengthening areas less adequately addressed in the 2003 Act. Another country-specific example is China.86 Also a tobacco producer that early ratified the FCTC, the Chinese government received 7.6% of its total revenue for 2005 from tobacco sales and taxes. The public health mandate to reduce demand and supply of tobacco products, represented by the FCTC, and the economic pressures to maintain and expand revenues, especially in areas where 50% of provincial income is from tobacco, create a strong tension. Support for action to implement provisions of the treaty globally has been committed by the Bloomberg Global Initiative to Reduce Tobacco Use, a $125 million
46894_CH14_395_430.qxd
424
2/25/10
4:10 PM
Page 424
CHAPTER 14 SMOKING AND OTHER TOBACCO USE
program that doubles the available funds for tobacco control in low- and middle-income countries. The WHO Report on the Global Tobacco Epidemic, 2008, subtitled The MPOWER Package, presents its six-part package of proven tobacco control policies to enable Member States to implement the FCTC.1 Its conclusion:1, p 59 “Unless urgent action is taken, more than 1 billion people could be killed by tobacco during this century. But this dire future can be changed by the leaders of governments and civil society. As the tobacco epidemic is entirely manmade, the end of the tobacco epidemic must also be manmade. We must act now.”
CURRENT ISSUES “Harm Reduction”: An Improbable Product? The concept that smoking might become safer through development of some less hazardous substitute has come to be discussed as “harm reduction.” One approach was to devise and market lower-tar or filtered cigarettes. The history of this approach is recounted in a 2005 review, with the conclusion that these products caused a higher prevalence of smoking and burden of disability and death than if they had not been introduced and then accepted by the public in the belief that risk was actually reduced.87 Issues in determining the true benefit, if any, of alternative tobacco products, were outlined: (1) How to assess risk reduction for the individual consumer—e.g., is it feasible to evaluate potential long-term individual health consequences of an alternative product? (2) How to assess population effects—e.g., how much increase in prevalence of exposure would be needed to offset any reduction in individual risk? (3) How to inform health professionals and the public about such products in ways to avoid nonusers of tobacco products from taking them up. Lastly, (4) How should such products be regulated—e.g., what should be the approach, who should pay, and what enforcement provisions would suffice? The lack of answers to these questions makes establishment of individual safety and public health acceptability of such products seem improbable. More recently, the case for medicinal nicotine products to replace tobacco use for heavily addicted users has also been reviewed.88 Although such products evidently have been incorporated in practice for smoking cessation, there remains controversy over use of these and smokeless tobacco products as a public health strategy.
Advancing the Science: Balanced Assessment? The role of the tobacco industry with respect to the science regarding secondhand smoke exposure and its health consequences has been described as attempting to undermine the evidence; fighting smoke-free regulations; and designing and interpreting their own cardiovascular studies based on preservation of corporate viability.89 Efforts of the industry to conceal and misrepresent their own science have been publicly exposed.90 On the other hand, it has been charged that “. . . the purported risks posed by ETS have been used to justify draconian regulations that criminalize and marginalize lawful citizens, pitting children against their parents, spouses against spouses, and people against people to the point of raising homicidal animosities against smokers.”91 It appears that translating the science of secondhand smoke reduction into practice, a major current issue in tobacco control, will call for continued attention to sound interpretation and accurate communication to the public and policy makers. Achieving Regulation: Global as Well as Local? Demonstration of the rapid and striking impact of clean indoor air regulations, from local to national levels in the United States and elsewhere, is highly encouraging and offers the prospect of widening coverage of populations by such protection. Still, progress is not to be assumed, and it is important to ask the question, Can the global tobacco pandemic be controlled? This is the global question of greatest importance concerning smoking and other tobacco use, because of the projected increases in smoking prevalence and its vast health consequences in the decades ahead. Pressures to reduce tobacco consumption in the United States and several other industrialized countries have had the negative effect of intensified marketing of tobacco products elsewhere, especially in developing countries. In many of these countries, efforts now bolstered by the FCTC are in progress to adopt legal controls on tobacco sales and imports. However, the countervailing influences are powerful and will be difficult to overcome. Still, as recently observed:92, p 1498 The first few years of the 21st century have made possible what was once considered impossible. In the face of an escalating pandemic, a global haze may be starting to lift. We are witnessing a public health evolution in which the once-extraordinary is rapidly becoming the social norm. Making smoking history moves us closer to reaffirming the right to the highest standard of human health for all.
46894_CH14_395_430.qxd
2/25/10
4:10 PM
Page 425
REFERENCES 425
REFERENCES 1. WHO Report on the Global Tobacco Epidemic, 2008: The MPOWER Package. Geneva, World Health Organization; 2008. 2. Peto R. Smoking and death: the past 40 years and the next 40. Br Med J. 1994;209:937–939. 3. Doll R, Peto R, Wheatley K, et al. Mortality in relation to smoking: 40 years’ observations on male British doctors. Br Med J. 1994;309: 901–911. 4. Asplund K. Smokeless tobacco and cardiovascular disease. Prog Cardiovasc Dis. 2003;45: 383–394. 5. US Department of Health and Human Services. The Health Consequences of Involuntary Exposure to Tobacco Smoke: A Report of the Surgeon General. Atlanta, GA: US Department of Health and Human Services, Centers for Disease Control and Prevention, Coordinating Center for Health Promotion, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health; 2006. 6. Tate C. Cigarette Wars: The Triumph of “The Little White Slaver.” New York, NY: Oxford University Press; 1999. 7. Kluger R. Ashes to Ashes. America’s HundredYear Cigarette War, the Public Health, and the Unabashed Triumph of Philip Morris. New York, NY: Vintage Books; 1997. 8. US Department of Health, Education and Welfare. Smoking and Health: Report of the Advisory Committee to the Surgeon General. Public Health Service Publication No. 1103. Washington, DC: US Department of Health, Education and Welfare, Public Health Service; 1964. 9. US Department of Health and Human Services. Reducing the Health Consequences of Smoking. 25 Years of Progress. A Report of the Surgeon General, 1989. Atlanta, GA: US Department of Health and Human Services, Public Health Service, Centers for Disease Control, Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health; 1989.
10. Murray CLJ, Lopez AD. Quantifying the burden of disease and injury attributable to ten major risk factors. In: Murray CJL, Lopez AD eds. The Global Burden of Disease: A Comprehensive Assessment of Mortality and Disability from Diseases, Injuries, and Risk Factors in 1990 and Projected to 2020. Boston, MA: The Harvard School of Public Health; 1996: 295–395. 11. Stanley K. Appendix A. Control of tobacco production and use. In: Jamison DT, Mosley WH, Measham AR, Bobadilla JL, eds. Disease Control Priorities in Developing Countries. Oxford (England): Oxford University Press; 1993:703–723. 12. Jha P, Chaloupka FJ, Moore J, et al. Tobacco addiction. In: DT Jamison et al., eds. Disease Control Priorities in Developing Countries. 2nd ed. International Bank for Reconstruction and Development/The World Bank. Washington, DC; 2006:869–885. 13. World Health Organization. WHO Framework Convention on Tobacco Control. Geneva (Switzerland): World Health Organization; 2003. 14. Reddy KS, Gupta PC. Report on Tobacco Control in India. New Delhi, India: Ministry of Health & Family Welfare; 2004. 15. Frost F, Tollestrup K, Starzyk P. History of smoking from the Washington state death certificate. Am J Prev Med. 1994;10:335–339. 16. Wagenknecht LE, Burke GL, Perkins LL, et al. Misclassification of smoking status in the CARDIA Study: a comparison of self-report with serum cotinine levels. Am J Public Health. 1992;82:33–36. 17. Repace JL. Exposure to secondhand smoke. In: Ott WR, Steineman AC, Wallace LA, eds. Exposure Analysis. Boca Raton, FL: CRC Press, Taylor & Francis Group; 2007. 18. National Cancer Institute. The Role of the Media in Promoting and Reducing Tobacco Use. Tobacco Control Monograph No. 19. NIH Pub. No. 07-6242. Bethesda, MD: US Department of Health and Human Services,
46894_CH14_395_430.qxd
426
2/25/10
4:10 PM
Page 426
CHAPTER 14 SMOKING AND OTHER TOBACCO USE
National Institutes of Health, National Cancer Institute; June 2008. 19. US Department of Health and Human Services. Preventing Tobacco Use among Young People: A Report of the Surgeon General. Atlanta, GA: Office on Smoking and Health, National Center for Chronic Disease Prevention and Health Promotion, Centers for Disease Control and Prevention, Public Health Service, US Department of Health and Human Services; 1994. 20. Ambrose JA, Barua RS. The pathophysiology of cigarette smoking and cardiovascular disease: an update. J Am Coll Cardiol. 2004;43: 1731–1737.
28. Tate C. The butt of controversy. Smithsonian. 1989;20:107–117. 29. Doll R. Smoking and health. www.pitt.edu/ ~super1/lecture/lec2047/001.htm. Accessed March 14, 2008. 30. US Department of Health and Human Services. The Health Consequences of Smoking: Cardiovascular Disease. A Report of the Surgeon General. Rockville, MD: US Department of Health and Human Services, Public Health Service, Office on Smoking and Health; 1983.
21. Bakhru A, Erlinger TP. Smoking cessation and cardiovascular disease risk factors: results from the Third National Health and Nutrition Examination Survey. PLoS Medicine. 2005; 2(6):e160.
31. US Department of Health and Human Services. The Health Consequences of Smoking: A Report of the Surgeon General. Atlanta, GA: US Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health; 2004.
22. Venn A, Britton J. Exposure to secondhand smoke and biomarkers of cardiovascular disease risk in never-smoking adults. Circulation. 2007;115:990–995.
32. Keys A. Seven Countries: A Multivariate Analysis of Death and Coronary Heart Disease. Cambridge, MA: Harvard University Press; 1980.
23. Centers for Disease Control and Prevention. Cigarette smoking among adults—United States, 2006. MMWR. 2007;56:1157–1161.
33. Pooling Project Research Group. Relationship of blood pressure, serum cholesterol, smoking habit, relative weight and ECG abnormalities to incidence of major coronary events: final report of the Pooling Project. J Chronic Dis. 1978;31:201–306.
24. Centers for Disease Control and Prevention. Cigarette use among high school students— United States, 1991–2007. MMWR. 2008;57: 686–688. 25. Centers for Disease Control and Prevention. Disparities in secondhand smoke exposure— United States, 1988–1994 and 1999–2004. MMWR. 2008;57:744–747. 26. Centers for Disease Control and Prevention. Global Youth Tobacco Surveillance, 2000–2007. Surveillance Summaries. MMWR. 2008;57 (No. SS-1). 27. National Center for Health Statistics. Health, United States, 2007. With Chartbook on Trends in the Health of Americans. Hyattsville, MD; 2007. www.cdc.gov/nchs/data/hus/hus07 .pdf. Accessed December 20, 2008.
34. Thun MJ, Myers DG, Day-Lally C, Namboodiri MM, et al. Age and the exposure-response relationships between cigarette smoking and premature death in Cancer Prevention Study II. In: Burns DM, Garfinkel L, Samet JM, eds. Changes in Cigarette-Related Disease Risks and Their Implication for Prevention and Control. Smoking and Tobacco Control, Monograph 8. NIH publication 97-4213. Bethesda, MD: National Cancer Institute, National Institutes of Health; 1997:383–413. 35. Kawachi I, Colditz GA, Stampfer MJ, et al. Smoking cessation and decreased risks of total mortality, stroke, and coronary heart disease incidence among women: a prospective cohort study. In: Burns DM, Garfinkel L, Samet JM, eds. Changes in Cigarette-Related Disease
46894_CH14_395_430.qxd
2/25/10
4:10 PM
Page 427
REFERENCES 427
Risks and Their Implication for Prevention and Control. Smoking and Tobacco Control, Monograph 8. NIH publication 97-4213. Bethesda, MD: National Cancer Institute, National Institutes of Health; 1997. 36. Powell JT. Smoking. In: Fowkes FGR, ed. Epidemiology of Peripheral Vascular Disease. London: Springer-Verlag; 1991:141–153. 37. Parish S, Collins R, Peto R, et al. Cigarette smoking, tar yields, and non-fatal myocardial infarction: 14,000 cases and 32,000 controls in the United Kingdom. Brit Med J. 1995;311: 471–477. 38. Teo KK, Ounpuu S, Hawken S, et al. Tobacco use and risk of myocardial infarction in 52 countries in the INTERHEART study: a casecontrol study. Lancet. 2006;368:647–658. 39. He J, Vupputuri S, Allen K, Prerost MR, Hughes J, Whelton PK. Passive smoking and the risk of coronary heart disease––a metaanalysis of epidemiologic studies. N Engl J Med. 1999;340:920–926. 40. Whincup PH, Gilg JA, Emberson JR, et al. Passive smoking and risk of coronary heart disease and stroke: prospective study with cotinine measurement. Brit Med J. doi:10.1136/ bmj.38146.427188.55 (published 30 June 2004). 41. Zieske AW, McMahan CA, McGill HC Jr, et al. Smoking is associated with advanced coronary atherosclerosis in youth. Atherosclerosis. 2005;180:87–92. 42. Critchley JA, Unal B. Is smokeless tobacco a risk factor for coronary heart disease? A systematic review of epidemiological studies. Eur J Cardiovasc Prevention Rehab. 2004;11: 101–112. 43. Asplund K, Nasic S, Janlert U, Stegmayr B. Smokeless tobacco as a possible risk factor for stroke in men. A nested case-control study. Stroke. 2003;34:1754–1759. 44. Doll R, Peto R, Boreham J, Sutherland I. Mortality in relation to smoking: 50 years’ observations on male British doctors. BMJ.
2004;328:1519. doi:10.1136/bmj.38142 .554479.AE (published June 22, 2004). 45. Mazzati E, Vander Hoorn S, Lopez AD, et al. Comparative quantification of mortality and burden of disease attributable to selected risk factors. In: Lopez AD, Mathers CD, Ezzati M, Jamison DT, Murray CJL, eds. Global Burden of Disease and Risk Factors. Washington, DC: The International Bank for Reconstruction and Development/The World Bank; 2006:241–396. 46. World Health Organization. The World Health Report 2002. Reducing Risks, Promoting Healthy Life. Geneva: World Health Organization; 2002. 47. Hozawa A, Okamura T, Kadowaki T, et al. Is weak association between cigarette smoking and cardiovascular disease mortality observed in Japan explained by low total cholesterol?–– NIPPON DATA80. Int J Epidemiol. 2007;36: 1060–1067. 48. Stamler J. Established major coronary risk factors. In: Marmot M, Elliott P, eds. Coronary Heart Disease Epidemiology: From Aetiology to Public Health. Oxford, England: Oxford University Press; 1992:35–66. 49. Reid DD, McCartney P, Hamilton PJS, Rose G, Jarrett RJ, Keen H. Smoking and other risk factors for coronary heart-disease in British civil servants. Lancet. 1976;November 6: 979–984. 50. Katz LN, Allen EV, Cherkasky M, et al. Cigarette smoking and cardiovascular diseases: report by the American Heart Association. Circulation. 1960;22:160–166. 51. World Health Organization Expert Committee. Smoking and Its Effects on Health. WHO Technical Report Series 568. Geneva (Switzerland): World Health Organization; 1975. 52. World Health Organization Expert Committee. Controlling the Smoking Epidemic. WHO Technical Report Series 636. Geneva (Switzerland): World Health Organization; 1979. 53. World Health Organization Expert Committee. Smoking Control Strategies in Developing
46894_CH14_395_430.qxd
428
2/25/10
4:10 PM
Page 428
CHAPTER 14 SMOKING AND OTHER TOBACCO USE
Countries. WHO Technical Report Series 695. Geneva (Switzerland): World Health Organization; 1983. 54. US Department of Health and Human Services. Reducing Tobacco Use: A Report of the Surgeon General. Atlanta, GA: US Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health, 2000. Reprinted with corrections, October 2000. 55. Backinger CL, O’Connell ME. Developing consensus on tobacco control and research. Am J Prev Med. 2007;33(6S):S311–S313. 56. IOM (Institute of Medicine). Ending the Tobacco Problem: A Blueprint for the Nation. Washington, DC: The National Academies Press; 2007. 57. Rose G, Hamilton PJS, Colwell L, Shipley MJ. A randomized controlled trial of anti-smoking advice: 10-year results. J Epidemiol Community Health. 1982;36:102–108. 58. Critchley JA, Capewell S. Mortality risk reduction associated with smoking cessation in patients with coronary heart disease. JAMA. 2003;290:86–97. 59. Ranney L, Melvin C, Lux L, McClain E, Lohr KN. Systematic review: smoking cessation intervention strategies for adults and adults in special populations. Ann Intern Med. 2006;145:845–856. 60. US Preventive Services Task Force. The Guide to Clinical Preventive Services 2006. Recommendations of the US Preventive Services Task Force. Washington, DC: Agency for Healthcare Research and Quality. 2006. www.ahrq.gov/clinic/uspstf/uspstbac.htm. Accessed October 14, 2007. 61. Okuyemi KS, Nollen NL, Ahluwalia JS. Interventions to facilitate smoking cessation. Am Fam Physician. 2006;74:262–271. 62. Nides M, Leischow S, Sarna L, Evans SE. Maximizing smoking cessation in clinical prac-
tice: pharmacologic and behavioral interventions. Prev Cardiol. 2007;10(2 Suppl 1):23–30. 63. Keller PA, Bailey LA, Koss KJ, Baker TB, Fiore MC. Organization, financing, promotion, and cost of U.S. quitlines. Am J Prev Med. 2007; 32:32–37. 64. Curry SJ, Emery S, Sporer AK, et al. A national survey of tobacco cessation programs for youths. Am J Public Health. 2007;97:171–177. 65. Tobacco. Guide to Community Preventive Services Website. Centers for Disease Control and Prevention. www.thecommunityguide.org/ tobacco/. Last updated June 14, 2005. Accessed September 25, 2007. 66. Hopkins DP, Fielding JE, Task Force on Community Preventive Services. The Guide to Preventive Services: Tobacco Use Prevention and Control. Reviews, Recommendations, and Expert Commentary. Am J Prev Med. 2001; 20(2 Suppl). 67. Secker-Walker RH, Gnich W, Platt S, Lancaster T. Community interventions for reducing smoking among adults. The Cochrane Database of Systematic Reviews 2002, Issue 2. Art. No.: CD001745. doi: 10.1002/14651858 .CD001745. 68. The COMMIT Research Group. Community intervention trial for smoking cessation (COMMIT), I: cohort results from a four-year community intervention. Am J Public Health. 1995;85:183–192. 69. The COMMIT Research Group. Community intervention trial for smoking cessation (COMMIT), II: changes in adult cigarette smoking prevalence. Am J Public Health. 1995;85:193–200. 70. Lando HA, Pechacek TF, Pirie PL, et al. Changes in adult cigarette smoking in the Minnesota Heart Health Program. Am J Public Health. 1995;85:201–208. 71. Popham WJ, Potter LD, Hetrick MA, et al. Effectiveness of the California 1990–1991 tobacco education media campaign. Am J Prev Med. 1994;10:319–326.
46894_CH14_395_430.qxd
2/25/10
4:10 PM
Page 429
REFERENCES 429
72. Guidelines for school health programs to prevent tobacco use and addiction. J School Health. 1994;64:353–360. 73. Centers for Disease Control and Prevention. Best Practices for Comprehensive Tobacco Control Programs—2007. Atlanta, GA: US Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health; October 2007.
82. WHO Expert Committee on Prevention in Childhood and Youth of Adult Cardiovascular Diseases. Prevention in Childhood and Youth of Adult Cardiovascular Diseases: Time for Action. Technical Report Series 792. Geneva (Switzerland): World Health Organization; 1990. 83. World Health Organization. WHO Framework Convention on Tobacco Control. Geneva (Switzerland): World Health Organization; 2003 (updated reprint 2005).
74. Myers ML. Protecting the public health by strengthening the Food and Drug Administration’s authority over tobacco products. New Engl J Med. 2000;343:1806–1809.
84. WHO Framework Convention on Tobacco Control. Full List of Signatories and Parties to the WHO Framework Convention on Tobacco Control. www.who.int/fctc/signatories_parties/ en/print.html. Accessed December 19, 2008.
75. Glantz LH, Annas GJ. Tobacco, the Food and Drug Administration, and Congress. New Engl J Med. 2000;343:1802–1806.
85. Reddy KS, Gupta PC, eds. Report on Tobacco Control in India. New Delhi, India: Ministry of Health & Family Welfare; 2004.
76. Pickett MS, Schober SE, Brody DJ, Curtin LR, Giovino GA. Smoke-free laws and secondhand smoke exposure in US non-smoking adults, 1999–2002. Tobacco Control. 2006;15: 302–307.
86. Wright AA, Katz IT. Tobacco tightrope— balancing disease prevention and economic development in China. New Engl J Med. 2007; 356:1493–1496.
77. Centers for Disease Control and Prevention. Reduced secondhand smoke exposure after implementation of a comprehensive statewide smoking ban—New York, June 26, 2003– June 30, 2004. MMWR. 2007;56:705–708. 78. Sargent RP, Shepard RM, Glantz SA. Reduced incidence of admissions for myocardial infarction associated with public smoking ban: before and after study. BMJ. 2004;328:977–980. 79. Glantz SA. Meta-analysis of the effects of smokefree laws on acute myocardial infarction: an update. (Letter) Prev Med. 2008;47: 452–453.
87. Warner KE. Will the next generation of “safer” cigarettes be safer? J Pediatr Hematol Oncol. 2005;27:543–550. 88. Britton J, Edwards R. Tobacco smoking, harm reduction, and nicotine product regulation. Published online October 5, 2007. www .thelancet.com doi:10.1016/S0140-6736(07) 61482-2. 89. Tong EK, Glantz SA. Tobacco industry efforts undermining evidence linking secondhand smoke with cardiovascular disease. Circulation. 2007;116:1845–1854. 90. Glantz SA, Barnes DE, Bero L, et al. Looking through a keyhole at the tobacco industry: the Brown and Williamson Documents: Special Communications. JAMA. 1995;274:219–224.
80. Pell JP, Haw S, Cobbe S, et al. Smoke-free legislation and hospitalizations for acute coronary syndromes. New Engl J Med. 2008;359: 482–491.
91. Gori GB. Stoking the rigged terror of secondhand smoke. Regulation. 2007;Spring:14–17.
81. Roemer R. Legislative Action to Combat the World Smoking Epidemic. Geneva (Switzerland): World Health Organization; 1982.
92. Koh HK, Joossens LX, Connolly GN. Making smoking history worldwide. New Engl J Med. 2007;356:1496–1498.
46894_CH14_395_430.qxd
2/25/10
4:10 PM
Page 430
46894_CH15_431_502.qxd
2/26/10
12:31 PM
Page 431
C H A P T E R
15 Other Personal Factors policy dilemma. Much, if not all, of the epidemiologic evidence supports such an association, and mechanisms of action such as increased concentration of high-density lipoprotein cholesterol make a causal relation plausible. However, no clinical trials have been conducted to provide the kind of evidence regarding alcohol that is usually required as a basis for recommending an intervention. There is some controversy, therefore, as to appropriate policy. More prominent on a global level is the concern and call to action to oppose marketing influences that both promote increased alcohol consumption and pose unquestioned health hazards and other social costs.
SUMMARY The major established cardiovascular risk factors, as well as the potential for discovering genetic factors of public health importance, are addressed in Chapters 7–14. Countless additional personal characteristics have been investigated with respect to their possible relation to atherosclerotic and hypertensive cardiovascular diseases. Hopkins and Williams found 246 of them to consider in a 1981 review, and the number has doubtless since increased, perhaps several fold (cited in Chapter 17, “What Causes Cardiovascular Diseases?”). Three areas of particular interest—alcohol consumption, adverse psychosocial factors, and hemostatic factors—warrant particular attention and are addressed here, necessarily in less detail than topics of the preceding chapters. Briefer still is comment on several other personal factors that may be considered as “evolving” (not entirely new, but with recent developments of interest) or “emerging” (more recently investigated or less clearly understood but potentially important). Beyond these individual-level factors, the final chapter of Part III addresses some population-level social and environmental conditions relevant to epidemiology and prevention of cardiovascular diseases.
2. Adverse Psychosocial Factors Like metabolic or physiologic factors, psychosocial factors pose their particular challenges for research. For population studies, standardization of definitions, classification, and methods of observation and measurement is essential for comparability across studies and for consistent interpretation of results. These qualities have been difficult to achieve in connection with many of the concepts investigated in the area of psychosocial factors. Nonetheless, much of the work to date in this area supports the general theoretical view that socially conditioned stimuli have psychological effects and, through these, neurohumoral and other mechanistic effects that may increase (or decrease) susceptibility to disease, both generally and specifically. Four areas most extensively or recently investigated are Type A behavior pattern, occupational stress, depression, and social support. The concepts underlying these areas of research and some of the results of particular studies are discussed as illustrations of this area of research as a whole. Many observations support the general theory as it applies
1. Alcohol Consumption Although alcohol consumption may be considered a component of diet, it is addressed separately owing to special interest in the association of “low to moderate” levels of intake, in comparison with abstinence or higher intakes, with reduced risks of coronary heart disease and ischemic stroke. Because of many known adverse effects of alcohol from health and social perspectives, this U- or J-shaped relation poses a 431
46894_CH15_431_502.qxd
432
2/26/10
12:31 PM
Page 432
CHAPTER 15 OTHER PERSONAL FACTORS
to cardiovascular diseases, but others do not, and their synthesis is difficult owing to the wide diversity of concepts and methods employed. Recent literature lacks the far-reaching and integrative character of some provocative papers of two decades or more ago. It reflects instead both a greater concentration on methodologic issues and further branching into new concepts. However, convergence of work in two areas seems promising: understanding of causation and prevention of depression, with its adverse cardiovascular effects, as one well-defined and extensively studied factor; and potential for wider adoption and replication of a standard composite instrument to assess multiple dimensions of psychosocial status in diverse populations. 3. Hemostatic Factors Blood coagulation is a protective process when blood loss threatens and an adverse one when thrombosis occurs within a vessel such as a coronary artery and obstructs blood flow. Thrombosis has been known for more than 80 years to be linked with coronary heart disease and other manifestations of advanced atherosclerosis. Only relatively recently, however, have epidemiologic studies of selected components of the hemostatic process been conducted. Three phases–– blood vessel status, blood platelet function, and coagulation phases contribute jointly to this process. Epidemiologic studies relate mainly to the last of these. Fibrinogen, factor VII, and other hemostatic factors have been found to be associated with many other established risk factors and with risks of coronary heart disease, stroke, and peripheral arterial disease. Genetic epidemiology has demonstrated geneenvironment interactions involving some coagulation factors. Associations found in some populations or groups may depend on such interactions and not be generalizable to others. Trials of drugs to modify platelet function or promote resolution of thrombi have shown benefit in high-risk persons, such as those actively developing an acute myocardial infarction or stroke or who are at high risk of recurrences. Less clear-cut is whether benefits of antiplatelet drugs, principally aspirin, outweigh risks of serious bleeding when applied in primary prevention. Much remains to be learned about the extent of an independent causal role for any of the coagulation factors, although the case for fibrinogen is fairly strong. Meanwhile, to apply current knowledge about antiplatelet and thrombolytic therapy in acute cardiovascular episodes, if the potential benefit of these interventions is to be realized for those eligible to receive them, requires changes in timeliness of the victim’s response and in physician and hospital practice.
4. Evolving and Emerging Factors There remain several personal factors or characteristics whose importance or special interest warrants brief review. The concept of combination pharmacotherapy, known best in reference to the “Polypill,” has been proposed for more than a decade and has received preliminary evaluation in a trial of acceptability, feasibility, intermediate effects, and tolerability with outcomes that support further investigation. Hormone replacement therapy (HRT) offers a recent example of well-established expectations of cardiovascular benefit based on observational epidemiologic studies, where evidence from randomized controlled trials indicates greater harm than benefit. Infection by a few specific agents continues to be investigated as an underlying etiology of atherosclerosis, and recognition of the toll of morbidity and mortality due to influenza virus infection among persons with cardiovascular disease calls attention to the special need for influenza immunization in this high-risk group. HIV/AIDS is found associated with metabolic and anthropometric factors that contribute to atherosclerosis and its complications and identifies persons with HIV/ AIDS, and especially those receiving highly active antiretroviral therapy, as a group of special concern in cardiovascular disease prevention. Antioxidants continue to be regarded as important in contributing to mechanisms of atherosclerosis but, as with HRT, expectations of benefit of antioxidant dietary supplements based on observational epidemiologic studies have not been supported by randomized controlled trials, in which either harm, or at best no benefit, has resulted. The expected value of reducing hyperhomocysteinemia through folate or B vitamin supplements has similarly been found by randomized controlled trials to be unfulfilled. Inflammation and C-reactive protein are areas of prominent research interest, with findings related to mechanisms of atherogenesis. Recent clinical trial evidence suggests that statins reduce already-low LDL-cholesterol and elevated C-reactive protein with marked benefits in reducing rates of vascular outcomes in healthy men and women. These findings raise new issues regarding the potentially widespread use of pharmacotherapy in primary prevention. Finally, “novel” markers continue to emerge from new and ongoing studies, and an approach to systematic evaluation of their contribution to cardiovascular risk prediction has now been proposed. This approach may serve to elevate priority for investigation of certain suggested markers and to put others in perspective as informative but less clinically applicable measures.
46894_CH15_431_502.qxd
2/26/10
12:31 PM
Page 433
1. ALCOHOL CONSUMPTION 433
1. ALCOHOL CONSUMPTION Overview Alcohol consumption has, in addition to its many known adverse health effects, properties that tend to reduce the risk of coronary heart disease. In some respects, alcohol-containing beverages simply constitute another one of many dietary constituents. Ascertainment of alcohol consumption entails similar procedures and encounters the same difficulties as with other aspects of dietary data collection. But special social and cultural considerations surrounding the use of alcohol compound these difficulties, add to problems in interpreting reported intakes, and complicate establishment and implementation of policies concerning its use. Alcohol consumption therefore warrants discussion separately from dietary imbalance, although its relation with other aspects of diet is also important.
semiquantitative units of intake (usually categorized as light, moderate, or heavy), and categories of user versus abuser or alcoholic. Time patterns of consumption may be distinguished as number of drinks per occasion of drinking. “Binge drinking” is concentration of consumption within a single occasion, not revealed by the average weekly intake. Finally, blood alcohol level (BAL) may be referred to, in units of parts alcohol/10,000 parts blood, with a ratio of 5/10,000 (BAL 0.05 ml ethanol/100 ml blood) associated with sensory and psychological symptoms and 10/10,000 (BAL 0.10) often legally defined as alcohol intoxication.1 General Reliability In a report on the relation of alcohol consumption and mortality among British physicians, Doll and others summarized succinctly the problem of assessing this characteristic:3, pp 911–912 Reliable quantitative evidence is, however, difficult to obtain. Information about drinking habits has to be obtained not from direct measurement but from answers provided by individual people about themselves or their close relatives and friends. Unless the amount usually drunk is close to zero it is intrinsically difficult to describe, and the description is peculiarly liable to bias. For many people, the consumption of alcohol has emotional and moral overtones, and respondents may underestimate the amount drunk from feelings of guilt or, perhaps less often, exaggerate it out of bravado. Moreover, the amount that a person normally drinks may vary substantially from one period to another, affecting the relevance of answers at one time to subsequent mortality.
Definitions and Measurement Qualitative Categories Alcohol consumption refers to individual practices regarding alcoholic beverages of all types, including beer, wine, fortified wine, and distilled spirits. Lifelong practice of abstinence from alcohol use characterizes persons or groups categorized as abstainers or teetotalers (from “T-total” to emphasize total abstinence). Use of alcohol may be classified as past, for persons reporting previous but not current use, or current, for persons who are then further classified as to beverage types, time pattern, and quantity of use. Quantities of alcohol are variously described in epidemiologic reports. Each type of beverage has its own unit of measure and conversion factor to express its ethanol content, as indicated in Table 15-1.1,2 Quantitative Units Other dimensions of alcohol intake include numbers of drinks per unit of time (usually daily or weekly),
The susceptibility to error in such self-reports is reminiscent of that confronting dietary assessment generally. The advantage of study of persons of middle or older age, whose habits may be rather stable, was noted in support of the use of such data in studies among adult physicians. It may be difficult to
Table 15-1
Quantities of Alcohol (Ethanol) Consumption Ethanol Type of Beverage Content (%) Unit of Measure Beer (US) 3.5 12-oz bottle, 355 ml Wine 12.1 3.5-oz glass, 104 ml Distilled spirits, 80 proof 40.0 1-oz shot, 30 ml a.
Ethanol Amount in oz (ml) 0.42 (12.43) 0.42 (12.58) 0.40 (12.00)
Conversion from oz of Beverage to oz of Ethanola 0.045 0.129 0.411b
Ethanol in grams 23 (ethanol in oz). Ethanol in spirits 0.411 (oz of spirits), not (volume of mixed drink).
b.
Source: Data from Committee on Diet and Health, National Academy of Sciences, Washington DC, 1989; Kuller LH, Alcohol and Cardiovascular Disease, In: Pearson TA, Criqui MH, Luepker RV, Oberman A, Winston M, eds, Primer in Preventive Cardiology, American Heart Association, Dallas, 1994.
46894_CH15_431_502.qxd
434
2/26/10
12:31 PM
Page 434
CHAPTER 15 OTHER PERSONAL FACTORS
judge the applicability of current information to past habits, however. In one investigation, consistency of reported alcohol consumption on two occasions 10 years apart was tested, and those whose current drinking on the second occasion was in the upper extreme recalled less drinking 10 years earlier than they had reported at that time.4 The British Regional Heart Study evaluated reported alcohol intake at baseline by the original reports and when participants were questioned about baseline intake at successive follow-up for up to 20 years.5 “Heavy” intake at baseline became rare, “moderate” decreased in frequency, “light” increased, and “none” became more frequent, with successively later recall. This suggested that alcohol intake might be systematically biased toward underreporting, with the degree of error increasing with time. The Augsburg, Germany, cohort of the WHO MONICA Project was interviewed on two occasions 3 years apart regarding alcohol intake.6 It found that, as with other measures with an important degree of intraindividual variability, combining data from two points in time improved the prediction of outcomes— here, fatal and nonfatal coronary heart disease and allcause mortality. Light to moderate intake appeared to be associated with the lowest risk of both outcomes. Cahalan reviewed concerns about survey methods regarding alcohol consumption.7 The Special Problem of Exdrinkers People classified as exdrinkers pose special problems of interpretation. Exdrinkers exhibit greater risk of coronary heart disease than those reported as consuming relatively low, or “moderate,” quantities of alcohol. It would appear that moderate intake is protective. An alternative interpretation of the J-shaped curve of cardiovascular disease risk with increasing alcohol intake could be that some proportion of selfreported nondrinkers stopped for health reasons and were at increased risk. A review by Criqui refuted this interpretation.8 First, any substantial degree of “migration” from high to low or no intake would have to occur, for example, before baseline classification in cohort studies, because later migration would weaken the association of higher intake with cardiovascular diseases. Second, the postulated shift would have to be specific to persons with cardiovascular diseases and not others, because the excess risk at the lowest levels of alcohol intake is mainly related to these diseases; however, this seems implausible. Third, some studies excluded persons from analysis who had known cardiovascular diseases at baseline, and the J-shape of the risk curve remained. Fourth, when those who report never hav-
ing consumed alcohol are compared with current drinkers, their risk is higher, and this cannot be explained by change in status among past drinkers. Notwithstanding these arguments, the issue continues to be debated. Determinants and Mechanisms Diet and Health reviewed both population-wide or cultural determinants and individual or personal determinants of alcohol consumption. Use of alcoholic beverages has a long history in many societies, with no record of alcohol intake as a social problem until the advent of distillation, reportedly near AD 1100.1 Some cultural factors promote alcohol use, but in Western societies, particular mores and values often determine how effects of alcohol are viewed. Alcoholism is distinct from alcohol consumption in general; genetic contributions to familial recurrence of alcoholism have been found, such that it is four times as common among sons of alcoholic fathers than those of nonalcoholic fathers, even if rearing is not by the biologic parents. More typical patterns of alcohol intake at the individual level are presumed to reflect beliefs about the benefits or risks involved, peer influences, and the marketing of alcoholic beverages. There are several mechanisms of action by which alcohol intake could affect risks of atherosclerotic and hypertensive diseases.1 Alcohol intake in France has been shown to be associated with different patterns of macronutrient intake and food consumption, according to both beverage type (beer, wine, or mixed consumption) and level of ethanol intake.9 This recalls to mind the role of alcohol as part of a dietary pattern in which, for example, high alcohol intake may displace other energy sources such as fruits, vegetables, and cereals. A short list of mechanisms purported to underlie a cardioprotective effect of alcohol, shown in Table 15-2, includes effects on HDL-cholesterol, platelets, insulin sensitivity, inflammation, LDL-cholesterol oxidation, fibrinogen, and psychological stress.10 Some of these mechanisms will be discussed briefly. High-Density Lipoprotein (HDL) Cholesterol The most frequently reported apparently favorable effect of alcohol consumption is raising HDL-cholesterol. This mechanism has been observed in several prospective studies and in small clinical experiments. For example, in the Honolulu Heart Program, which studied more than 8000 Japanese American men, mean values of HDL-cholesterol increased from 42.2 mg/dl in nondrinkers to 56.7 mg/dl in those reporting drink-
46894_CH15_431_502.qxd
2/26/10
12:31 PM
Page 435
1. ALCOHOL CONSUMPTION 435
Table 15-2
Mechanism Purported to Underlie the Cardioprotective Effects of Moderate Consumption of Alcoholic Beverages Mechanism Biologic Response Strength of Evidence (Class) HDL raising Decreased atherosclerosis I Antiplatelet effects Decreased thrombosis IIb Increased insulin sensitivity38–40 Decreased type II DM IIa Anti-inflammatory effects Decreased CRP59 IIb Antioxidant properties of other Decreased LDL oxidation32 other effects? III components of alcoholic beverages Decreased fibrinogen Decreased thrombosis IIa Decreased psychological stress III CRP, C-reactive protein; DM, diabetes mellitus; LDL, low-density lipoprotein. Source: Reprinted with permission from The American Journal of the Medical Sciences, Vol 329, JA Hill, p 130.
ing 20 or more oz of alcohol weekly.11 These and similar findings in other US centers participating in the Collaborative Lipoprotein Phenotyping Study are shown in Figure 15-1.12 In analyses designed to test the hypothesis that the role of alcohol in modifying coronary risk was mediated through its effect on HDL-cholesterol, Criqui observed that approximately one-half the influence of alcohol was through this pathway, on the basis of reduction of the regression coefficient for alcohol when HDL-cholesterol concentration was included in the statistical model.8 In the Honolulu Heart Program, low-density lipoprotein (LDL) cholesterol concentration decreased from 147.0 to 97.7 mg/dl over the same range of alcohol intake as described for HDL-cholesterol.11 In addition, however, other studies have shown that triglyceride concentration increases with alcohol intake, as does the degree of increase in blood fats after fatty meals.1 Blood Pressure Also on the adverse side, blood pressure appears to be increased among drinkers of alcohol, perhaps to a greater extent during the withdrawal phase after intake than during or immediately after drinking. This could explain the reported pattern among heavy drinkers of exhibiting high blood pressure after hospital admission, with gradual return to lower values during hospital confinement.13 More generally, both blood pressure distributions and the prevalence of high blood pressure are associated with reported alcohol intake in numerous epidemiologic studies, usually with a dose-response gradient. For example, male workers in the Chicago Western Electric Company showed differences in mean values of systolic and diastolic blood pressure from 132.9 to 146.5 mm Hg and from 85.8 to 94.3 mm Hg, respectively, from the categories of occasional drinks or none to six-
or-more-beer equivalents per day, without adjustment for other factors.14 The proportion of men with high blood pressure also increased over this range of alcohol intake, from 20.0 to 47.4%. Similar results were reported for men in the Honolulu Heart Program10 and, in the Kaiser-Permanente experience, for both Black and White men and women.15 It appears that not all effects of alcohol on cardiovascular disease are beneficial. Genetic Influences Even the beneficial effect of alcohol may be restricted to a subset of the population with a specific genotype, according to a recent case-comparison study of myocardial infarction in the setting of the World Health Organization MONICA Project centers in Ireland and France (Étude Cas-Témoin de l’Infarctus du Myocarde, or ECTIM).16 Details of this study were discussed in Chapter 7, “Genes and Environment.” In brief, a gene that controls activity of the cholesteryl ester transfer protein (CETP) gene, prominently involved in HDL-cholesterol metabolism, has two alleles, B1 and B2. The B2 allele was found to interact with alcohol consumption in determining blood concentrations of both CETP and HDL-cholesterol. However, unusually high alcohol consumption (75 g/ day, corresponding to more than 3 oz ethanol or nearly eight drinks per day) was necessary for this mechanism to increase HDL-cholesterol. Hemostasis It has been suggested that alcohol consumption relates positively to hemostatic function.17 In the Physicians’ Health Study, alcohol intake was directly associated with concentration of tissue-type plasminogen activator (t-PA), an enzyme involved in reversal of blood clot formation (fibrinolysis). Among persons classified as drinking alcohol with decreasing frequency (daily,
46894_CH15_431_502.qxd
436
2/26/10
12:31 PM
Page 436
CHAPTER 15 OTHER PERSONAL FACTORS
Figure 15-1 Mean High Density Lipoprotein (HDL) Cholesterol Levels, According to Reported Habitual Alcohol Intake in the Cooperative Lipoprotein Phenotyping Study. Source: Reprinted with permission from SB Hulley and S Gordon, Circulation, Vol 64 (Suppl III), p III-59, © 1981.
weekly, monthly, and rarely or never), t-PA concentrations decreased from 10.9 to 9.7, 9.1, and 8.1 nanograms/ml, respectively (p for trend 0.0002). This effect was independent of HDL-cholesterol concentration and offers a distinct mechanism by which alcohol intake could reduce the risk of coronary heart disease. The quantity of intake was not determined except that those classified as daily drinkers included persons reporting consumption of up to two or more drinks daily. Also, studies in US middle-aged and young adults, among others, have shown fibrinogen concentrations to be inversely related to alcohol intake.18,19 In the Atherosclerosis Risk in Communities Study, a 100 ml/week increment of ethanol intake was equivalent to a statistically significant decrement of fibrinogen of 2.2 mg/dl; but adjustment for other factors (lipids, insulin, and leukocyte count) reduced
the change in fibrinogen to 0.9 mg/dl, no longer statistically significant. Attenuation of the Alcohol Association Analysis of experience in the Nurses Health Study and Health Professionals Follow-Up Study identified associations of alcohol intake with HDL-cholesterol, hemoglobinA1C (a measure of glucose intolerance), and fibrinogen levels.20 Lowest risks of myocardial infarction occurred among men and women who reported drinking 3 to 7 days per week, in analyses adjusted for major risk factors. Further adjustment for these three additional factors greatly attenuated the association among women, reducing the alcohol effect by 75%, and eliminated it altogether in men. Publication of these results stimulated an accompanying editorial noting the evident confounding of the
46894_CH15_431_502.qxd
2/26/10
12:31 PM
Page 437
1. ALCOHOL CONSUMPTION 437
alcohol–coronary association by these three characteristics and calling for the necessary next scientific step to clarify the true role of alcohol, a randomized controlled trial.21 The additional finding that educational status and income influenced the association between beverage type and all-cause mortality led investigators in Copenhagen to suggest that socioeconomic factors should be taken more fully into account in future studies.22 Distribution US Adults The National Health Interview Survey provides data on alcohol intake of adults aged 18 years and older, categorized as “current drinker,” “former drinker,” or “lifetime abstainer” and as “heavier drinker” (men reporting more than 14 drinks and women reporting more than 7 drinks per week), “five or more drinks in a day on at least 1 day in the past year,” and “five or more drinks in a day on at least 12 days in the past year” (Table 15-3).23 Data shown are for both sexes combined, by age, race/ethnicity, and percent of poverty level. Survey years 1997, 2000, 2005, and 2006 indicate recent trends. In 2006, 60.8% of adults were self-reported as drinkers, 14.1% as former drinkers, and 25.0% as lifetime abstainers. The oldest age groups, especially at age 65 years and older, were less frequent current drinkers and more often lifetime abstainers than younger adults. Whites were more often current drinkers and Asians were more often lifetime abstainers than other groups. Current drinking was more frequent above than below the 200% of poverty line, and lifetime abstention was less so. Males were current drinkers more frequently than females (67.4 versus 54.9%). Heavier drinkers were 5.0% of adults, whereas 20.2% were drinkers of five or more drinks in a day at least 1 day, and 9.2% at least 12 days, in the past year. Heavier drinking was similar by sex, but males reported five or more drinks in a day substantially more often than females. Trends were generally in the direction of less frequent current drinking, and males were less often heavier drinkers in 2006 than in 1997, whereas for females, each category of heavy drinking increased in frequency. Diet and Health indicated that in the United States, 18 million persons aged 18 years or older were estimated to have alcohol-related problems as of the mid-1980s.1 Of these, 41% were classified as alcohol abusers and 59% as alcoholics, with the distinction being made on the basis of attributed social or health problems, including job loss, arrest, illness, loss of
behavioral control, symptoms of alcohol withdrawal, and others. The high prevalence of drinking especially among some groups of Native American adults was noted. This was not apparent in current drinking status but was suggested by the heavier drinking categories, in Table 15-3. US Youth The National Survey on Drug Use and Health provides data on youth as well as on adults, beginning with age 12 years and age-specific by two-year strata to age 18 years, then 18–25, 26–34, and 35 years and older, similarly by sex and by race/ethnicity (Table 15-4).23 Alcohol use was reported by 3.9% of 12–13 year olds, reached 29.7% by age 17–18 years, and doubled for the 18–25 age group. “Binge alcohol use” in this survey refers to consumption of five or more drinks on one occasion in the past 30 days, and “heavy drinking” here is similar intake on 5 or more days in the past 30 days. At ages 16–17 years, 20.0% were binge drinkers and 5.6% were heavy drinkers. Alcohol use was more frequent, binge drinking was twice as frequent, and heavy alcohol use was three times as frequent among males as among females. Current drinking was most frequent among Whites, and binge and heavy drinking were most frequent among American Indian or Alaska Natives. Only slight changes were observed in these patterns over the period from 2002 to 2006; no comparable data were available for earlier years. Study of the alcohol content of beer, wine, and spirits sold in the United States permits estimation of per capita alcohol consumption from each source.24 Traced from 1950–2002, the trends show an irregular pattern with a general increase from 0.2 to 0.3 gallons of alcohol from wine; an increase in alcohol intake from spirits from about 0.7 gallons to a peak of 1.1 gallons around 1975 and decrease to 0.6 gallons per capita; and an increase in beer alcohol from 1.15 gallons to a peak of 1.5 gallons in 1982 and slight decline to 1.3 gallons. As of 2002, beer accounted for more than twice as much alcohol consumed as spirits and more than four times as much as wine. Variation in Prevalence Among Populations The INTERSALT Study included 52 cross-sectional sample surveys from countries around the world in men and women who were classified, as part of their nutritional assessment, according to both weekly alcohol intake (in ml) and categories of heavy versus lesser degrees of drinking.25 The proportion of persons who were heavy drinkers was very strongly correlated (r 0.97) with population mean levels of alcohol
Alcohol Consumption by Adults 18 Years of Age and Over, by Selected Characteristics: United States, Selected Years 1997–2006 [Data are based on household interviews of a sample of the civilian noninstitutionalized population] Lifetime Alcohol Drinking Status1 Current Drinker Former Drinker Lifetime Abstainer Characteristic 1997 2000 2005 2006 1997 2000 2005 2006 1997 2000 2005 Percent of Adults 18 years and over, 63.1 61.4 61.4 60.8 15.7 14.4 14.3 14.1 21.2 24.2 24.4 age-adjusted2 18 years and over, 63.4 61.6 61.4 60.8 15.5 14.3 14.3 14.3 21.1 24.1 24.3 crude Both sexes Age All persons: 18–44 years 69.4 67.3 66.4 65.8 10.6 9.7 8.8 9.3 19.9 23.1 24.8 18–24 years 62.2 59.1 58.1 59.3 5.9 5.2 4.4 4.6 31.8 35.7 37.5 25–44 years 71.6 69.9 69.2 68.1 12.0 11.1 10.3 10.9 16.4 19.1 20.5 45–64 years 63.3 62.0 62.6 61.5 18.5 16.8 17.3 17.2 18.3 21.1 20.1 45–54 years 67.1 65.1 66.3 64.9 16.8 15.0 15.1 14.6 16.1 20.0 18.6 55–64 years 57.3 57.3 57.3 56.8 21.1 19.7 20.4 20.8 21.6 22.9 22.2 65 years and over 43.4 42.1 43.1 43.7 26.7 25.0 25.8 23.6 29.9 33.0 31.1 65–74 years 48.6 47.0 47.7 48.2 24.8 23.8 25.1 22.2 26.6 29.3 27.3 75 years and over 36.6 36.2 38.0 38.5 29.1 26.4 26.6 25.1 34.3 37.4 35.4 Race2,3 White only 66.0 64.5 64.4 63.8 15.2 14.2 14.1 14.0 18.7 21.3 21.5 Black or African 47.8 46.7 46.4 48.5 21.0 17.1 16.4 16.0 31.1 36.1 37.2 American only American Indian or 53.9 54.2 50.0 52.8 22.9 21.7 18.9 19.9 23.2 *24.1 31.1 Alaska Native only Asian only 45.8 43.0 42.9 43.0 8.8 9.2 9.5 9.3 45.3 47.8 47.6 Native Hawaiian or Other Pacific --* * * --* * * --* * Islander only 2 or more races --61.4 51.6 55.0 --19.5 22.6 26.6 --19.1 25.8 Hispanic origin and race2,3 Hispanic or Latino 53.4 52.4 50.8 50.5 14.7 12.4 13.8 13.6 32.0 35.2 35.4 Mexican 53.0 51.0 48.1 49.0 14.4 13.4 14.8 14.5 32.6 35.6 37.0
Table 15-3
35.9 36.4
18.4
*
47.7
27.3
22.2 35.5
12:31 PM
24.9 36.1 21.0 21.3 20.4 22.4 32.7 29.5 36.4
2/26/10
24.9
438
25.0
2006
46894_CH15_431_502.qxd Page 438
CHAPTER 15 OTHER PERSONAL FACTORS
65.9 46.7
45.3 50.6 66.3
67.5 47.8
46.1 52.8 68.7
67.1
43.3 49.2
67.0 46.0
62.9
65.9
46.0 51.2
66.4 48.4
62.5
13.9
20.2 20.1
15.4 21.0
15.8
12.9
18.8 17.9
14.4 17.1
14.6
12.6
19.2 18.2
14.1 16.5
14.3
12.4
19.1 18.0
14.0 16.1
14.2
17.4
33.6 27.1
17.1 31.2
20.1
20.8
35.9 31.5
19.7 36.2
22.8
20.3
37.5 32.6
18.9 37.5
22.7
21.7
34.9 30.8
19.6 35.4
23.3
Data from Health, United States, 2008, pp 305–310, National Center for Health Statistics, Centers for Disease Control and Prevention, Department of Health and Human Services.
Source: CDC/NCHS, National Health Interview Survey, family core and sample adult questionnaires.
12:31 PM
*Estimates are considered unreliable. Data preceded by an asterisk have a relative standard error (RSE) of 20%–30%. Data not shown have an RSE of greater than 30%. - - - Data not available. 1 Lifetime alcohol drinking status categories are based on self-reported responses to questions about alcohol consumption. Current drinkers had at least 12 drinks in their lifetime and at least one drink in the past year. Former drinkers had at least 12 drinks in their lifetime and none in the past year. Lifetime abstainers had fewer than 12 drinks in their lifetime. See Appendix II, Alcohol consumption. 2 Estimates are age-adjusted to the year 2000 standard population using four age groups: 18–24 years, 25–44 years, 45–64 years, and 65 years and over Age-adjusted estimates in this table may differ from other age-adjusted estimates based on the same data and presented elsewhere if different age groups are used in the adjustment procedure. See Appendix II, Age adjustment. 3 The race groups, White, Black, American Indian or Alaska Native, Asian, Native Hawaiian or Other Pacific Islander, and 2 or more races, include persons of Hispanic and non-Hispanic origin. Persons of Hispanic origin may be of any race. Starting with 1999 data, race-specific estimates are tabulated according to the 1997 Revisions to the Standards for the Classification of Federal Data on Race and Ethnicity and are not strictly comparable with estimates for earlier years. The five single-race categories plus multiple-race categories shown in the table conform to the 1997 Standards. Starting with 1999 data, race-specific estimates are for persons who reported only one racial group; the category 2 or more races includes persons who reported more than one racial group. Prior to 1999, data were tabulated according to the 1977 Standards with four racial groups and the Asian only category included Native Hawaiian or Other Pacific Islander. Estimates for single-race categories prior to 1999 included persons who reported one race or, if they reported more than one race, identified one race as best representing their race. Starting with 2003 data, race responses of other race and unspecified multiple race were treated as missing, and then race was imputed if these were the only race responses. Almost all persons with a race response of other race were of Hispanic origin. See Appendix II, Hispanic origin; Race. 4 Percent of poverty level is based on family income and family size and composition using US Census Bureau poverty thresholds. Missing family income data were imputed for 26%–30% of adults 18 years of age and over in 1997–1998 and 32%–35% in 1999–2006. See Appendix II, Family income; Poverty. Notes: Standard errors are available in the spreadsheet version of this table. Available from: http//www. cdc gov/nchs/hus htm. Data for additional years are available. See Appendix III.
62.6
64.1
2/26/10
Not Hispanic or Latino White only Black or African American only Percent of poverty level2,4 Below 100% 100%–less than 200% 200% or more
46894_CH15_431_502.qxd Page 439
1. ALCOHOL CONSUMPTION 439
46894_CH15_431_502.qxd
440
2/26/10
12:31 PM
Page 440
CHAPTER 15 OTHER PERSONAL FACTORS
Table 15-4
Use of Alcohol in the Past Month Among Persons 12 Years of Age and Over, by Age, Sex, Race, and Hispanic Origin: United States, 2002–2006 [Data are based on household interviews of a sample of the civilian noninstitutionalized population 12 years of age and over] Age, Sex, Race, and Alcohol Use Binge Alcohol Use4 Heavy Alcohol Use5 Hispanic Origin 2002 2005 2006 2002 2005 2006 2002 2005 2006 Percent of Population 12 years and over 51.0 51.8 50.9 22.9 22.7 23.0 6.7 6.6 6.9 Age 12–13 years 4.3 4.2 3.9 1.8 2.0 1.5 0.3 0.2 0.2 14–15 years 16.6 15.1 15.6 9.2 8.0 8.9 1.9 1.7 1.2 16–17 years 32.6 30.1 29.7 21.4 19.7 20.0 5.6 5.3 5.6 18–25 years 60.5 60.9 61.9 40.9 41.9 42.2 14.9 15.3 15.6 26–34 years 61.4 62.5 61.8 33.1 32.9 34.2 9.0 9.6 10.0 35 years and over 52.1 53.3 51.8 18.6 18.3 18.4 5.2 4.7 5.1 Sex Male 57.4 58.1 57.0 31.2 30.5 31.2 10.8 10.3 10.7 Female 44.9 45.9 45.2 15.1 15.2 15.2 3.0 3.1 3.3 Age and sex 12–17 years 17.6 16.5 16.6 10.7 9.9 10.3 2.5 2.4 2.4 Male 17.4 15.9 16.3 11.4 10.4 10.7 3.1 3.0 2.8 Female 17.9 17.2 17.0 9.9 9.4 9.9 1.9 1.8 1.9 Hispanic origin and race3 Not Hispanic or Latino: White only 55.0 56.5 55.8 23.4 23.4 24.1 7.5 7.4 7.8 Black or African 39.9 40.8 40.0 21.0 20.3 19.1 4.4 4.2 4.6 American only American Indian or 44.7 42.4 37.2 27.9 32.8 31.0 8.7 11.5 9.0 Alaska Native only Native Hawaiian or Other * 37.3 36.7 25.2 25.7 24.1 8.3 5.3 11.0 Pacific Islander only Asian only 37.1 38.1 35.4 12.4 12.7 11.8 2.6 2.0 2.4 2 or more races 49.9 47.3 47.1 19.8 20.8 22.8 7.5 5.6 6.3 Hispanic or Latino 42.8 42.6 41.8 24.8 23.7 23.9 5.9 5.6 5.7 See footnotes at end of table. 3 Persons of Hispanic origin may be of any race. Race and Hispanic origin were collected using the 1997 Revisions to the Standards for the Classification of Federal Data on Race and Ethnicity. Single-race categories shown include persons who reported only one racial group. The category 2 or more races includes persons who reported more than one racial group. See Appendix II, Hispanic origin; Race. 4 Binge alcohol use is defined as drinking five or more drinks on the same occasion on at least one day in the past 30 days. Occasion is defined as at the same time or within a couple of hours of each other. See Appendix II, Binge drinking. 5 Heavy alcohol use is defined as drinking five or more drinks on the same occasion on each of five or more days in the past 30 days. By definition, all heavy alcohol users are also binge alcohol users. Notes: The National Survey on Drug Use & Health (NSDUH), formerly called the National Household Survey on Drug Abuse (NHSDA), began a new baseline in 2002 and cannot be compared with previous years. Because of methodological differences among the National Survey on Drug Use & Health, the Monitoring the Future Study (MTF), and the Youth Risk Behavior Survey (YRBS), rates of substance use measured by these surveys are not directly comparable. See Appendix I, MTF, NSDUH, and YRBS. Data for additional years are available. See Appendix III. Source: Substance Abuse and Mental Health Services Administration, Office of Applied Studies, National Survey on Drug Use & Health. Available from http://www.oas.samhsa.gov/nsduh.htm. Data from Health, United States, 2008, pp 300–304, National Center for Health Statistics, Centers for Disease Control and Prevention, Department of Health and Human Services.
intake. This finding was interpreted by Rose as evidence that the greater the average alcohol intake in a population, the more frequent would be behavior at the upper extreme. (The median value rather than the mean would be clearer evidence for this point, which enters into discussions of alcohol policies.) The European Prospective Investigation into Cancer and Nutrition (EPIC) demonstrated wide
variation among study populations throughout Europe in levels and types of alcoholic beverages consumed.26 A lifestyle questionnaire was for assessment of alcohol intake retrospectively, when each participant was age 20, 30, 40, or (in two of 21 centers) 50 years. Evidence of changes in quantity and type of alcohol consumed led the investigators to conclude that lifetime history of alcohol intake should be incorporated in epidemiologic studies, especially
46894_CH15_431_502.qxd
2/26/10
12:31 PM
Page 441
1. ALCOHOL CONSUMPTION 441
where variation over time could influence comparisons within a multicenter study. Criqui and Ringel analyzed published statistics data on alcohol use by type of beverage among 21 economically developed countries in Europe, North America, Australia, and Japan; food disappearance data centering on the same years; and data on mortality from all causes and from coronary heart disease at ages 35–74 years for 1965, 1970, 1980, and 1988.27 Their purpose was to investigate the “French paradox,” the observation that France’s presumed adverse dietary characteristics were accompanied by low coronary mortality, possibly explained by its high rank among countries in wine intake. Gross mea-sures of population use of ethanol in wine, beer, and spirits and of animal fat, vegetables, and fruits were presented for each country. Regarding international variation in alcohol intake, national per capita ethanol consumption in liters per capita from wine, beer, and spirits ranged from 0.6 (Israel) to 9.1 (France), 0.3 (Iceland) to 5.9 (Austria), and 1.0 (Italy) to 3.2 (Finland), respectively. Total ethanol intake for France was 13.1 liters per capita per year, rivaled only by Spain with 12.2 liters; the total for the United States was 7.2 liters. Rates and Risks Between-Population Differences Whether differences in alcohol intake between countries are related to coronary mortality was studied most thoroughly in this investigation of the French paradox. Marked between-population differences in alcohol consumption were apparent, by both amounts and sources. As of 1988, the last year of the analysis by Criqui and Ringel, France exceeded all other countries in ethanol consumption from wine and ranked above the median for ethanol from spirits but below for beer. Whether high alcohol consumption was the most likely explanation of the favorable coronary mortality remained to be addressed.27 Notably, other aspects of the French diet were favorable: Animal fat intake of 25.7% of calories ranked among the lowest among the 21 countries, especially in 1965 and 1970, whereas vegetable intake ranked among the highest, with only three countries having higher intakes in 1988. Figure 15-2 indicates the relation between coronary heart disease death rates and use of wine ethanol, total ethanol, and the three food components, each separately, for the 1988 data. A positive correlation with animal fat and negative correlations with vegetable and fruit consumption were accompanied by negative correlations with total and wine ethanol, the latter being the strongest
negative relation with a correlation coefficient, R, of –0.66. The volume of wine consumed was no more strongly correlated with coronary mortality than wine ethanol, so wine ingredients other than ethanol were judged not to be the basis for the correlation. The impression from this analysis that ethanol from wine is specifically related to reduced coronary mortality was interpreted as reflecting the dominance in the analysis of data for France, which had exceptionally high wine intake and low coronary mortality, whereas in most countries wine was not the main ethanol source. This probably accounts for the variation in reports from within-population studies regarding the ethanol source that relates to reduced coronary heart disease risk. Especially important was the further observation that ethanol from beer was significantly positively associated with all-cause mortality. There was no overall indication of lower total mortality in relation to ethanol. France, with the highest per capita ethanol intake among the 21 countries, did not rank lowest in total mortality, because of noncardiovascular causes associated with ethanol consumption that offset any national benefit from reduced coronary mortality. In a further study of alcohol intake and cardiovascular disease in France, a comparison with Ireland was undertaken in a five-year prospective study of middle-aged men free of coronary heart disease at entry.28 Relations of alcohol and coronary heart disease differed between the two cohorts, with greater reduction in relative risk at each level of drinking in France and no such trend in Ireland. Within-Population Comparisons Within-population analysis of alcohol consumption and coronary mortality has been reported in a number of cohort and case-comparison studies. The study of British doctors cited previously showed for ischemic heart disease (ICD 9 codes 410–414) a decrease in rates from 12.3 to 7.1/1000 per year across categories of alcohol intake from none to 15–21 units per week. Rates increased with further increments of alcohol intake to 9.2 and 8.9/1000 per year at 29–42 and 43 or more units per week, respectively.3 The data were evaluated first by categorical comparison of the combined groups from 1–14 units per week versus none and, second, by testing the significance of the trend over paired categories of drinkers with intakes of 1–14, 15–28, or 29 or more units per week. The minimum risk was at the level of 8–14 or 15–21 units of alcohol per week (one to three drinks per day, as defined in this study). For both ischemic heart disease and the category, other known causes of
46894_CH15_431_502.qxd
442
2/26/10
12:31 PM
Page 442
CHAPTER 15 OTHER PERSONAL FACTORS
Figure 15-2 Correlations of Age-Adjusted Coronary Heart Disease (CHD) Mortality Rates per 100,000 Population with Dietary Items in 1988. Source: Reprinted with permission from MH Criqui and BL Ringel, Lancet, Vol 344, p 1721, © 1994, The Lancet, Ltd.
death, rates were somewhat higher with lower intake; only for ischemic heart disease was this effect sufficient to make the linear trend nonsignificant. For “alcohol-augmented causes,” only the linear trend was significant, indicating a generally continuous increase in risk from lowest to highest categories of intake, without the J- or U-shaped relation. Also noteworthy was the significant trend of risk for cerebrovascular disease, attributable to the highest intake group, a finding reported elsewhere as well. As the authors concluded, ischemic heart disease mortality was reduced among alcohol users in contrast to nonusers, and they judged the effect to be “largely irrespective of amount.”3, p 911 Two studies, however, illustrate the observation that changes in intake may be associated with changes in risk. In one case, increased intake by 12.5 g/day during follow-up in the
course of longitudinal observation was associated with a significantly reduced relative risk of fatal or nonfatal coronary heart disease.29 In another, decreasing consumption was associated with increased risk and increasing consumption with reduced risk of all-cause mortality.30 A report from the Physicians Health Study indicated that sudden cardiac death shared the apparent benefit of risk reduction with as little as two to four or as much as five to six drinks per week.31 The Atherosclerosis Risk In Communities (ARIC) Study in the United States examined risk of incident coronary heart disease by sex and race in middle-aged adults followed for nearly 10 years.32 A positive association with alcohol intake was found for Black men in contrast to an inverse association for White men. The authors questioned whether the putative
46894_CH15_431_502.qxd
2/26/10
12:31 PM
Page 443
1. ALCOHOL CONSUMPTION 443
cardioprotective effect of alcohol is real or reflects lifestyle characteristics of drinkers. Other reports have found corresponding associations of alcohol consumption with reduced progression of coronary atherosclerosis, heart failure due to coronary artery disease, noncoronary heart failure in persons with diabetes mellitus, and dementia.33–35
Data from several studies were compiled by Rehm and others to investigate the relation between categories of alcohol intake and cardiovascular and other chronic conditions (Table 15-5).36 In contrast to abstainers, consumers of alcohol from level I to III (see legend for sex-specific definitions) experienced continuously increasing risk of malignancies, neuropsy-
Table 15-5
Relative Risk for Major Chronic Disease Categories, by Gender and Average Drinking Category Females Males ICD–9 ICD–10 Drinking Category* Disease Code Code I II III I II Malignant neoplasms 140–208 C00–C97 Mouth and oropharynx cancers 140–149 C00–C14 1.45 1.85 5.39 1.45 1.85 Esophagus cancer 150 C15 1.80 2.38 4.36 1.80 2.38 Liver cancer 155 C22 1.45 3.03 3.60 1.45 3.03 Breast cancer 174 C50 1.14 1.41 1.59 Under 45 years of age 1.15 1.41 1.46 45 years and over 1.14 1.38 1.62 Other neoplasms 210–239 D00–D48 1.10 1.30 1.70 1.10 1.30 Diabetes mellitus 250 E10–E14 0.92 0.87 1.13 1.00 0.57 Neuropsychiatric conditions 290–319, F01–F99, 324–359 G06–G98 Unipolar major depression 300.4 F32–F33 RR not available: AF could not be determined otherwise (Rehm et al., in press b) Epilepsy 345 G40-G41 1.34 7.22 7.52 1.23 7.52 Alcohol use disorders 291, 303, F10 AF** AF AF AF AF 305.0 100%† 100% 100% 100% 100% Cardiovascular diseases (CVD) 390–459 I00–I99 Hypertensive disease 401–405 I10–I13 1.40 2.00 2.00 1.40 2.00 Coronary heart disease 410–414 I20–I25 0.82 0.83 1.12 0.82 0.83 Cerebrovascular disease 430–438 I60–I69 Ischemic stroke 0.52 0.64 1.06 0.94 1.33 Hemorrhagic stroke 0.59 0.65 7.98 1.27 2.19 Other CVD causes 415–417, I00, I26–I28, 1.50 2.20 2.20 1.50 2.20 423–424, I34–I37, 426–429, I44–I51, 440–448, I70–I99 451–459 Digestive diseases 530–579 K20–K92 Cirrhosis of the liver 571 K70, K74 1.26 9.54† 9.54† 1.26 9.54†
III 5.39 4.36 3.60
1.70 0.73
6.83 AF 100% 4.10 1.00 1.65 2.38 2.20
9.54†
Note: Relative risk estimates are shown to quantify the effect size of the risk relationships. For example, females in drinking category I have a relative risk of 1.14, compared with female abstainers, of breast cancer. A relative risk of 1.14 corresponds to a 14-percent higher risk. For females in drinking category III, the relative risk is 1.59, or about one and one-half times as large as for female abstainers. The same relationship can also be expressed as a risk increase of 59 percent. Varying numbers of studies were used to report on the different diseases. Measurement problems for outcomes affected the reliability of the data for some endpoints, especially the different subtypes at strokes and the unspecified categories such as “other cardiovascular disease” or “other neoplasms.” The result for these categories should be regarded with caution. *Definition of drinking categories. Category I: for females, 0–19.99 g pure alcohol daily; for males, 0–39.99 g pure alcohol daily. Category II: for females, 20–39.99 g pure alcohol daily; for males, 40–59.99 g pure alcohol daily. Category III: for females, 40 g or more pure alcohol; for males, 60 g or more pure alcohol. **AF attributable fraction—that is the proportion of disease under consideration that is attributable to alcohol. † For liver cirrhosis, a combined estimate was derived for drinking categories II and III. Sources: Unless otherwise specified, Gutjahr et al., 2001; for breast cancer and stroke, Ridolfo and Stevenson 2001; for hypertension Corrao et al., 1999; for CHD, drinking category III, Corrao et al., 2000. Reprinted with permission from Alcohol Research and Health: The Journal of the National Institute on Alcohol Abuse and Alcoholism, Vol 27, J Rehm, G Gmel, CT Sempos, M Trevisan, p 41.
46894_CH15_431_502.qxd
444
2/26/10
12:31 PM
Page 444
CHAPTER 15 OTHER PERSONAL FACTORS
chiatric disorders, hypertensive disease, “other” cardiovascular conditions, and cirrhosis of the liver. Only for diabetes mellitus, coronary heart disease, and stroke were relative risks below 1.00 except at level II or above—and for men, all intake levels were associated with increased risk for hemorrhagic stroke. The authors noted other studies indicating that binge drinking, regardless whether average ethanol intake was light to moderate, was associated with increased risk of coronary events. Marmot and Brunner reviewed reports on moderate alcohol use (five or fewer units per day) compared with the experiences of nondrinkers.37 Results were mixed but were in several instances consistent with a protective effect of this level of alcohol use, with no study showing a significant excess risk at this level. Incident ischemic stroke was investigated in the Health Professionals Follow-up Study of US males.38 Over a period of 14 years, 412 cases were documented, with relative risks of 0.99 for light drinkers (0.1–9.9 g/day), 1.26 for moderate drinkers (10.0–29.9 g/day), and 1.42 for heavier drinkers ( 30.0 g/day). Confidence limits around these estimates were wide, but the trend of increasing risk was statistically sig-
Table 15-6 Characteristic
nificant. When subclassified by age and other characteristics (Table 15-6), point estimates of relative risks were below 1.00 for younger age, higher BMI, aspirin use, and higher folate intake, but few of the respective confidence limits excluded 1. Overall, increased stroke risk appeared at levels of intake of more than two drinks/day, but there was no clearly reduced risk at lower levels. Global Impact Given that the 1983 declaration by the World Health Assembly that alcohol-related problems are among the world’s major health concerns, alcohol consumption has escalated in developing countries.39 Awareness of the global impact of alcohol has increased during the current decade. This is evident in reviews of alcohol use in China, India, and Russia; in discussion regarding low- and middle-income countries in the Disease Control Priorities for Developing Countries Project (DCP2); and in a series of articles on this topic in The Lancet in mid-2009.40–44 Room and others attributed causation of more than 60 different medical conditions and 4% of the global bur-
Multivariate-Adjusted Relative Risk for Ischemic Stroke According to Updated Alcohol Consumption, Stratified by Selected Clinical Characteristics* Multivariate-Adjusted Relative Risk According to p Value Alcohol Consumption (95% Cl) for Trend 0 g/d 0.1–9.9 g/d 10.0–29.9 g/d 30.0 g/d
Age 40–59 y 60 y Body mass index Below median level Above median level Aspirin use Yes No Hypertension Yes No Folate intake Below median level Above median level Caffeine intake Below median level Above median level
1.00 1.00
0.84 (0.40–1.79) 1.03 (0.72–1.47)
1.15 (0.51–2.57) 1.31 (0.90–1.89)
2.05 (0.83–5.07) 1.44 (0.94–2.21)
0.03 0.02
1.00 1.00
1.11 (0.68–1.80) 0.89 (0.58–1.37)
1.44 (0.87–2.38) 1.13 (0.72–1.77)
1.82 (1.02–3.23) 1.15 (0.68–1.94)
0.01 0.2
1.00 1.00
0.80 (0.51–1.27) 1.19 (0.75–1.88)
1.16 (0.73–1.84) 1.33 (0.82–2.15)
0.93 (0.52–1.65) 2.00 (1.18–3.40)
0.2 0.006
1.00 1.00
0.95 (0.59–1.51) 0.92 (0.59–1.44)
1.22 (0.76–1.97) 1.08 (0.67–1.75)
1.09 (0.64–1.86) 1.49 (0.84–2.64)
0.2 0.07
1.00 1.00
1.32 (0.80–2.18) 0.77 (0.50–1.17)
1.78 (1.06–2.98) 0.91 (0.58–1.44)
1.79 (1.01–3.18) 1.16 (0.68–1.99)
0.03 0.18
1.00 1.00
1.06 (0.71–1.59) 1.02 (0.59–1.76)
1.24 (0.80–1.93) 1.45 (0.84–2.51)
1.60 (0.95–2.70) 1.53 (0.84–2.81)
0.04 0.04
*Multivariate-adjusted relative risks adjusted for age; smoking; body mass index; geographic region; parental history of myocardial infarction; physical activity; hypercholesterolemia; aspirin use; diabetes; and intake of vitamin E, folate, energy, saturated fat, trans fats, potassium, magnesium, ω-3 fatty acids, and dietary fiber, except for the stratifying characteristic. Source: Reprinted with permission from Annals of Internal Medicine, Vol 142, KJ Mukamal, A Ascheiro, MA Mittleman, et al., p 16, © 2005 American College of Physicians.
46894_CH15_431_502.qxd
2/26/10
12:31 PM
Page 445
1. ALCOHOL CONSUMPTION 445
den of disease to alcohol, nearly equivalent to the impacts of tobacco and of hypertension.45 Rehm and others as part of DCP2 addressed the global health impact of alcohol-related disease on both regional and world-wide levels.43 Alcohol-related vascular diseases are considered to contribute about one-tenth of all alcohol-attributable DALYs worldwide, after an offsetting deduction for reduced vascular risk in high-income countries, presumably due to the relation of alcohol intake to coronary heart disease risk. Prevention and Control Individual Measures In the context of cardiovascular disease prevention, moderation of intake is generally recommended for persons who consume alcohol; initiation for the purpose of risk reduction is not recommended. The US Preventive Services Task Force (USPSTF) recommends screening and behavioral counseling interventions to reduce alcohol misuse by adults in primary care settings.46 The USPSTF has found evidence insufficient to support this intervention for adolescents. Community- or Population-Wide Measures Approaches recommended by the Guide to Community Preventive Services include regulation of alcohol outlet density, maintaining limits on days of sale, increasing alcohol taxes, and enhanced enforcement of laws prohibiting sales to minors.47 Global Strategies Policy Development—1990s. An International Symposium on Moderate Drinking and Health met in 1993 under sponsorship of several Canadian organizations and identified both policy implications and “best advice” for individuals.48 Policy implications were that alcohol control policies should not be relaxed, that substitution of lower-risk drinking for higher-risk drinking should be encouraged, and that educational messages should not suggest adoption of regular drinking by those who currently drink only irregularly or not at all. At the individual level, 10 points of advice were suggested, including that two standard drinks per day, omitting 1 day per week, should be the maximum intake and should not be adopted for health reasons by those who currently drink less; that intoxication should be avoided; that special circumstances have separate requirements, such as abstention during pregnancy; and that those planning to increase alcohol intake for health reasons should first consult their physicians, who may
identify contraindications or alternative means of risk reduction. Long-standing policies of 18 US health organizations and agencies with respect to prevention of substance abuse in adolescence were compiled by the American Medical Association and published in 1994, but these do not address the question of the possible beneficial effect of alcohol on atherosclerosis or its risk factors.49 A report in the United Kingdom, Sensible Drinking, reviewed the basis for official alcohol policy.50 It traced the development of this policy from its inception in 1976 through the report date of 1995. Current policy was to recommend that drinking less than 21 units weekly by men and 14 units by women was unlikely to damage health (one unit in United Kingdom usage 8 g ethanol, equivalent to one-half pint of beer or lager, a small glass of wine, or a standard measure of spirits). The lower levels for women reflected the evidence that metabolism of alcohol and tissue compartments differ by sex so as to generate substantially greater blood levels of alcohol in women at the same level of intake as for men. The report included a detailed review of studies on beneficial and adverse health effects of alcohol, as well as its own recommendations. The new report changed the advice on “sensible drinking” for the United Kingdom to indicate that for men at all ages regular drinking of between three and four units (between 24 and 32 g ethanol) daily and for women drinking between two and three units (between 16 and 24 g ethanol) daily will not accrue significant health risk. It was indicated that the “maximum health advantage” would be attained for both men and women at a level between one and two units (between 8 and 16 g ethanol) daily. Current Policies in 30 Countries. Brand and others conducted a comparative analysis of alcohol control policies in 30 countries comprising the Organization for Economic Cooperation and Development (OECD), including countries in Europe, Asia, and North America as well as Australia.51 By use of an Alcohol Policy Index devised to rate national policies, the investigators considered domains of physical availability of alcohol, drinking context, alcohol prices, alcohol advertising, and motor vehicle operator limits or testing for alcohol consumption. A wide range of scores indicated variable strength of alcohol control policies across these countries. Weaker policies were strongly associated with greater annual per capita ethanol consumption. World Health Organization. Following earlier reports on the alcohol problem, activity across the WHO Regions, and initiatives developed to intervene
46894_CH15_431_502.qxd
446
2/26/10
12:31 PM
Page 446
CHAPTER 15 OTHER PERSONAL FACTORS
in primary care settings, a new report from the WHO Expert Committee on Problems Related to Alcohol Consumption, published in 2007, recognized the alcohol-attributable burden as “becoming a priority area for international public health” and proposed that the burden could be reduced by proven strategies.52–54 Mechanisms of harm were reviewed as well as availability and consumption of alcohol throughout the world and the harm attributable to it in terms of both health and social effects. The following effective strategies and interventions were identified: measures to reduce availability of alcohol, alcohol price and taxes, restricting sale of alcohol, regulation of the drinking context, restrictions on alcohol marketing, drinking-driving countermeasures, education and persuasion, and early intervention and treatment services. Guidance in policy development across sectors and at different jurisdictional levels was included. The ten recommendations that conclude the report make no reference to alcohol use in relation to reduction of coronary heart disease risk. Disease Control Priorities in Developing Countries. Cost-effectiveness of potential interventions is discussed in detail by Rehm and others, cited previously.43 Their focus is prevention of high-risk drinking, defined as drinking 20 grams/day or more of ethanol on average for females and 40 grams/day for males, noting that a bottle of table wine contains about 70 g ethanol. Interventions evaluated for costeffectiveness included drunk-driving legislation and testing, taxation, reduced hours of sale, advertising bans, and brief clinical advice or counseling. It was considered that known interventions are cost-effective and could reduce the related burden by up to 25%. A high level of sensitivity and caution was advised in implementation of such policies because of local cultures, conditions, and interests. Current Issues Conflicting Attitudes and Judgments The dilemma addressed previously concerning policy and practice is the most immediate issue with respect to alcohol consumption. Strongly held opinions are discordant with respect to whether alcohol use should be positively recommended, whether potential risks outweigh potential benefits, and whether one form of alcoholic beverage may convey greater or lesser benefit than others, for example: The Group concludes . . . that light to moderate consumption of alcohol confers a protective effect against a number of serious diseases,
in particular CHD and ischemic stroke, and also against cholesterol gallstones.50, p 28 This report [Sensible Drinking, 1995] comes less than 6 months after the Royal Colleges of Physicians, Psychiatrists, and General Practitioners jointly concluded that low to moderate alcohol consumption protected against coronary heart disease but confirmed that the sensible limits of 21 and 14 units should continue because to increase them would do more harm than good.55, p 1643 The potential reduction in risk of CHD from drinking modest amounts of alcohol must be balanced against the fact that there is no level of drinking which is without risk of adverse consequences.48, p 13 [W]ell-intentioned information about moderate alcohol consumption and CHD could be used as an excuse to abuse alcohol. . . . [M]oderate alcohol consumption for the prevention of CHD as a public health policy would be irresponsible. . . . However, in selected patients at elevated CHD risk, responsible use of alcohol may offer some benefit. Advice beyond this limited arena could well do more harm than good, particularly given the evidence that average alcohol consumption in a given population faithfully predicts the extent of alcohol abuse, with a correlation coefficient of 0.97.8, p 138 [P]rudence suggests a lowering of or abstention from alcohol consumption.1, p 451 As pointed out by Kuller: It is important to note that no experimental clinical trials have shown that either increasing or decreasing alcohol consumption changes the risk of heart attack.2, p 230 It has been remarkable to participate in socialscientific gatherings of cardiovascular epidemiologists and other health professionals in recent years and to observe the now commonplace toast to HDLcholesterol raised by successive clusters of attendees (the present author, perhaps, among them). The contrast to objectivity, or skepticism, which greets virtually every other proposed intervention to reduce population rates and individual risks of cardiovascular diseases, requiring not one but multiple experimental tests of efficacy and safety, is extraordinary.
46894_CH15_431_502.qxd
2/26/10
12:31 PM
Page 447
2. ADVERSE PSYCHOSOCIAL FACTORS 447
The seeming obstacles to feasibility of such experiments are not unique to alcohol, but the dismissal of concern for the lack of such evidence is exceptional. In view of the apparently narrow “window of benefit”8 in relation to known adverse effects of exposure, the readiness in many quarters to embrace as public health policy the recommendation to take up or increase one’s alcohol use for prophylactic purposes will probably continue to be controversial. A Question of Substance In the further debate, it may be useful to ask, “Is atherosclerosis an alcohol-deficiency disease?” If the answer is no, then the admonition of Rose perhaps should be recalled: In mass prevention each individual has usually only a small expectation of benefit, and this benefit can easily be outweighed by a small risk. . . . This makes it important to distinguish two approaches. The first is the restoration of biological normality by the removal of an abnormal exposure. . . ; here there can be some presumption of safety. This is not true for the other kind of preventive approach, which leaves intact the underlying causes of incidence and seeks instead to interpose some new, supposedly protective intervention. . . . Here the onus is on the activists to produce adequate evidence of safety.56, p 38 Global Policy Based on a current global assessment of the burden of disease, injury, and economic cost attributable to alcohol use and alcohol-use disorders, Rehm and others have concluded that “alcohol consumption is one of the major avoidable risk factors, and action to reduce burden and costs associated with alcohol should be urgently increased.”44, p 2223 As a call to action, Casswell and Thamarangsi wrote:57, p 2247 Cost effective and affordable interventions to restrict harm exist, and are in urgent need of scaling up. Most countries do not have adequate policies in place. Factors impeding progress include a failure of political will, unhelpful participation of the alcohol industry in the policy process, and increasing difficulty in free-trade environments to respond adequately at a national level. An effective national and international response will need not only governments,
but also non-governmental organizations to support and hold government agencies to account. International health policy, in the form of a Framework Convention on Alcohol Control, is needed to counterbalance the global conditions promoting alcohol-related harm and encourage national action.
2. ADVERSE PSYCHOSOCIAL FACTORS Overview Psychosocial Factors Hemingway and Marmot provided a general definition of a “psychosocial factor” as “a measurement that potentially relates psychological phenomena to the social environment and to pathophysiological changes.”58 By its emphasis on measurement and on individual-level phenomena that link the social environment and pathophysiology, this definition avoids what its authors describe as “the unhelpful term of ‘stress.’” It implies objective assessment of psychological characteristics for which there is evidence, or a plausible hypothesis, of a demonstrable mediating role. Under this definition, the authors presented their systematic review of prospective cohort studies of psychosocial factors and coronary heart disease. The review identified studies that they categorized in four groups of psychosocial factors: “psychological traits (type A behaviour, hostility), psychological states (depression, anxiety), psychological interaction with the organisation of work (job control-demands-support), and social networks and social support.”58, p 1462 Qualifying reports in each area were distinguished as addressing either etiology of coronary heart disease in healthy populations or prognosis among persons with existing coronary heart disease. Table 15-7 summarizes their overall findings from the studies reviewed in each area. The inconsistency of findings in each area is noteworthy, as is the number of studies reporting moderate or stronger associations. These areas are the focus of the present review as well, on the basis of their relative prominence in this area of cardiovascular epidemiology and the diversity of psychological phenomena that they represent. These and related concepts of psychological factors associated with coronary heart disease were also reviewed by Theorell.59 He underscored the multidisciplinary nature of research in this area, involving epidemiology and social science, psychology and be-
46894_CH15_431_502.qxd
448
2/26/10
12:31 PM
Page 448
CHAPTER 15 OTHER PERSONAL FACTORS
Table 15-7
Summary of Prospective Studies Investigating Psychosocial Factors and CHD Number of Reports of Number of Reports of Etiological Studies (n 70) Prognostic Studies (n 92) 0 0 Type A behavior/hostility 1 11 5 1 3 10 1 1 Depression 0 7 6 9 0 16 7 11 Anxiety 0 4 1 3 1 9 4 4 Work characteristics 0 3 5 5 0 2 2 0 Social support 0 3 4 2 0 7 4 10 Summary of effect The extent to which the paper supports the hypothesis that adverse psychosocial characteristics increase risk of, or mortality from, CHD, is summarized in a single symbol (, 0, or ). The description of the summary symbols is as follows –Relative risk 0.75. “finding counter to hypothesis.” 0 Relative risk 0.75–1.50. ‘lack of clear association.’ Relative risk 1.50 and 2.00. ‘moderate association in line with hypothesis.’ Relative risk 2.00 ‘strong association in line with hypothesis’ Source: Reprinted with permission from Seminars in Vascular Medicine, Vol 2, No 3, H Kuper, M Marmot, H Hemingway, p 309, © 2002 by Thieme Medical Publishers, Inc.
havioral science, and medicine. Others have described as an “emerging field” the area of “behavioral cardiology, which is based on the understanding that psychosocial and behavioral risk factors for CAD [coronary artery disease] are not only highly interrelated, but also require a sophisticated health care delivery system to optimize their effectiveness.”60, p 637 Matthews, by contrast, focused specifically on what she termed “psychological science” and new opportunities for this discipline to contribute to further understanding of coronary heart disease and its development.61 Her attention was concentrated on opportunities for research, with no reference to its potential implications for clinical practice in psychology. The several cited sources provide detailed reviews, from varying perspectives, of the extensive research on psychological factors related to coronary heart disease. A brief discussion follows regarding the theoretical background of this work and some conceptual schemes to represent the implicated determinants and mechanisms. Theoretical Background The 1960s and 1970s was a period of especially active development in this area. Much of the literature from that period addressed difficulties in conceptualizing ideas, formulating hypotheses, and operationalizing research in this area. The necessity to communicate across several disciplines was noted as part of the difficulty. For example, in 1974 Cassel presented a theoretical formulation for research on
psychosocial processes and stress. He noted that epidemiologists and others commonly misinterpreted the medical concept of stress originally expressed by Selye and Wolff—in the original formulation, stress was a bodily state and not a component of the environment.62 From this point of view, misconception of stress by epidemiologists exclusively as an external agent of disease, disregarding the internal state of stress, led to inappropriate expectations––that is, causation of particular disorders rather than more general or diffuse health consequences. Consequently, this view led to the wrong questions and hypotheses regarding psychological factors and health. Cassel and others argued, to the contrary, that psychosocial processes should be expected to modify susceptibility to disease risks generally, through the well-known mechanism of their effects on neuroendocrine balance. In addition, they should be understood as culturally dependent. Therefore different social groups, as well as individuals, might differ in their characteristic psychological responses to particular external stimuli. (By implication, differences in findings, rather than consistency, might be expected across multiple population studies.) Finally, it was proposed that both beneficial and adverse effects of psychosocial processes should be recognized. Interventions to prevent disease might thus focus on strengthening the beneficial psychosocial influences, thereby reducing the influence of adverse factors. This view contributes to the idea that attention to patterns of psychosocial influences, rather than iso-
46894_CH15_431_502.qxd
2/26/10
12:31 PM
Page 449
2. ADVERSE PSYCHOSOCIAL FACTORS 449
lated components, may be most fruitful. Application of these concepts led one observer to consider those aspects of social stress that might contribute to the marked differences in coronary heart disease rates between the United States and Japan.63 The many stressful features of Japanese life appeared to be balanced by many institutionalized features for relief of stress. This suggested that overall stress might contribute less to coronary heart disease risks in Japan than in the United States, where stress-relieving features were thought to be lacking. This might offer a partial explanation of the striking population differences in coronary mortality between the two countries. Measurement Measurements specific to a particular factor are discussed as follows. Generally, the problem of measurement of psychosocial factors concerns the wide range of phenomena to be considered and the plethora of approaches used in addressing them. Defining the particular psychological characteristic to be studied and determining what questions will elicit relevant responses are challenging tasks. Cultural variation among individuals and across groups, noted previously, compounds the difficulty. Standardization is often lacking, although some questionnaire instruments have been widely used, such as the Minnesota Multiphasic Personality Inventory (MMPI). Everson-Rose and Lewis, for example, considered the scope of interest much as did Hemingway and Marmot, and identified 11 published scales for assessing selected psychosocial factors and presented from two to seven questions to illustrate each one.64 Topics were: depressive symptoms; hopelessness; anxiety (trait); hostility/cynical distrust; anger-in; anger-out; social connections; emotional support; availability of emotional support/attachment; job strain; and effort-reward imbalance. Familiarity with the source for each instrument is necessary to understand the concepts at issue, means of deriving the items, availability of reference data, and other issues. Even when a particular instrument is used in successive studies, a subset of items is often selected in preference to use of the entire instrument. This practice, usually motivated by concern about respondent burden when multiple factors are being studied, introduces issues about interpretation and comparability of results among studies. Clear understanding of the relevant methods is important in interpreting data on psychosocial characteristics. As an example, data are presented here on prevalence of self-reported “serious psychological distress” among US adults, aged 18 years and older,
as determined through the National Health Interview Survey, for selected years from 1997–1998 to 2005–2006 (Table 15-8).65 Without knowledge of the questions asked and methods of scoring responses, the prevalence data are meaningless. This condition is defined by a score of 13 or more points on a 24point scale representing the sum of item scores from 0 (none of the time) to 4 points (most of the time) on each of six questions regarding feelings experienced during the past 30 days—sadness, nervousness, restlessness/fidgetiness, hopelessness, worthlessness, or “that everything was an effort.” This “K6 instrument” was designed to identify, with as few questions as possible, persons with “potentially diagnosable mental illness,” that is, persons scoring at or above the threshold of 13 points. With this understanding, the striking gradient in prevalence of serious psychological distress by poverty status becomes informative. This distribution appears consistent with that of depression (see below), which might be expected given the nature of the K6 item content. Determinants and Mechanisms Stress and Stress Research The concept of competing or balancing psychological influences, noted previously, was incorporated explicitly by House in the schematic representation of stress research shown in Figure 15-3.66 Stress is understood here as a subjective phenomenon, perhaps less clearly a “bodily state” than in the view of Selye, Wolff, and Cassel. Although House’s context was occupational stress, the figure could be applied more broadly to encompass each of the psychosocial factors addressed here. Consistent with the definition presented by Hemingway and Marmot, this schematic view represents responses to stress at physiological, cognitive, and behavioral levels as intermediate between social conditions conducive to stress and outcomes, which could include coronary events as “physiological.” Perceived stress is interposed as a necessary link between social conditions and responses. The perception of stress arises from the interaction of “particular objective social conditions and particular personal characteristics.”66, p 14 Reverse processes in response to stress are coping, which ameliorates the effects of underlying social conditions, and defenses, which reduce the perception of stress. These components of the responses to stress, as well as production of outcomes, are all subject to influence by unspecified “conditioning variables,” described as individual or situational. This qualification suggests the possibility of important sources of
46894_CH15_431_502.qxd
450
2/26/10
12:31 PM
Page 450
CHAPTER 15 OTHER PERSONAL FACTORS
Table 15-8
Serious Psychological Distress Among Adults 18 Years of Age and Over, by Selected Characteristics: United States, Average Annual 1997–1998 to 2005–2006 [Data are based on household interviews of a sample of the civilian noninstitutionalized population] Characteristic 1997–1998 1999–2000 2001–2002 2003–2004 2005–2006 Percent of Persons with Serious Psychological Distress1 18 years and over, age-adjusted2,3 18 years and over, crude3 Age 18–44 years 18–24 years 25–44 years 45–64 years 45–54 years 55–64 years 65 years and over 65–74 years 75 years and over Sex2 Male Female Race2,4 White only Black or African American only American Indian or Alaska Native only Asian only Native Hawaiian or Other Pacific Islander only 2 or more races Hispanic origin and race2,4 Hispanic or Latino Mexican Not Hispanic or Latino White only Black or African American only Percent of poverty level2,5 Below 100% 100%–less than 200% 200% or more Hispanic origin and race and percent of poverty level2,4,5 Hispanic or Latino: Below 100% 100%–less than 200% 200% or more Not Hispanic or Latino: White only: Below 100% 100%–less than 200% 200% or more Black or African American only: Below 100% 100%–less than 200% 200% or more
3.2 3.2
2.6 2.6
3.1 3.1
3.1 3.1
3.0 3.0
2.9 2.7 3.0 3.7 3.9 3.4 3.1 2.5 3.8
23 2.2 2.4 3.2 3.5 2.6 2.4 2.3 2.5
2.9 2.8 3.0 3.9 4.2 3.4 2.4 2.4 2.4
2.9 2.8 2.9 3.9 3.9 3.9 2.4 2.3 2.5
2.7 2.1 2.9 3.8 3.9 3.6 2.4 2.2 2.5
2.5 3.8
2.0 3.1
2.4 3.8
2.3 3.9
2.3 3.6
3.1 4.0 7.8 2.0
2.5 2.9 *7.2 *1.4
3.0 3.5 8.1 *1.8
3.1 3.4 *5.5 *1.8
2.8 3.7 *4.7 2.3
* 4.8
* 5.0
* 9.1
* 5.5
5.0 5.2 3.0 2.9 3.9
3.5 2.9 2.5 2.4 2.9
4.0 3.8 3.1 3.0 3.5
3.9 3.6 3.1 3.0 3.3
3.3 3.3 2.9 2.8 3.7
9.1 5.0 1.8
6.8 4.4 1.6
8.4 5.2 2.0
8.8 5.2 1.8
7.6 5.2 1.7
8.6 5.4 2.9
6.1 3.8 2.2
7.5 4.1 2.9
7.4 3.7 2.3
5.1 3.2 2.6
9.6 5.2 1.8
7.8 4.9 1.6
9.2 5.9 2.0
10.4 6.1 1.7
8.9 6.1 1.6
8.7 4.3 1.6
6.0 3.6 1.3
7.2 4.9 1.7
7.4 4.1 1.5
7.4 5.0 1.7
— —
46894_CH15_431_502.qxd
2/26/10
12:31 PM
Page 451
2. ADVERSE PSYCHOSOCIAL FACTORS 451
Table 15-8
Serious Psychological Distress Among Adults 18 Years of Age and Over, by Selected Characteristics: United States, Average Annual 1997–1998 to 2005–2006—continued Characteristic 1997–1998 1999–2000 2001–2002 2003–2004 2005–2006 Geographic region2 Northeast 2.7 1.9 2.8 2.9 2.6 Midwest 2.6 2.5 2.9 2.7 2.8 South 3.8 2.9 3.5 3.5 3.4 West 3.3 2.8 3.0 3.0 2.6 Location of residence2 Within MSA6 3.0 2.3 3.0 2.9 2.8 Outside MSA6 3.9 3.5 3.8 3.8 3.9 *Estimates are considered unreliable. Data preceded by an asterisk have a relative standard error (RSE) of 20%–30%. Data not shown have an RSE greater than 30%. - - - Data not available. 1 Serious psychological distress is measured by a six-question scale that asks respondents how often they experienced each of six symptoms of psychological distress in the past 30 days. See Appendix II, Serious psychological distress. 2 Estimates are age-adjusted to the year 2000 standard population using five age groups: 18–44 years, 45–54 years, 55–64 years, 65–74 years, and 75 years and over. See Appendix II, Age adjustment. 3 Includes all other races not shown separately. 4 The race groups, White, Black, American Indian, or Alaska Native, Asian, Native Hawaiian or Other Pacific Islander, and 2 or more races, include persons of Hispanic and non-Hispanic origin. Persons of Hispanic origin may be of any race. Starting with 1999 data, race-specific estimates are tabulated according to the 1997 Revisions to the Standards for the Classification of Federal Data on Race and Ethnicity and are not strictly comparable with estimates for earlier years. The five single-race categories plus multiple-race categories shown in the table conform to the 1997 Standards. Starting with 1999 data, race-specific estimates are for persons who reported only one racial group; the category 2 or more races includes persons who reported more than one racial group. Prior to 1999, data were tabulated according to the 1977 Standards with four racial groups and the Asian only category included Native Hawaiian or Other Pacific Islander. Estimates for single-race categories prior to 1999 included persons who reported one race or, if they reported more than one race, identified one race as best representing their race. Starting with 2003 data, race responses of other race and unspecified multiple race were treated as missing, and then race was imputed if these were the only race responses. Almost all persons with a race response of other race were of Hispanic origin. See Appendix II, Hispanic origin; Race. 5 Percent of poverty level is based on family income and family size and composition using US Census Bureau poverty thresholds. Missing family income data were imputed for 26%–30% of persons 18 years of age and over in 1997–1998 and 32%–35% in 1999–2006. See Appendix II, Family income; Poverty. 6 MSA is metropolitan statistical area. Starting with 2005–2006 data, MSA status is determined using 2000 census data and the 2000 standards for defining MSAs. For data prior to 2005, see Appendix II, Metropolitan statistical area (MSA) for the applicable standards. Notes: Standard errors for selected years are available in the spreadsheet version of this table. Available from: http://www.cdc.gov/nchs/hus.htm. Data for additional years are available. See Appendix III. Sources: CDC/NCHS, National Health Interview Survey, family core questionnaire. Data from Health, U.S., 2008, pp 291–292, National Center for Health Statistics, Centers for Disease Control and Prevention, Department of Health and Human Services.
CONDITIONING VARIABLES: Individual or Situational
RESPONSES TO STRESS 1. Physiological 2. Cognitive/Affective 3. Behavioral 3
ing)
(d
PERCEIVED STRESS 2
ef en
se s)
(cop
SOCIAL CONDITIONS CONDUCIVE TO STRESS 1
5
OUTCOMES 1. Physiological 2. Cognitive/Affective 3. Behavioral 4
Figure 15-3 A Paradigm of Stress Research. Solid arrows between boxes indicate presumed causal relationships among variables. Dotted arrows from the box labeled “Conditioning Variables” intersect solid arrows, indicating an interaction between the conditioning variables and the variables in the box at the beginning of the solid arrow in predicting variables in the box at the head of the solid arrow. Source: Reprinted with permission from J House, A Paradigm of Stress Research, Journal of Health and Social Behavior, p 13, © American Sociological Association.
46894_CH15_431_502.qxd
452
2/26/10
12:31 PM
Page 452
CHAPTER 15 OTHER PERSONAL FACTORS
variability in response to social conditions and perceived stress. Such variability might appear systematically between, for example, demographic groups within a population. Jenkins, in proceedings of the National Workshop Conference on Socioenvironmental Stress and Cardiovascular Disease, included health-related behaviors such as poor dietary habits, physical inactivity, and smoking among the consequences of socioenvironmental stresses, not limiting this concept to “those [stimuli] that repeatedly arouse the autonomic nervous system.”67, p 149 Such effects would constitute an important set of mediating or conditioning variables through which social conditions lead to physiological and other outcomes. Jenkins also implied that stresses need not be perceived to produce responses, as they may operate even if psychological defenses prevent their conscious recognition. Study of stress that is not perceived would pose further methodologic difficulties. Triggers and Their Effects The preceding discussion implicitly concerns chronic social stress and its consequences. A distinct scheme to represent acute psychological effects on coronary heart disease is depicted in Figure 15-4.68 According to this scheme, emotional triggers can lead to acute
clinical cardiovascular events—arrhythmias, myocardial infarction, or unstable angina. Intervening physiological responses include a wide array of mechanisms, and their immediate pathophysiological effects lead directly to the outcomes identified. Other representations of these phenomena, whether chronic or acute in nature, have been developed in greater or lesser detail as, for example, by Everson-Rose, by Rozanski, and by Krantz and McCeney.60,64,69 Their common focus on “stress,” the concept eschewed as “unhelpful” by Hemingway and Marmot, is supported by others. For example, Bairey Merz and colleagues identified psychosocial stress as:70, p 141 “a newly recognized (nontraditional) risk factor that appears to contribute to all recognized mechanisms underlying cardiac events, specifically, (a) clustering of traditional cardiovascular risk factors, (b) endothelial dysfunction, (c) myocardial ischemia, (d) plaque rupture, (e) thrombosis, and (f) malignant arrhythmias.” Whether “stress” is best considered as distinct from the four more specific factors addressed as follows or as related to them in some yet to be specified way is unclear. Favorable Psychosocial and Behavioral Attributes Lastly, by way of general background, psychosocial and behavioral attributes can also be considered as
Emotional trigger
Physiological responses Coronary vasoconstriction
Autonomic dysfunction
Neuroendocrine activation
Inflammatory response
Hemodynamic response
Prothrombolic response
Pathophysiological effects Cardiac electrical instability
Myocardial ischemia
Plaque disruption
Thrombus formation
Clinical events Ventricular tachycardia/ fibrillation
Myocardial infarction
Unstable angina
Figure 15-4 Hypothesized Links Between Acute Emotional Triggers and Cardiac Events Mediated Through Physiological Response and Pathophysiological Effects. Source: Reprinted with permission from Progress in Cardiovascular Diseases, Vol 49, No 5, MR Bhattacharyya, A Steptoe, p 354, © Elsevier Inc.
46894_CH15_431_502.qxd
2/26/10
12:31 PM
Page 453
2. ADVERSE PSYCHOSOCIAL FACTORS 453
having potential favorable effects in relation to cardiovascular risk. The scheme illustrated in Figure 15-3 suggests that both positive coping skills and relevant defenses against adverse effects of social stress should be counted as personal psychological assets, whereas their lack represents a liability. Individual or group differences in these attributes may be innate or instead attributable to learned skills and behaviors, whether as a result of life experience in general or of more specifically directed health education, counseling, or other interventions. In the area of health-promoting behavior and adherence to preventive or therapeutic programs, favorable psychosocial factors may be considered influential both in protecting from effects of adverse social conditions and ensuring the effectiveness of preventive measures.71 Against this background, four of the more prominently investigated aspects of psychosocial profile and related conditions are reviewed here: Type A behavior pattern, depression, occupational stress, and social support. Type A Behavior Pattern Development of the Concept In 1959, Friedman and Rosenman reported that they had identified an “overt behavior pattern” that their results suggested was largely responsible for the marked increase in frequency of coronary heart disease found in one of three groups of men studied.72 In a preliminary study, they surveyed more than 200 business executives and physicians and asked them their opinions about major causes of coronary heart disease. They found “chronic exposure to emotional trauma” to be a dominant theme in the responses, the trauma being attributed to conditions such as competitiveness, consciousness of deadlines, and related characteristics. The concept that this set of conditions might cause coronary heart disease led to a further study in which they examined a group of accountants before, during, and after episodes of such trauma. The accountants exhibited acute increases in blood cholesterol concentration and blood coagulability during the exposure period. On the basis of these two studies, the investigators then defined three categories of overt behavior pattern, designated patterns A, B, and C. Pattern A incorporated the following characteristics:72, p 1286 (1) an intense, sustained drive to achieve selfselected but usually poorly defined goals, (2) profound inclination and eagerness to compete, (3) persistent desire for recognition and advancement, (4) continuous involvement in mul-
tiple and diverse functions constantly subject to time restrictions (deadlines), (5) habitual propensity to accelerate the rate of execution of many physical and mental functions, and (6) extraordinary mental and physical alertness. Pattern B was defined as lacking all features of pattern A. Pattern C differed from pattern B only in being accompanied by chronic anxiety. Initial Findings To test the prediction that men exhibiting pattern A would have a greater tendency to manifest coronary heart disease and some of its risks, Friedman and Rosenman used nonmedical volunteers to recruit groups of men who conformed to these respective patterns.72 They interviewed the men to document the degree to which they exhibited the patterns and examined them with respect to clinical evidence of coronary heart disease (by history and electrocardiography) and other characteristics. Men in groups A and B were categorized as exhibiting a “completely” or “incompletely” developed pattern on the basis of the interviews. The frequencies of clinical coronary heart disease in the five groups were as follows: A (complete), 34%; A (incomplete), 28%; B (complete), 0; B (incomplete), 4%; and C, 4%. The Western Collaborative Group Study (WCGS) Of the many studies of Type A behavior that followed, the most direct follow-up to the initial investigation was the WCGS, organized by Rosenman and Friedman with participation of 10 California companies.73 From 1960 to 1961, the WCGS enrolled 3154 healthy men age 39–59 years at entry. Through a structured interview, each participant was classified as complete or incomplete Type A or B or neither Type A nor B. The few remaining “indeterminate” individuals were grouped separately. The main results were reported after 8.5 years of follow-up. The overall frequencies of new events by age at entry were, for the age group 39–49 years, 95/1067 (8.9%) in Type A and 50/1182 (4.2%) in Type B and, for the age group 50–59 years, 83/522 (15.9%) in Type A and 29/383 (7.6%) in Type B. The relation of behavior pattern and other risk factors to the incidence of coronary heart disease in multivariate analysis indicated significant associations with age, smoking, systolic blood pressure, total cholesterol concentration, and behavior pattern for men in the age group 39–49 years and for the same factors except age and systolic blood pressure for the older group. The estimated relative risks for Type A behavior pattern after adjustment for other factors were
46894_CH15_431_502.qxd
454
2/26/10
12:31 PM
Page 454
CHAPTER 15 OTHER PERSONAL FACTORS
1.87 and 2.16 in the two age groups, or an increase of 1.37 and 1.46 per standard deviation unit of the two-point score from B 0 to A 1. The investigators interpreted the results as indicating direct effects of Type A behavior pattern, perhaps through neurohumoral pathways, in addition to any effect through the other risk factors. Modifiability of Type A Behavior Pattern In subsequent studies, intervention to alter Type A behavior was undertaken among 862 men who previously experienced myocardial infarction.74 One group of 592 patients received group counseling concerning Type A behavior in addition to cardiology counseling, whereas 270 patients received only cardiology counseling. Intervention for Type A behavior was described as including muscle relaxation techniques, behavioral training to recognize and modify Type A manifestations, environmental changes, and changes in values and goals. After 3 years of study, Friedman and colleagues reported net reduction in scores for Type A behavior that were greater in the intervention than in the control group by each of several measures used, thus supporting the view that Type A behavior could be modified. Survival analysis indicated greater duration of follow-up time free of recurrent myocardial infarction or death in the intervention group. Analysis of the groups as randomly allocated, irrespective of treatment adherence, indicated recurrence rates of 7.2% versus 13.2% in treated versus control men (p 0.005). Extended Follow-Up of Cases in the WCGS However, subsequent study of the relation of Type A behavior to recurrent coronary heart disease added to inconsistencies in the accumulating observations. Prospective assessment of the association of Type A behavior with risk of coronary heart disease was undertaken in the setting of the Multiple Risk Factor Intervention Trial (MRFIT). No increase in risk was found in men classified at baseline as Type A.75 To study further experience of the WCGS, 257 men who had experienced coronary events in the initial 8.5 years of follow-up were evaluated in greater detail to determine the relation of Type A behavior to mortality from coronary events.76 Of the 257 men, 135 had symptomatic acute myocardial infarction or sudden death as the initial coronary event. Of these, 18 of 93 Type A men (19.4%) and 8 of 42 Type B men (19.0%) died within 24 hours, indicating no relation between Type A and 24-hour mortality. Of 231 men who survived the first 24 hours, subsequent mortality was 19.1% for Type A men and 31.7% for Type B men (p 0.04). Figure 15-5 demonstrates this result as the cumulative proportion of each group dy-
ing of coronary heart disease through a total of 22 years of follow-up. This result indicated that men with coronary heart disease and Type A behavior experienced a decreased risk of death from coronary heart disease relative to men classified as Type B, opposite the predicted result. Hostility and Anger Some questioned whether the Type A behavior pattern was sufficiently specific to identify increased coronary risk in diverse populations. Perhaps within the overall Type A construct, an element such as hostility or anger was the specific component of interest. Accordingly, increasing emphasis has been placed on this characteristic. For example, a group of 255 physicians was followed for 25 years after questionnaire assessment of health characteristics through the MMPI, which includes items on hostility.77 The median score divided the group into higher and lower levels of hostility with coronary events at rates of 4.5/1000 and 0.9/1000 person-years, respectively (p 0.0005). Episodes of anger within 2 hours preceding onset of myocardial infarction, or over the preceding year, were the focus of an innovative study design to evaluate triggering of acute coronary events.78 The comparison was between the occurrence of such episodes among cases and among controls during corresponding periods in relation to the time of onset of the case. The estimated relative risk of myocardial infarction within 2 hours of an episode of anger among 39 patients, in accordance with a scale devised for this study, was 2.3 (95% confidence interval [CI] 1.7–3.2). The experience through the middle to late 1970s with the Type A behavior pattern, the self-administered Jenkins Activity Survey (JAS) devised to identify some key components of Type A, and other studies in this area were reviewed extensively by Jenkins and in the report of a Forum on Coronary-Prone Behavior.79–81 These reviews identified many of the issues in this area of research prior to its more recent focus on component factors such as hostility and anger and methodologic aspects of these studies. More recently, several factors including two derived from the Framingham Type A Questionnaire were investigated in relation to risk of hypertension in the Coronary Artery Risk Development in Young Adults (CARDIA) Study.82 The Type A components were termed “time urgency/impatience (TUI)” and “achievement striving/competitiveness (ASC).” Other measures were taken from different scales and included hostility, depression, and anxiety. TUI and hostility were both associated with 15-year incidence of hypertension, whereas ASC, depression, and anxiety were not associated. Findings were generally sim-
46894_CH15_431_502.qxd
2/26/10
12:31 PM
Page 455
2. ADVERSE PSYCHOSOCIAL FACTORS 455
Cumulative Proportion Dying of CHD
0.40
0.30
0.20
Type A = 0.10 Type B =
0.00 0
2.2
4.4
6.6
8.8
11.0
13.2
15.4
17.6
19.8
22.0
Years after the Initial CHD Event
Figure 15-5 Cumulative Case Fatality Rates Among 231 Patients with Coronary Heart Disease (CHD) Who Survived for 24 Hours, by Behavior Pattern. Source: Reprinted with permission from DR Ragland and RJ Brand, Type A Behavior and Mortality from Coronary Heart Disease, The New England Journal of Medicine, Vol 318, pp 65–69, © 1998, The Massachusetts Medical Society. All rights reserved.
ilar across sex and race/ethnic groups. Such findings have tended to sustain interest in identifying a consistent and interpretable indicator of what the Type A behavior pattern represented. Studies in Adolescents All the foregoing observations were based on coronary events among adults. However, relations between Type A behavior pattern, hostility, and anger and their suppression or expression have been studied in children and adolescents as well for their possible relation to known risk factors for atherosclerosis. A review of this literature found inconsistent results, possibly because of methodologic issues.83 Aronowitz, in his chapter “The Rise and Fall of the Type A Hypothesis,” considered the nature of this hypothesis as part of the risk factor approach to understanding disease. He concluded:84, p 165 While the particular formulation of the relationship between body and mind that the Type A hypothesis represented no longer plays a significant role in mainstream medicine, and new or sometimes merely fashionable options for
understanding such relationships are ceaselessly provided by the laboratory and the study of populations, investigators, doctors, and patients continue to negotiate cause and responsibility for health and illness within historically familiar and culturally determined boundaries. Occupational Stress Concepts of Occupational Stress and Job Strain Conditions of work affect employed persons during a significant portion of their lives and could, if they influence health adversely, be expected to be associated with commonly occurring health problems, including coronary heart disease. This area has received attention for several decades and was reviewed by Theorell, a major contributor to the field.59 Concepts now common to discussions in this area include demand, control, decision latitude, and others. These aspects of the work situation are traceable in part to the work of Alfredsson and colleagues, who devised a simple conceptual scheme in the form of a 2 2 table.85 Control and possibility for growth and development
46894_CH15_431_502.qxd
456
2/26/10
12:31 PM
Page 456
CHAPTER 15 OTHER PERSONAL FACTORS
were high or low, and demands were high or low. “Job strain” was represented by the adverse combination of high demand and low control. Studies of Job Strain and Iso-Strain In a case-comparison study of Swedish men younger than age 65 years, occupational data for 334 cases of fatal and nonfatal myocardial infarction and 882 controls were analyzed.85 Occupations of the men were classified by their psychological characteristics on the basis of a national survey among workers in those occupations. For young men, younger than age 50 years, the dual characteristic of “hectic” work and lack of control over the tempo of work (i.e., strain) resulted in approximately a twofold risk of myocardial infarction, though details of the analysis were not provided. The relation of job strain to prevalence of indicators of previous myocardial infarction was studied in data from two US national health surveys conducted in 1960–1962 and 1971–1975.86 In this study, jobs characterized as having job strain were associated with substantially greater prevalence of myocardial infarction, as shown in Figure 15-6. Relative risk estimates for job strain as assigned to the reported occupations were 3.8 and 4.8 in the highest versus the lowest deciles of the scale in the two respective surveys. The magnitude of these estimates was similar to relative risks found for smoking and cholesterol concentrations. Despite limitations of this cross-sectional analysis, implications were drawn for organization of the work environment and consideration of decisionmaking powers of workers. Prospective data are illustrated by a Swedish study of 7219 working men with 9 years’ followup.87 Cardiovascular diseases, including coronary heart disease, stroke, and peripheral vascular disease, accounted for 193 deaths over this period. Workers were classified according to a composite index termed “iso-strain,” to reflect both social isolation at work and job strain, as previously defined. Figure 15-7 indicates, for blue-collar workers, the probability of cardiovascular death in 9 years for three strata of the population—those with high, intermediate, and low iso-strain. For all age groups, those with least isolation and job strain had the lowest probability, whereas both the intermediate and highest groups had high probabilities of cardiovascular death. For whitecollar workers, similar curves were presented, although at lower probabilities of mortality at all ages. Future studies were proposed in which social class would be taken more fully into account to investigate variation in contributions of iso-strain among different social and occupational strata of the population.
Subclinical atherosclerosis was used as the cardiovascular outcome in a further study of the simpler 2 2 classification of work status described previously.88 Carotid artery intima-media thickness (IMT) and prevalence of carotid plaque were measured and compared across the four strata of more than 2600 Swedish working men and women, aged 46 to 65 years. No consistent pattern was apparent to differentiate workers with job strain in comparison with those in the low demand/high control group, taken as the reference. The major risk factors compared across groups showed only very few statistically significant differences. It has been suggested that an association of job strain with cardiovascular outcomes in some studies might result from confounding by pre-employment health status. To address this question, analysis of data from the Cardiovascular Risk in Young Finns Study compared carotid IMT of workers at ages 33–39 years who had been examined regarding risk factors at ages 12–18 years.89 It was found that IMT was predicted by pre-employment risk factor levels, but that job strain carried a dose-response gradient beyond the risk-factor prediction. It was proposed on the basis of these results that the suspected confounding did occur but did not fully account for the association with job strain and that intervention studies should be undertaken to test this association more rigorously. Depression A substantial body of epidemiologic evidence addresses depression in relation to coronary heart disease from as early as the mid-1960s, with growing numbers of studies in the 1990s and early 2000s. A number of critical reviews include a particularly insightful one by Frasure-Smith and Lesperance in 2005.90 Their review began with the observation that behavioral medicine specialists, but not cardiologists or their major national (North American) organizations, believed depression to be an important cardiovascular risk factor. To assess reasons why the evidence might not be persuasive outside behavioral medicine, they searched the literature and identified 21 etiologic and 43 prognostic publications and noted, pointedly, that these primary research reports were outnumbered by 79 published review articles. Definitions of depression were identified in this review as a fundamental difficulty. They found 23 approaches to measurement within the set of studies qualifying for review. Further, “These counts do not include studies using different cut-off points for the same measure, those using cut-off points versus continuous scores, those operationalizing depression us-
46894_CH15_431_502.qxd
2/26/10
12:31 PM
Page 457
2. ADVERSE PSYCHOSOCIAL FACTORS 457
HANES 9 Other
Percent MI
High Strain 6
3
0 Age
N (OTHER) N (HIGH STRAIN)
25–34
35–44
45–54
55–64
65+
490 151
390 108
531 149
366 102
165 31 30.7
HES 9 Other
Percent MI
High Strain 6
3
0 Age
<25
N (OTHER) 194 N (HIGH STRAIN) 113
25–34
35–44
45–54
55–64
65+
466 141
491 146
378 103
273 57
91 13
Figure 15-6 Prevalence of Myocardial Infarction by Age and Job Strain Among Employed Males, US Health and Nutrition Examination Survey and Health Examination Survey. Source: From Chronic Disease Epidemiology and Control, © 1993 by the American Public Health Association. Reprinted with permission.
ing different combinations of the same measures, or those measuring increases in depression over time versus baseline depression.”90, p S19 Another problem in assessing the literature is occurrence of multiple publications from the same study—not duplicated reports of the same material, but different measures,
approaches to analysis, time periods of follow-up, or outcomes. Inability to identify truly independent reports limits confidence in weighing the results from multiple publications. Variation in outcome measures was also striking in that 66 different outcomes were studied, with inadequate definitions of terms
46894_CH15_431_502.qxd
458
2/26/10
12:31 PM
Page 458
CHAPTER 15 OTHER PERSONAL FACTORS
.20
.18
.16
Probability of CVD Mortality
.14
.12
.10
.08
.06
.04
.02
0 25
30
35
40
45
50
55
60
65
Age Groups
Figure 15-7 Estimates of Age Trends in Nine-Year Mortality from Cardiovascular Disease (CVD) Within a Random Sample of BlueCollar Working Men in Sweden (N 4235) by Low ( ), Medium ( ), and High ( ) Levels of Iso-Strain (i.e., Combination of Social Isolation and Job Strain. Source: Reprinted with permission from JV Johnson, EM Hall, and T Theorell, Scandinavian Journal of Work and Environmental Health, Vol 15, p 276, © 1989, University of Massachusetts, Lowell.
and insufficient documentation of outcomes. Selection of covariates for inclusion varied greatly among studies, such that results of analysis with covariate adjustment could not be compared across studies. The plethora of reviews was taken by FrasureSmith and Lesperance to result from subjectivity and personal bias in reviewing this literature, such that many investigators are stimulated to publish their own interpretations. The authors of this review concluded:90, p S23 Despite multiple methodological differences from study to study, the data from prospective adequately powered etiologic and prognostic
studies with objective outcome measures and recognized indices of depression are remarkably consistent in their support of depression as a risk factor for both the development of and the worsening of CHD. A conceptual scheme, or “biobehavioral model,” to link depression with clinical events (e.g., acute myocardial infarction, sudden cardiac death) was presented by Lett and others.91 Depression was seen as leading to two distinct but intersecting pathways, each involving an explicit set of multiple “plausible” factors. One pathway was through behavioral risk factors—smoking, alcohol, medical adherence, and
46894_CH15_431_502.qxd
2/26/10
12:31 PM
Page 459
2. ADVERSE PSYCHOSOCIAL FACTORS 459
physical activity. The other was through physiological risk factors—platelet activity, hypothalamicpituitary axis dysregulation, autonomic nervous system dysregulation, inflammation, and traditional risk factors (diabetes, obesity, and hypertension). Depression was thus a multipotent influence leading, in their assessment, to an overall relative risk of 1.5 to 2.0 for onset of coronary artery disease and of 1.5 to 2.5 for fatal and nonfatal outcomes among persons with existing coronary disease. Other perspectives are informative as well. Rugulies, for example, distinguished between the effects of depressive mood and clinical depression.92 Standardized psychometric scales were used to rate depressive mood (10 studies), and clinical procedures identified clinical depression (three studies). A stronger relation with coronary heart disease outcomes (either fatal or combined fatal and nonfatal) was found with clinically defined depression (Figure 15-8). The overall relative risk (based on the most extensively adjusted estimate from each study) was 1.49 (1.16, 1.92) for depressive mood and 2.69 (1.63, 4.43) for clinical depression. Rugulies noted the joint relation of depression and risk of coronary heart dis-
ease with lower social class and the potential importance of a “larger bio-psychological-social model” of coronary heart disease. Under such a framework, research would include both social and psychological constructs, rather than only one of these domains, in addition to biomedical aspects. Stansfield and Rasul addressed the question whether “psychological distress,” usually evidenced by depressive symptoms, reflects underlying physical ill health that is the more direct predictor of coronary risk.93 Eight prospective studies were reviewed, predominantly community-based and typically with 5 to 15 years of follow-up for fatal and nonfatal coronary outcomes. After adjustment for subclinical disease and self-reported physical illness at baseline, psychological distress was found to be independently related to coronary outcomes. The Global Burden of Disease Study (GBDS) identified a major contribution of depression to disability, on a global scale.94 By taking disability (measured in DALYs) into account, in addition to mortality alone, the impact of unipolar major depression could be estimated. In the GBDS projection, this condition would advance from fourth to second place among
First Author, Year (Reference) 1
Anda, 1993 (35)
1.5
2.3 3.36
Aromaa, 1994 (31) 1.23
Barefoot, 1996 (32) Ferketich, 2000 -women- (41)
0.4 0.74
1.48 1.28
Ferketich, 2000 -men- (41) Ford, 1998 (42) Mendes de Leon, 1998 -men- (33)
4.16
1.49 1.34
Schwartz, 1998 (34) 0.4
2.02 1
Whooley, 1998 (45)
3
1.7
1.29 1.64
All Studies (n ⫽ 13)
4.59
1.88
0.9
11.62
3.71
2.33
0.77
Sesso, 1998 (44)
2.08
1.16 1.49 1.92
Depressive Mood Only (n ⫽ 10)
1.63
Clinical Depresssion Only (n ⫽ 3) 0
2.9
1.67 1.42
Pratt, 1996 (43)
Wassertheil-Smoller, 1996 (36)
4.06
2.12
0.96 0.34 0.7
3.38
2.08
1.11
Mendes de Leon, 1998 -women- (33)
2.97
1.91
0.5
1
1.5
2.69
2
2.5
4.43
3
3.5
4
4.5
5
Relative Risk
Figure 15-8 Individual and Overall Relative Risks (Most-Adjusted) for Coronary Heart Disease in Depressed Subjects. Source: Reprinted with permission from American Journal of Preventive Medicine, Vol 23, R Rugulies, p 56, © 2002, American Journal of Preventive Medicine.
46894_CH15_431_502.qxd
460
2/26/10
12:31 PM
Page 460
CHAPTER 15 OTHER PERSONAL FACTORS
the 15 leading causes of burden of disease, from 1990 to 2020. The sequel to the GBDS, the Global Burden of Disease and Risk Factors, updated the information on which such projections could be based by adding available survey data regarding severity of unipolar depressive disorders.95 In all, 56 studies were available from all World Bank regions, with resulting prevalence estimates applied to the population sizes estimated for the respective regions and subregions. These data contributed to assessment of the burden of several major psychiatric disorders under the Disease Control Priorities Project—schizophrenia, bipolar disorder, depression, and panic disorder.96 Among these four conditions, depression was by far the greatest contributor, causing 51.8 million DALYs or 30.8% of the total neuropsychiatric disease burden and 3.4% of the total disease burden globally. Social Support Concepts and Definitions Several concepts of interpersonal relationships and their effects on health are linked with the term “social support.” Berkman emphasized the distinction between social networks and social support, with the admonition that an individual’s identification with a social network should not be presumed to constitute social support.97 It was therefore considered important to recognize and assess several dimensions of each. Social networks were defined as “the web of social ties that surrounds an individual,”97, p 414 with the following structural dimensions: density and complexity, that is, measures of interaction within the group; group size; symmetry or reciprocity, defined as equality of supports and obligations among members; geographic proximity or dispersion; homogeneity, assessed with respect to age, social class, and religion; and accessibility. Social support, in turn, was defined as “the emotional, instrumental, and financial aid that is obtained from one’s social network.”97, p 415 The difficulty of separating objective, external measures of support from the subjective, psychological states of emotion makes this definition problematic. A more detached concept of social support, attributed to House, involves “a transaction of (a) emotional concern, (b) instrumental aid (goods and services), (c) information, or (d) appraisal (information relevant to self-evaluation).”97, p 415 Berkman discussed methodologic aspects of research in this area were review, including matters of definition and measurement. Several studies of social networks, social support, and physical health were also summarized, with the criticism that measure-
ment of either social networks or social support was generally inadequate. Findings of Selected Studies House and colleagues reviewed social support under a perhaps broader concept, “social relationships,” using a scale of “level of social integration” (both undefined).98 Their review cited the seminal contributions of Cassel and of Cobb. Each of them had assembled prior work on stress and psychosocial factors, as studied in both human and animal experiments, up to the mid-1970s, and formulated theoretical frameworks for future research. Several early studies were summarized by House and others. All were prospective studies of mortality, representing populations in Finland and Sweden, as well as both Blacks and Whites in the United States. In each study population, though at different levels of mortality, higher mortality was associated with low levels of social integration. Relative risks of lowest versus highest categories ranged, for men, from 1.08 in Evans County, Georgia, Blacks to 4.00 in Gothenburg, Sweden, and, for women, from 1.07 in Evans County Whites to 2.81 in Alameda County, California. Also, the report briefly discussed conclusions from animal experiments and human clinical studies that suggested potential neuroendocrine pathways to link psychosocial phenomena with physiologic responses and their adverse health effects. In the view of House and colleagues:98, p 543 The evidence on social relationships is probably stronger, especially in terms of prospective studies, than the evidence which led to the certification of the Type A behavior pattern as a risk factor for coronary heart disease. The evidence regarding social relationships and health increasingly approximates the evidence in the 1964 Surgeon General’s report that established cigarette smoking as a cause or risk factor for morbidity and mortality from a range of diseases. The report noted the need for research in three areas: determinants of “exposure” to social relationships, mechanisms linking such exposures with health, and means of intervention against “relative social isolation.” House and colleagues presented prospective studies only because these provided direct evidence of time sequence between social circumstances and the health outcomes assessed. However, work by Reed and colleagues in the Honolulu Heart Program, also prospective, resulted in different conclusions.99 Among 4251 men of Japanese ancestry classified ac-
46894_CH15_431_502.qxd
2/26/10
12:31 PM
Page 461
2. ADVERSE PSYCHOSOCIAL FACTORS 461
cording to a variety of psychosocial characteristics in 1971, 7-year incidence of all diseases combined was unrelated to social networks, and persons in the highest categories of mobility and status inconsistency evidenced no lesser health outcomes in the presence than in the absence of social networks. Only coronary heart disease incidence, interestingly, was associated with the social network scale, the rates ranging in a consistent gradient from 45/1000 men over the seven years in the lowest to 30/1000 in the highest category (p 0.05). After multivariate analysis taking known risk factors into account, this association was no longer significant, except marginally so for nonfatal myocardial infarction alone. Whether this suggests confounding of the association by other risk factors or operation of social network effects through those risk factors is unresolved. Reed and colleagues concluded that their measures of social networks did not indicate an effect on general susceptibility to illness. They noted, however, that cultural differences among particular groups should perhaps be expected to yield varying findings and offered several suggestions for strengthening epidemiologic research in this area. Such cultural specificity as may differentiate Japanese men in Hawaii from other groups studied previously was investigated within a single, biethnic community, Corpus Christi, Texas, as reported more recently.100 Farmer and colleagues devised a social support scale based on data previously collected in hospital interviews of persons admitted with acute myocardial infarction. The available items were related to marital status, solitary living conditions, and history of advice from others to seek help in the course of the immediate illness. A scale was constructed with five levels, from low to high social support. Survival time by social support category was determined over 55 months of follow-up for 596 Mexican-American and non-Hispanic White survivors of acute myocardial infarction, beginning after initial survival for the first 28 days from onset of symptoms. In the total study population, low social support was associated with lesser survival over most of the follow-up period, although by 55 months the groups had experienced similar mortality. Of particular interest was the finding of an ethnic group difference in the association between low social support and mortality, as shown in Table 15-9a and b. For Mexican Americans the relative risk of death in 55 months for those with low versus high social support was 3.38 (CI 1.73–6.62), whereas for non-Hispanic Whites the corresponding relative risk was 1.21 (CI 0.64–2.30). Thus, the same measure of social support appeared to
Table 15-9a
Low Social Support and Other Factors and Relative Risk of Death in 55 Months After Initial Survival of Hospitalized Myocardial Infarction in Mexican Americans (Panel a) Variable RR 95% CI Low social support 3.38 1.73–6.62 60 years of age 1.73 1.19–2.51 Employed vs unemployed 0.38 0.15–0.92 Smoker vs non-smoker 1.50 0.86–2.61 Diagnosed Diabetes mellitus 1.98 1.10–3.25 Hypertension 0.40 0.23–0.70 Note: The relative risk and 95% confidence interval were calculated after excluding from the regression model the variables that were not statistically significant. RR, relative risk; CI, confidence interval. Source: Behavioral Medicine, Vol 22, pp 59–66, 1996. Reprinted with permission of the Helen Dwight Reid Educational Foundation. Published by Heldref Publications, 1319 Eighteenth St, NW, Washington, DC 20036-1802. Copyright © 1996.
Table 15-9b
Low Social Support and Other Factors and Relative Risk of Death in 55 Months After Initial Survival of Hospitalized Myocardial Infarction in Non-Hispanic Whites (Panel b) Variable RR 95% CI Low social support 1.21 0.64–2.30 60 years of age 2.30 1.43–3.70 Employed vs unemployed 0.59 0.29–1.23 Smoker vs non-smoker 1.45 0.80–2.61 Diagnosed Diabetes mellitus 1.86 1.07–3.24 Hypertension 0.89 0.51–1.55 Note: The relative risk and 95% confidence interval were calculated after excluding from the regression model the variables that were not statistically significant. RR, relative risk; CI, confidence interval. Source: Behavioral Medicine, Vol 22, pp 59–66, 1996. Reprinted with permission of the Helen Dwight Reid Educational Foundation. Published by Heldref Publications, 1319 Eighteenth St, NW, Washington, DC 20036-1802. Copyright © 1996.
have quite different health implications between two ethnic groups in the same community, a finding that underscores the comments of Reed and colleagues. In a recent assessment of epidemiologic work in the area of social support, Lett and others identified eight reports on coronary heart disease outcomes in persons initially free of disease and 19 reports on outcomes in coronary patients. They found overall support for a relative risk from 1.5 to 2.0 for both first and recurrent coronary heart disease outcomes. Their conclusion was qualified, however:101, p 869
46894_CH15_431_502.qxd
462
2/26/10
12:31 PM
Page 462
CHAPTER 15 OTHER PERSONAL FACTORS
Although low levels of support are associated with increased risk for CHD events, it is not clear what types of support are most associated with clinical outcomes in healthy persons and CHD patients. The development of a consensus in the conceptualization and measurement of social support is needed to examine which types of support are most likely to be associated with adverse CHD outcomes. Other Characteristics A number of other psychosocial characteristics have been addressed in epidemiologic studies of cardiovascular diseases. Each contributes to the persisting view that health in general or cardiovascular health in particular reflects the influence of interpersonal relations and other psychosocial processes and the mechanisms that may link them with human physiology and pathology. These factors include stressful life events, perception of illness, disparities in social status with one’s spouse or other associates, discontinuities between current life situation and upbringing, marital and employment status, John Henryism (hard work and determination to overcome psychosocial stressors), and the stress of experiencing natural disasters, among others. The methodologic issues in research on these topics have much in common with those addressed in connection with Type A behavior pattern, occupational stress, depression, and social support. Some of these are addressed in the review by Theorell cited previously.59 The INTERHEART Study, involving more than 11,000 cases of acute myocardial infarction and 13,000 controls from 52 countries, included measures of psychosocial factors.102 For this cross-cultural and multilingual study, an instrument was needed to provide a “standard yet simple set of questions that inquired about psychosocial conditions during the previous 12 months.” The topics addressed, with their corresponding odds ratios and population attributable fractions (“PAR” in the table), are shown in Table 15-10: stress at work, stress at home, general stress, financial stress, stressful life events, locus of control, feeling depressed, and depression. Feeling depressed and clinical depression were distinguished from each other by scores on a standard scale adapted for this study. Feeling depressed, but not the number of items scored positive, was associated with myocardial infarction. Locus of control, or perceived ability to control the circumstances of life, was protective in the sense of a relative risk progressively less than 1.0 with increasingly supportive responses. Data were presented for each region of the world repre-
sented by the study populations and were, in the main, consistent across regions and ethnic groups for both men and women. The authors concluded that:102, p 953 “Presence of psychosocial stressors is associated with increased risk of acute myocardial infarction, suggesting that approaches aimed at modifying these factors should be developed.” In an exceptional editorial in Psychosomatic Medicine, Sheps and others (including Frasure-Smith, cited previously) referred to publication of this “landmark” report from the INTERHEART Study by The Lancet as “evidence that psychosocial factors are important to the mainstream medical community. . . . Biobehavioral factors play important roles in health and illness, and must therefore play an equally important role in medical research.”103, p 798 Prevention and Control Evidence was noted above that the Type A behavior score could be modified experimentally, with a statistically significant relative reduction in recurrence rates for coronary events in the intervention versus the control group.74 In a more general discussion of psychosocial factors in the practice of cardiology, Rozanski and others presented the rationale for concern about these factors and approaches to screening and management.60 Methods of intervention included exercise training, nutritional counseling, relaxation training, stress management, social support, and health information. Several trials were cited that evaluated cardiovascular outcomes of intervention among patients with existing coronary heart disease—the Recurrent Coronary Prevention Project Study, Ischemic Heart Disease Study, and Enhancing Recovery in Coronary Heart Disease Study (ENRICHD). Results were mixed. It was difficult in some studies to achieve target levels of psychological improvement, or differential improvement between intervention and control groups, sufficient to have the intended impact. Psychopharmacotherapy has also been investigated, such as use of selective serotonin uptake inhibitors (SSRIs), a class of antidepressant drugs, in eligible patients. Additional reviews are cited by Graves and Miller, who advocate larger and more diverse study populations in future trials to provide more influential evidence of efficacy of psychosocial interventions.104 Depression has been a particular focus of intervention studies, as reviewed by Lett and others and by Lichtman and others in a recent AHA Science Advisory.91,105 Lett and others found a paucity of evidence for efficacy of available antidepressant interventions, through the approaches indicated previously, in improving coronary outcomes. Lichtman and others
Odds Ratio (99% CI) 1 0.95 (0.84–1.08) 1.38 (1.19–1.6l) 2.14 (1.73–2.64) 1 1.05 (0.97–1.13) 1 52 (1.34–1.72) 2.12 (1.68–2.65) 1 1.05 (0.96–1.14) 1.45 (1.30–1.61) 2.17 (1.84–2.55) 1 119 (1.11–1.29) 1.33 (1.19–1.48) 1 1.23 (1.13–1.34) 1.48 (1.33–1.64) 1 0.89 (0.80–0.98) 0.72 (0.65–0.79) 0.68 (0.61–0.76) 1 1.55 (1.42–1.69) 1 1.50 (1.21–1.86) 1.65 (1.47–1.85) 1.44 (1.27–1.65)
Number of Controls (%) 1768 (23.9%) 3923 (53.1%) 1324 (17.9%) 372 (5.0%) 5343 (39.2%) 6873 (50.4%) 1179 (8.6%) 253 (1.9%) 3688 (27.0%) 7193 (52.7%) 2183 (16.0%) 584 (4.3%) 6628 (48.6%) 5361 (39.3%) 1659 (12.2%) 8528 (62.5%) 3349 (24.5%) 1771 (13.0%) 2619 (19.2%) 3265 (23.9%) 4839 (35.5%) 2925 (21.4%) 11244 (82.4%) 2404 (17.6%) 11244 (82.4%) 298 (2.2%) 1145 (8 4%) 961 (7.0%)
9% (7–10)
16% (10–22)
10% (8–13)
11% (7–14)
Source: Reprinted with permission from The Lancet, Vol 364, A Rosengren, S Hawken, S Ôunpuu, et al., for the INTERHEART Investigators, p 958.
12:31 PM
12% (7–17)
8% (4–12)
9% (1–18)
PAR (99% CI)
2/26/10
All associations are significant at p 0.0001, Q quartile, where Q1 is lowest locus. *Includes both working and non-working participants. †Felt sad, blue, or depressed for more than 2 consecutive weeks in past year.
Psychosocial Risk Factors in Cases and Controls Number of Cases (%) Stress at work (n 12,813) Never 1138 (21.0%) Some of the time 2499 (46.1%) Several periods 1249 (23.0%) Permanent 540 (10.0%) Stress at home (n 24,767) Never 4086 (36.8%) Some of the time 5361 (48.2%) Several periods 1288 (11.6%) Permanent 384 (3.5%) General stress* (n 24,767) Never 2777 (25.0%) Some period, home or work 5352 (48.1%) Several periods, home or work 2139 (19.2%) Permanent, home or work 851 (7.7%) Financial stress (n 24,767) Little or none 4872 (43.8%) Moderate 4625 (41.6%) Severe 1622 (14.6%) Stressful life events (n 24,767) None 6425 (57.8%) 1 2904 (26.1%) 2 or more 1790 (16.1%) Locus of control (n 24,767) Q1 2620 (23.6%) Q2 2938 (26.4%) Q3 3614 (32.5%) Q4 1947 (17.5%) Feeling depressed† (n 24,767) No 8446 (76.0%) Yes 2673 (24.0%) Depression (n 24,767) Not depressed 8446 (76.0%) 0–1 items 346 (3.1%) 2–4 items 1369 (12.3%) 5 or more items 958 (8.6%)
Table 15-10
46894_CH15_431_502.qxd Page 463
2. ADVERSE PSYCHOSOCIAL FACTORS 463
46894_CH15_431_502.qxd
464
2/26/10
12:31 PM
Page 464
CHAPTER 15 OTHER PERSONAL FACTORS
provided a detailed evaluation algorithm to identify and manage depression in patients with coronary heart disease. Separately, Lett and others found little evidence that improving low social support—often associated with depression—reduces coronary event rates. A Cochrane review (see Chapter 19, “Evidence and Decision Making”) was undertaken by Rees and others “to determine the effectiveness of psychological interventions, in particular stress management interventions, on mortality and morbidity, psychological measures, quality of life, and modifiable cardiac risk factors, in patients with coronary heart disease (CHD).”106, p 3 A total of 36 qualifying trials were identified, 18 of which addressed stress management. The authors concluded that, overall, the trials showed small effects of intervention on anxiety and depression and none on coronary outcomes. Poor quality, heterogeneity, and evidence of positive publication bias limited the utility of available studies. Current Issues Convergence Versus Divergence The upsurgence of interest in psychosocial factors and cardiovascular disease that characterized the 1960s and 1970s has been followed by a large number of studies on diverse yet often interrelated aspects of psychology, health behavior, and social interaction. It has been difficult to evaluate the status of this area of research, or even subtopics within it, in part because of the seeming elusiveness of standard concepts, definitions, classification, and methods of observation and measurement. The field has appeared to be continuing in a predominantly divergent phase, suggesting that greater value has been accorded to innovation than to integration. One area where convergence may be especially promising is the topic of depression, increasingly supported as a factor in incidence and prognosis of coronary heart disease. Another may be the potential for wider application of an approach like that of the INTERHEART Study assessment of psychosocial status. Methodologic research on the correspondence of findings between this and more extensive and longerestablished instruments could help to advance standardization in psychosocial research on cardiovascular diseases. A Public Health Perspective This section of the chapter has addressed mainly individual-level aspects of psychosocial factors, a somewhat artificial separation from the social aspects. Because the social context has special importance
from the perspective of epidemiology and public health, it remains instructive to read such thoughtful essays as those of Cassel and of others cited in Chapter 16, “Social and Physical Environment,” which follows, as well as work on social support.62,107,108 These and other theoretical writings continue to reinforce the relevance of psychosocial factors to health and particularly to the chronic diseases. Keeping the social aspect of psychosocial factors prominently in view may serve to broaden public health interest, support, and engagement in this arena.
3. HEMOSTATIC FACTORS Overview The term “hemostatic factors” encompasses the three phases of hemostasis, which is the internal regulatory process by which the flow of blood in a vessel can be arrested: the vascular phase, the platelet phase, and the coagulation phase.109 Hemostasis is highly regulated, involving a complex and delicate balance of biochemical and physiological interactions that maintain blood flow under normal conditions but have the capacity to interrupt blood loss rapidly, under conditions such as gross injury to a vessel. The relation of thrombosis (formation of a blood clot within a vessel) to the occlusive events of advanced atherosclerosis was first recognized in the early 1900s. The dynamics of the atherosclerotic plaque and the intricacies of the biochemical and hemostatic response to plaque disruption, noted in Chapter 3, “Atherosclerosis,” are recent advances. Thrombosis is a pathological occurrence, which could be understood as either a normal response to adverse stimuli arising from the underlying vascular pathology of atherosclerosis or a disorder of hemostasis itself that compounds that pathology. Of the three phases of hemostasis, the coagulation phase has received much more extensive epidemiologic investigation than have the vascular and platelet phases, a development within the past 30 years. Coagulation as used here is intended to encompass thrombogenesis (clot formation) and thrombolysis (clot dissolution). The following discussion focuses first on prothrombotic factors, those promoting coagulation, as studied in populations. Aspects of hemostasis most directly relevant to prevention and control, by contrast, relate either to the thrombolytic side of coagulation—promoting clot dissolution in the context of acute case management for occlusive coronary or cerebrovascular events—or to inhibiting platelet function to prevent thrombosis in approaches to preventing first or recurrent vascular occlusions.
46894_CH15_431_502.qxd
2/26/10
12:31 PM
Page 465
3. HEMOSTATIC FACTORS 465
Definitions and Measurement Coagulation Factors The coagulation factors have been designated variously over the many years of their study and description, as illustrated in Table 15-11.110 In some cases (e.g., factor VII), reference is commonly to the Roman numeral, whereas in others (e.g., fibrinogen), the functional descriptive term is used more often. Each factor can be characterized in some detail with respect to its molecular type and size, site of biosynthesis, biologically effective half-life in the circulation, and function, according to current understanding of the coagulation process. Although several of these factors and additional related enzymes or intermediate products have been addressed in some epidemiologic studies, the greatest attention to date has been directed to fibrinogen. Fibrinogen Fibrinogen is the precursor of fibrin in the coagulation process and thus provides the material that becomes the physical framework of a thrombus when the process is fully developed. Fibrinogen is synthesized in the liver and is mainly found circulating in the blood, its concentration in plasma ranging typically from 160 to 415 mg/dl. When clot lysis occurs, fibrinogen is enzymatically reduced to smaller units that can
Table 15-11 Roman Numeral I II III IV V VII VIII IX X XI XII XIII — —
be identified in plasma. High plasma fibrinogen concentration may result from active subclinical progression of atherosclerosis or other pathophysiological processes. Interpreting values of fibrinogen concentration in individual measurements requires caution because of its within-person variability. In addition, a high concentration of fibrinogen may occur in the acute course or immediate aftermath of myocardial infarction or stroke, as a consequence rather than a precursor of the event. Measurement Between-laboratory standardization is limited or lacking altogether for all of the coagulation factors, but results from even the best laboratory procedure may be invalidated by the influence of the blood-drawing technique because disturbances of hemostasis caused by venipuncture may activate some of the hemostatic factors.111 Full biological activity of some factors may not be preserved in frozen storage of samples, so specimens held long term may not provide valid results. Fibrinogen can be assayed by allowing its conversion to fibrin, which can then be determined quantitatively by any of several techniques that are potentially practical for use in most clinical laboratories. For several of the hemostatic factors, bioassays measure their relative functional activity, for which results are expressed as a percentage of normal.
Nomenclature and Synonyms for Coagulation Factors Preferred Descriptive Name Synonyms Fibrinogen Prothrombin Tissue factor Thromboplastin Calcium ions Proaccelerin Labile factor, accelerator globulin (AcG), thrombogen Proconvertin Stable factor, serum prothrombin conversion accelerator (SPCA) Antihemophilic factor (AHF) Antihemophilic globulin (AHG), antihemophilic factor A, platelet cofactor 1, thromboplastinogen Plasma thromboplastin Christmas factor, antihemophilic factor B, component (PTC) autoprothrombin II, platelet cofactor 2 Stuart factor Prower factor, autoprothrombin III, thrombokinase Plasma thromboplastin Antihemophilic factor C antecedent (PTA) Hageman factor Glass factor, contact factor Fibrin stabilizing factor (FSF) Laki-Lorand factor (LLF), fibrinase, plasma transglutaminase, fibrinoligase Prekallikrein Fletcher factor HMW kininogen High molecular weight kininogen, contact activation cofactor, Fitzgerald factor, Williams factor, Flaujeac factor, Reid factor, Washington factor
Source: Reprinted with permission from TC Bithell, Thrombosis and Antithrombotic Therapy, in Wintrobe’s Clinical Hematology, 9th ed, Vol 1, GR Lee et al., eds, p 567, © 1993, Lea & Febiger.
46894_CH15_431_502.qxd
466
2/26/10
12:31 PM
Page 466
CHAPTER 15 OTHER PERSONAL FACTORS
Measurement of hemostatic factors was investigated in the Atherosclerosis Risk in Communities (ARIC) Study.112 Contributions of within-person variability, processing and assay variability, and the ratio of between-person variability to total variance were reported in determinations of seven hemostatic factors—including fibrinogen, factors VII and VIII, and von Willebrand factor. Reliability coefficients were high for factor VIII and intermediate for the other three of these among the seven factors studied. Determinants and Mechanisms Theoretical Schemes of Hemostasis Concentrations of fibrinogen or other coagulation factors are presumed to influence the regulatory balance of hemostasis in the direction of their dominant mode of action. It is therefore necessary to consider the context in which these factors operate. Figure 15-9 indicates the relationships among coagulation, platelet function, and endothelial function.113 Above the horizontal lines are the prohemostatic components, in-
volving fibrin, proteolytic enzymes, and activated (a) factors V and VII. Independently, substances called thromboxanes and PAF (platelet-activating factor) are released by blood platelets and function in the same direction. All of these factors and their attendant processes are balanced by opposing influences of the vascular endothelium. Here, heparin and thrombomodulin contribute to the function of plasminogen activators and fibrinolysis, AT-III (antithrombin III), HCF-II (heparin cofactor II), and proteins C and S. Thus, the coagulation phase is only one aspect of a complex set of regulatory interactions. Loss of balance in these interactions may result in either failure or excess of hemostatic function. Any dysfunctional state or condition that fosters thrombogenesis or inhibits thrombolysis is of concern. Coagulation Pathways For appreciating the roles of various components investigated in population studies, it is helpful to examine their pathways of action.110 The coagulation
COAGULATION
Fibrin
PLATELETS
Proteolytic Enzymes
Factors Va and VIIIa
AT-III HCF-II
Proteins C and S
Thromboxanes PAF
Fibrinolysis
Plasminogen Activators
Heparin
Prostacyclin
Thrombomodulin
ENDOTHELIUM
Figure 15-9 Interactions of Prohemostatic and Antihemostatic Mechanisms. Thrombosis may result when an imbalance of these phenomena occurs. The interactions indicated mainly occur in the microvasculature, which, because of the relatively large endothelial surface in proportion to the small volume of the flow, may act to clear the circulation of prothrombotic debris. Source: Reprinted with permission from TC Bithell, Thrombosis and Antithrombotic Therapy, in Wintrobe’s Clinical Hematology, 9th ed, Vol 2, GR Lee et al., eds, p 1516, © 1993, Lea & Febiger.
46894_CH15_431_502.qxd
2/26/10
12:31 PM
Page 467
3. HEMOSTATIC FACTORS 467
pathway (Figure 15-10) comprises two distinct processes by which coagulation can be initiated. The “intrinsic pathway” involves one set of factors that act upon sequentially to convert inactive specific protein molecules into activated enzymes. This process begins with factor XII and proceeds through factor X to the ultimate production of fibrin. Separately, the “extrinsic pathway” originates with tissue factor and factor VII to activate factor X and proceeds thereafter as does the intrinsic pathway. The “common pathway” can be activated by either the intrinsic or the extrinsic pathway. The process of fibrinolysis, by which a clot is dissolved, includes several inhibitors that can limit the anticoagulation process (Figure 15-11). For example, the inhibitors of plasminogen activators, or plasminogen activator inhibitor (PAI), counter fibrinolysis. PAI-1, for example, has been investigated in several epidemiologic studies to determine whether relatively high concentrations are associated with arterial thrombosis or other manifestations of atherosclerosis. Thus, a high concentration of PAI-1 is interpreted as reflecting a high level of opposition to clot lysis or, in ef-
fect, of promotion of progressive thrombus formation. Once activated, plasminogen becomes plasmin, in either bound or free form. In “physiologic” proteolysis, bound plasmin acts to break down fibrin. In “pathologic” proteolysis, free plasmin breaks down fibrinogen and other proteins. Fibrinogen is not only a prothrombotic factor but also has direct proatherogenic effects in its role as a marker of atherosclerotic inflammation, a property shared with lipoprotein (a) (see Chapter 11, “Adverse Blood Lipid Profile”) (Figure 15-12).114 Together, these factors potentiate one another through multiple pathways of action. Pearson and others proposed a model to link hemostatic factors with cardiovascular disease that brings coagulation factors and fibinolytic factors together (Figure 15-13).115 “Thrombotic predisposition” is represented as a balance of these two groups of factors. As suggested by others, thrombotic predisposition leads both to thrombotic occlusion and, with other risk factors, to atherosclerosis and thereby risk of plaque fissuring, all with coronary, cerebrovascular, and peripheral vascular consequences.
Note: PF-3, platelet factor 3; Pre-K, prekallikrein; HMWK, high molecular weight kininogen.
Figure 15-10 Pathways of Coagulation. Source: Reprinted with permission from TC Bithell, Thrombosis and Antithrombotic Therapy, in Wintrobe’s Clinical Hematology, 9th ed, Vol 1, GR Lee et al., eds, p 580, © 1993, Lea & Febiger.
46894_CH15_431_502.qxd
468
2/26/10
12:31 PM
Page 468
CHAPTER 15 OTHER PERSONAL FACTORS
PLASMA PROACTIVATORS
ENDOTHELIUM OTHER TISSUES
1
2
Inhibitors
PLASMINOGEN ACTIVATORS
Inhibitors
Fibrin formation
3
BOUND PLASMINOGEN (Fibrin)
4
BOUND PLASMIN (Fibrin)
7
“PHYSIOLOGIC” PROTEOLYSIS (Fibrinolysis)
FREE PLASMIN
8
"PATHOLOGIC" PROTEOLYSIS
PLASMINOGEN FREE PLASMINOGEN (Plasma)
5
(Plasma) 6
Note: Solid arrow, transformation; dashed arrow, action.
(Fibringenolysis and degradation of other proteins)
ANTIPLASMINS (Plasma)
Figure 15-11 The Physiology of Fibrinolysis. The steps numbered within arrows are discussed in the text. “Fluid phase” activation of plasminogen is produced by the conversion of several plasma proenzymes (“proactivators”) into proteolytic enzymes that act as plasminogen activators (step 1). Activators derived from endothelium are probably more important physiologically (step 2). Source: Reprinted with permission from TC Bithell, Thrombosis and Antithrombotic Therapy, in Wintrobe’s Clinical Hematology, 9th ed, Vol 1, GR Lee et al., eds, p 593, © 1993, Lea & Febiger.
At the same time, however, Pearson and others pointed to several conceptual issues whose resolution is needed to establish confidence in the model they proposed. As shown in Figure 15-14, the relationships established to date could be interpreted in any of six ways: first, as a true causal pathway but, alternatively, reflecting confounding; an extended causal pathway; interaction with other factors; prevalenceincidence bias; or paradoxical reactions. Together with measurement issues regarding hemostatic factors, discussed previously, these considerations suggest caution in ascribing causality to the associations so far demonstrated. Koenig noted that as an acute phase reactant, fibrinogen can be increased in concentration secondary to action of all cells involved in the atherosclerotic process because of their cytokine production.116 He identified six properties through which fibrinogen could contribute to cardiovascular risk: substrate for
thrombin formation and the last step in coagulation; an essential role in platelet aggregation; modulator of endothelial function; promoter of smooth muscle cell proliferation and migration; mediator of plasmin and receptor binding; and major acute phase protein. Nonetheless, “Whether or not fibrinogen is causally involved in atherothrombogenesis still remains to be determined . . .”116, p 601 Similarly, Shah expressed the view that:117, p 15 . . . a circulating thrombogenic state, as measured by plasma factors and their genetic variants, has not been consistently shown to be strongly linked to atherothrombotic events, with the possible exception of plasma fibrinogen levels. . . . The predominant factors contributing to arterial thrombosis in atherosclerosis have to do with disruption of the plaque itself (plaque rupture and superficial endothelial erosions).
46894_CH15_431_502.qxd
2/26/10
12:31 PM
Page 469
3. HEMOSTATIC FACTORS 469
FIBRINOGEN ⫹ Adhesion molecules
⫹ Blood viscosity
Fibrin formation
⫹
⫹
Platelet aggregation (Gp IIb/IIIa) ⫹
Endothelial dysfunction (↑perm.,↓NO) ⫹
⫹ Plasmin
⫺
PAI-1
⫹
TFP ⫹ inhibitor ⫺
Foam cells
Plasminogen activators Plasminogen ⫺ Fibrin degradation products Prothrombotic effect
⫹
Binding to LDL
⫹
⫹
VSMC proliferation ⫹
LIPOPROTEIN (a)
Proatherogenic effect
Figure 15-12 Flow Diagram Showing the Pathogenetic Inter-Relations Between Fibrinogen and Lipoprotein (a) [Lp(a)]. Both of these molecules possess prothrombotic and proatherogenic properties. Fibrinogen is a major determinant of blood viscosity, promotes platelet aggregation by binding to glycoprotein IIb/IIIa receptors, and is the biochemical substrate for fibrin formation under the action of thrombin. Lipoprotein(a) binds and inactivates the tissue factor pathway (TFP) inhibitor, a potent inhibitor of the tissue factor-mediated coagulation cascade that is produced by endothelial cells, platelets and monocytes, leading to increased fibrin formation. Lipoprotein(a) has also important inhibitory effects on fibrinolysis both by inhibition of plasmin formation due to its significant structural homology with plasminogen and enhanced expression of plasminogen activator inhibitor-1 (PAI-1). Both fibrinogen and lipoprotein(a) have been shown to activate expression of adhesion molecules, induce endothelial dysfunction by increasing permeability and decreasing nitric oxide (NO) production, respectively, and activate the migration and proliferation of vascular smooth muscle cells (VSMC). Moreover, both molecules can be found in the wall of atherosclerotic vessels where fibrinogen binds to lipoproteins and Lp(a) is internalized by vascular macrophages where it accumulates. LDL, Low-density lipoprotein.Source: Reprinted with permission from Journal of Hypertension, Vol 23, C Catena, M Novello, R Lapenna et al., p 1620, © 2005, Lippincott Williams and Wilkins.
Coagulation Factors (⫹)
Fibrinolytic Factors (⫺)
Thrombotic Predisposition Thrombotic Occlusion Plaque Fissuring
Atherogenic Risk Factors
Myocardial Infarction Unstable Angina Atherothrombotic Stroke Peripheral Occlusive Disease
Atherosclerosis
Note that coagulation factors and fibrinolytic factors can act either by promoting atherosclerosis or by causing thrombotic occlusion on a plaque fissure.
Figure 15-13 A Model for the Causal Association of Hemostatic Factors and Cardiovascular Disease. Note that coagulation factors and fibrinolytic factors can act either by promoting atherosclerosis or by causing thrombotic occlusion on a plaque fissure. Source: Reprinted with permission from American Journal of Clinical Nutrition, Vol 65(Supp1), TA Pearson, J LaCava, HFC Weil, p 1675S, © 1997 American Society for Clinical Nutrition.
46894_CH15_431_502.qxd
470
2/26/10
12:31 PM
Page 470
CHAPTER 15 OTHER PERSONAL FACTORS
1)
Causal Pathway ↑
2)
CVD
Hemostatic Factor
Confounding Other Factors
CVD
Hemostatic Factors 3)
Extended Causal Pathway Other Risk Factors → Hemostatic Factor → CVD
4)
Interaction with Other Factors Other Risk Factors Hemostatic Factors → CVD
5)
Prevalence Incidence Bias
6)
CVD
↑ Hemostatic Factor
CVD
↓ Hemostatic Factor
Paradoxical Reactions CVD → ↑ Coagulation Factors
→ ↑ Fibrinolytic Factors
Figure 15-14 Six Methodologic Issues in the Study of Hemostatic Factors and Cardiovascular Disease (CVD). In addition to a causal relation (1), hemostatic factors can be associated with disease via confounding (2), an extended causal pathway (3), an interaction with other risk factors (4), prevalence-incidence bias (5), or paradoxical reactions (6). Source: Reprinted with permission from American Journal of Clinical Nutrition, Vol 65(Supp1), TA Pearson, J LaCava, HFC Weil, p 1675S, © 1997 American Society for Clinical Nutrition.
Because of multiple redundancies in coagulation mechanisms, some processes that would appear to be rate-limiting may be dominated by others. Further, a high concentration of one component may downregulate production or activation of other components, with a net effect opposite to its direct mode of action. For these reasons it is difficult to select with confidence among factors available for study. Timing of their measurement, in relation to the course of the atherosclerotic process or to other conditions that may distort the picture, is an important consideration in study design. Interpretation of observations on one or another factor in isolation from others that are unmeasured may be erroneous. Further, intervention to alter the balance of the hemostatic process raises concerns about safety, given the importance of normal physiologic function of blood coagulation. Associations with Established Risk Factors At another level of consideration, epidemiologic studies have shown fibrinogen concentrations to be associated with many other factors related to atherosclerosis and its manifestations. A summary of the findings by Folsom is presented in Table 15-12.118 The factors studied included age, sex, and race; diet and physical activity or fitness; obesity; blood lipids; blood pressure; diabetes, serum insulin, and microalbuminuria; smoking; social class; seasonal varia-
tion; and several others. The direction and relative strength of association with each factor are indicated, from strongly or moderately positive to moderately or strongly negative, with two factors lacking association also included. Because of the nonspecific pattern of association with these diverse factors, it is unclear whether fibrinogen has an independent causal role. A related question is which, if any, of these associations indicate pathways through which fibrinogen may be influenced. This direction of inquiry may show how recognized risk factors promote atherogenesis or precipitate acute thrombotic events. Similar questions apply to other associated hemostatic factors. Interest in the relation between blood lipids or dietary fats and hemostatic factors has also received attention. One report suggested that one mechanism by which lipid-lowering therapy with statins reduces cardiovascular risk is through reduction in fibrinogen, PAI-1, and activated factor VII levels.119 The same report proposed prothrombotic effects of LDLcholesterol and antithrombotic effects of HDLcholesterol. Another report referred to “accumulating evidence” of a relationship between dietary fatty acids and “emerging hemostatic CVD risk factors, although much of this evidence is incomplete or conflicting.”120, p 410 Suggested mechanisms were increased bleeding time and reduced risk of thrombosis with marine n-3 fatty acid intake; reduction in factor VII coagulant activity with reduced saturated fatty acid
46894_CH15_431_502.qxd
2/26/10
12:31 PM
Page 471
3. HEMOSTATIC FACTORS 471
Table 15-12
Age
Factors Associated with Plasma Fibrinogen Concentration Positive Association No Association Strong Modest Winter season Dietary fat
Female gender Cigarette smoking Inflammatory conditions Stress Obesity Diabetes Menopause Oral contraceptives Leukocyte count
African-American ethnicity Blood pressure
Nicotine gum
LDL cholesterol Triglycerides Lipoprotein(a) Homocysteine Serum insulin Microalbuminuria
Negative Association Modest Strong Alcohol intake Social class Education level Physical activity, Japanese ethnicity fitness Fish oil Estrogen replacement HDL cholesterol Birth weight
Note: LDL, low density lipoprotein; HDL, high density lipoprotein. Source: Reprinted from European Heart Journal, Vol 6A, AR Folsom, Epidemiology of Fibrinogen, p 22, © 1995, by permission of the publisher WB Saunders Company Limited London.
intake; and an adverse effect of postprandial triglyceridemia on activation of factor VII. A link between hemostatic and social factors was observed in a cross-sectional analysis of fibrinogen concentration and other risk factors in the Kuopio Ischemic Heart Disease Risk Factor Study.121 Both age-adjusted and covariate-adjusted mean values of fibrinogen concentration were presented for each stratum of five socioeconomic indices. In age-adjusted analyses, significantly higher fibrinogen concentrations were observed in the following comparisons: for both blue-collar workers and farmers, in contrast to white-collar workers; for each of the lowest four strata versus the highest stratum according to income; for those with less than high school education versus high school or more; and for those with the lowest score for material possessions. In some cases, covariate adjustment rendered the results nonsignificant. Genetic Factors Studies of genetic variation in fibrinogen and factor VII loci indicate that the manner in which specific alleles control concentrations of hemostatic factors depends on additional factors, such as smoking status or triglyceride concentrations. These are potentially important demonstrations of gene–environment interactions. An implication of such findings is that comparison of risks of atherosclerosis and its complications between some populations or subgroups may be misleading unless both the genotypes and the other relevant personal or environmental characteristics are taken into account. Much remains to be learned in order to identify pertinent genetic variants
and determine the population distributions of these genotypes to their possible contribution to population differences in atherosclerosis. This aspect of the fibrinogen association with coronary heart disease was investigated by Keavney and others, who conducted a genetic analysis in a large case-control study, the International Studies of Infarct Survival (ISIS) Study, and a pool of several additional studies.122 Polymorphism in the -fibrinogen promoter gene was described in cases and controls with and without coronary heart disease. Associations were found between genotype and fibrinogen levels and between fibrinogen levels and coronary heart disease status, but the latter was greatly attenuated and became nonsignificant after adjustment for multiple other risk factors. The authors concluded that:122, p 935 Genotypes that produce lifelong differences in fibrinogen concentrations do not materially influence coronary disease incidence. As these genotype-dependent differences in fibrinogen were allocated randomly at conception (Mendelian randomization), this association is not likely to be confounded by other factors. Consequently, these genetic results provide strong evidence that long-term differences in fibrinogen concentrations are not a major determinant of coronary disease risk. Distribution General Frequency Distributions Substantial contributions to the epidemiology of hemostatic factors have come from the Northwick Park
46894_CH15_431_502.qxd
472
2/26/10
12:31 PM
Page 472
CHAPTER 15 OTHER PERSONAL FACTORS
(England) Heart Study (NPHS), in which several factors have been investigated.123 Population distributions for each of eight hemostatic factors and platelet characteristics were reported in some 2000 men and women, aged from 18 years to the late 50s or mid-60s. Relations of these characteristics with several other cardiovascular risk factors were also addressed. Population Differences: Japan and the United States Population differences in hemostatic variables were compared among small samples of Japanese men in rural and urban Japan and the United States and White men in the United States. The aim was to test the hypothesis that these factors would indicate lower risk of coronary heart disease for Japanese men in Japan than the other men, consistent with previous knowledge of marked differences in coronary heart disease rates among these populations.124 Fibrinogen concentration was highest in Whites in contrast to the other three groups (290 mg/dl versus a range from 223 to 250 mg/dl). Factors VII and VIII, measured functionally as coagulation activities, were greater for US Japanese and Whites than for Japanese in Japan (only factor VII was significantly higher). Von Willebrand factor did not differ among the four groups. In all samples, in addition, fibrinogen concentration was significantly greater in smokers than in nonsmokers. Variation at Menopause and with Estrogen Replacement A study of hemostatic factors in 207 eligible US women around the age of menopause (mean age 52 years) indicated a generally higher prohemostatic profile, or thrombogenic predisposition, after menopause than before. This difference was attenuated among women using hormone replacement therapy after menopause (predominantly conjugated estrogen plus progesterone).125 However, whereas hormone replacement appeared to avert increase in fibrinogen concentrations, factor VII activity was increased more among women with hormone replacement than without. This seemingly discrepant finding could be explained by the effect of hormone replacement therapy to increase triglyceride concentration, which in turn causes an increase in concentration of factor VII.126 Correlates of Impaired Fibrinolysis The converse of exaggerated coagulation is impairment of fibrinolysis. This type of hemostatic imbalance predisposes to thrombogenesis and to progression from plaque disruption to arterial occlusion. This tendency results from increased fibrinogen and factor VII levels and decreased fibrinolysis, which occur in the
presence of obesity, hyperlipidemia, diabetes, smoking, and emotional stress.127 Effects on Rates and Risks The Northwick Park Heart Study Several cohort studies of hemostatic factors and coronary heart disease have been conducted, perhaps earliest among them the NPHS cited previously. By 1986, that group found that both factor VII coagulant activity and fibrinogen concentration were important predictors of coronary events over 5 to 10 years of follow-up among White men age 40–64 years at entry.128 Factor VII, fibrinogen, cholesterol concentration, and systolic blood pressure were evaluated for possible independent associations with fatal, nonfatal, and total coronary events. The strongest predictive relations were found for fatal events within 5 years. On the basis of standard deviation units for each variable, each unit increase in age, factor VII, and fibrinogen significantly increased the risk by 92%, 55%, and 67%, respectively. Fibrinogen concentration appeared to have an intermediate role between smoking and risk of coronary heart disease. Overall, Meade and colleagues concluded:128, p 537 “There is increasing reason to consider the prevention of thrombosis as an effective approach to the prevention of IHD [ischemic heart disease]. The case for doing so is strengthened by the possibility that the biochemical disturbance in IHD may lie at least as much in the coagulation system as in the metabolism of cholesterol.” The ARIC Study The ARIC Study also found significantly increased risks of incident coronary heart disease events in relation to four hemostatic factors (fibrinogen, factor VIII, von Willebrand factor, and white blood cell count) among men and women followed for 5.2 years.129 Fibrinogen concentration and white blood cell count were the most consistent predictors of coronary events among non-Black men and both Black and non-Black women, in both younger ( 55 years) and older ( 55 years) men and older women. Factor VIII was associated with increased risk in Black women alone. Multivariate adjustment for other risk factors reduced the relative risks for coronary events to nonsignificant levels except for fibrinogen in both men and women and white blood cell count in women alone. Notably, all four hemostatic factors were significant predictors of all-cause mortality. The ARIC investigators concluded:129, p 1107 From a preventive medicine point of view, only measurement of fibrinogen (and not the other hemostatic factors) contributed anything beyond
46894_CH15_431_502.qxd
2/26/10
12:31 PM
Page 473
3. HEMOSTATIC FACTORS 473
traditional risk factors in the prediction of CHD. A fibrinogen measurement costs approximately the same as a lipid profile, so it could be considered for risk factor screening. However, there is no universal standardization system for the fibrinogen assay, and the independent contribution of fibrinogen to prediction of risk appears to be modest. There also has been no clinical trial yet to demonstrate that lowering fibrinogen will prevent CHD. These facts suggest that routine screening for elevated fibrinogen in healthy adults is currently not warranted.
1.2–2.5) to 4.1 (2.3–6.9), with an overall summary odds ratio of 2.3 (1.9–2.8). Many factors influence fibrinogen concentrations, as discussed previously. Nonetheless, the authors described fibrinogen as a major cardiovascular risk factor that should be included in future studies of risk factor modification. A subsequent meta-analysis of cohort studies on fibrinogen and other inflammatory markers and coronary heart disease included 12 population-based studies and six studies among persons with existing cardiovascular disease.131 Figure 15-15 presents the relative risks of coronary heart disease (not further defined), based on comparison of the top versus bottom third of the distribution of fibrinogen in each study. The overall risk ratio, 1.8 (1.6–2.0), differed little between the two groups of studies. Studies were further described regarding the extent of adjustment for other factors in the reported results, from () representing adjustment for age and sex only to () for these plus smoking, other standard risk factors, social class, and other chronic disease at baseline. The
Meta-Analyses of Cohort Studies In addition to the NPHS, five other studies were included in a meta-analysis of cohort studies of fibrinogen and coronary heart disease reported in 1993.130 The odds ratios for the highest versus lowest onethird of the distribution of fibrinogen concentration in individual studies ranged from 1.8 (95% CI
Type of Cohort No. of and Source Cases Population Based 422 Rumley et al, 19973 571 Sweetnam et al, 19964 5 348 Folsom et al, 1997 270 Kannel et al, 1987,6 19927 Wilhelmsen et al, 1984,8 19929 216 Meade et al, 199310 183 Junker et al, 199711 130 107 Cremer et al, 199112 Feskens and Kromhout, 199713 41 40 Stone and Thorp, 198514 581 Tunstall-Pedoe et al, 199715 16 Lowe et al, 1997 235 3144 Subtotal Preexisting Vascular Disease Benderly et al, 199617 Toss et al, 199718 Thompson et al, 199519 Haines et al, 198320 Martin et al, 199121 Kostis et al, 198222 Subtotal Total Coronary Heart Disease
Degree of Adjustment
Risk Ratio and Confidence Intervals (Top Third vs Bottom Third)
⫹⫹⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹ ⫹
Confidence Intervals 99%
421 138 106 63 126 20
⫹⫹⫹⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹ ⫹ NR
95%
874 1.8 (95% CI, 1.6–2.0)
4018 0.5
1
2
4
8
Figure 15-15 Prospective Studies of Fibrinogen and Coronary Heart Disease. Risk ratios compare top and bottom thirds of baseline measurements. For all figures, black squares indicate the risk ratio in each study, with the square size proportional to the number of cases and the horizontal lines representing the 99% confidence intervals (CI). The combined risk ratio and its 95% CI are indicated by unshaded diamonds for subtotals and by shaded diamonds for grand totals. NR indicates not reported; +, adjustment for age and sex only; , for these plus smoking; , for these plus some other standard vascular risk factors; , for these plus markers of social class; and , for these plus information on chronic disease at baseline. Source: Reprinted with permission from JAMA, Vol 279, J Danesh, R Collins, P Appleby, R Peto, Association of Fibrinogen, C-reactive Protein, Albumin, or Leukocyte Count with Coronary Heart Disease. Meta-Analysis of Prospective Studies, p 1479, © 1998, American Medical Association.
46894_CH15_431_502.qxd
474
2/26/10
12:31 PM
Page 474
CHAPTER 15 OTHER PERSONAL FACTORS
latter two levels of adjustment were found with the lowest risk ratios and 99% confidence bounds below 1. Quite similar and generally consistent results were found for serum albumin and C-reactive protein concentration and white blood cell count. The authors concluded that despite lack of clear mechanisms underlying these associations, their strength and consistency warrant continued investigation. Other Factors and Outcomes Studies of fibrin breakdown products and both coronary and peripheral arterial disease suggest that the level of fibrin turnover has prognostic significance for progression of disease. Endogenous tissue-type plasminogen activator (tPA) has also been studied in connection with both myocardial infarction and stroke. For myocardial infarction, for example, baseline concentrations of tPA antigen were greater in 231 men free of evident coronary heart disease at entry who developed myocardial infarction during follow-up than in 231 control subjects, drawn from the (United States) Physicians’ Health Study.132 This association is interpreted to indicate that high concentrations of tPA reflect elevated PAI-1 concentrations and impaired fibrinolysis. However, after analysis controlling for other risk factors, mainly high-density lipoprotein (HDL) cholesterol, the relation of tPA was no longer significant. Prevention and Control Attention to hemostatic dysfunction in connection with prevention of cardiovascular events or their complications is limited to individual-level measures and focuses mainly on antiplatelet and thrombolytic therapy. In the first area, antiplatelet therapy has been evaluated for both its role in acute myocardial infarction and stroke and its potential value in primary and secondary prevention. The trials addressing these issues were summarized in the mid-1990s by the Antiplatelet Trialists’ Collaboration, an international cooperative group.133 Altogether 174 trials were reviewed, comprising some 70,000 high-risk and 30,000 low-risk participants, the latter in primary prevention trials, and an additional 10,000 participants in trials in which multiple treatments were compared. The intervention most commonly tested was aspirin, in doses ranging from 75 to 325 mg/day. The clinical status of eligible participant groups was: prior, but not acute, myocardial infarction; suspected or definite acute myocardial infarction; prior stroke or transient ischemic attack (TIA); and low-risk individuals with none of the preceding conditions. Endpoints for outcome evaluation were nonfatal myocardial infarction or reinfarction, non-
fatal stroke or stroke recurrence, vascular death, and any death. Benefits of antiplatelet therapy were statistically significant for all outcomes in all high-risk groups and for nonfatal myocardial infarction but not the other outcomes in the low-risk, primary prevention stratum. Reduction in vascular events in the highrisk groups was approximately 25 percent and was reported to be statistically significant in men and women, middle and older age, hypertensive and normotensive patients, and patients with or without diabetes. Treatment in low-risk persons was accompanied by an apparent increase in risk of stroke, which was not statistically significant, and that the benefit was only about four events prevented per 1000 persons treated for 5 years. The authors concluded that benefit was clear for high-risk patient groups, that wider use of antiplatelet therapy was warranted, but that use for primary prevention could not be recommended in view of current uncertainty about risks and benefits. Goff and others subsequently demonstrated that use of thrombolytic therapy in acute coronary events greatly increased the proportion of cases surviving for up to 56 months from the event.134 However, only one in four cases even among those arriving at a hospital in 2–4 hours or less received it, and for the majority of cases delay time was not indicated in the hospital record. For its efficacy in acute and long-term prevention of multiple cardiovascular outcomes, use of acetylsalicylic acid, or aspirin, and other antiplatelet therapy has been investigated. Trials in high-risk patients, with acute or previous vascular disease or other predisposing condition, were the subject of a metaanalysis by the Antithrombotic Trialists’ Collaboration reported in 2002.135 Aspirin at daily doses from 75–150 mg, as well as other agents, was found to be protective against occlusive cardiovascular events at rates that outweighed the absolute risks of major extracranial bleeding. Primary prevention, in persons free of prior cardiovascular disease, would probably present a different balance of benefits and risks. Based on evidence available in 2002, the US Preventive Services Task Force recommended “strongly” that “clinicians discuss aspirin chemoprevention with adults who are at increased risk for coronary heart disease (CHD).”136, p 59 The importance of discussion of bleeding risks was emphasized, and it was noted that evidence for benefits and harms might be less reliable for women than for men. Uncontrolled hypertension was considered to attenuate the benefits and increase the risks of aspirin use for CHD prevention.
46894_CH15_431_502.qxd
2/26/10
12:31 PM
Page 475
3. HEMOSTATIC FACTORS 475
The report of the Women’s Health Initiative trial of low-dose aspirin for prevention of cardiovascular disease in women, in 2005, led to several independent analyses of the total of six primary prevention trials available by that time.137–139 The most extensive of these new analyses was that of the Antithrombotic Trialists’ (ATT) Collaboration, which unlike the others obtained individual-level data from all six of the primary prevention trials and included a corresponding analysis for 16 secondary prevention trials. Analysis confirmed the findings of previous metaanalyses with about a 20% reduction in both total stroke and coronary event rates with aspirin. However, in primary prevention, whereas a corresponding reduction in nonfatal coronary events was observed, hemorrhagic stroke was increased and the net stroke effect was nonsignificant, total vascular mortality did not differ between aspirin and control, and major gastrointestinal and extracranial bleeds increased with aspirin. It was noted that on the basis of this analysis, proportionate reductions in all serious vascular events were similar for men and women. Importantly, the authors concluded:139, p 1849 “In primary prevention without previous disease, aspirin is of uncertain net value as the reduction in occlusive events needs to be weighed against any increase in major bleeds. Further trials are in progress.” Other aspects of anticoagulant therapy in highrisk patients include use of “dual antiplatelet therapy” combining aspirin with agents having other mechanisms of action, as in patients with coronary artery stents.140 Long-term prophylaxis against thromboembolic events in atrial fibrillation and prevention of deep vein thrombosis and pulmonary embolism are other situations in which chronic anticoagulation may be indicated (see Chapter 6, “Related Conditions”). Thrombolytic Therapy The second broad area of intervention is thrombolytic therapy, which has been evaluated in the acute stage of myocardial infarction and stroke.141 Among 58,600 patients with acute myocardial infarction studied in large trials, mortality in 35 days from onset was reduced by 18 percent among those treated (from 11.4% in controls to 9.6% with treatment). Of particular practical importance, treatment effectiveness declined sharply beyond 1–3 hours, requiring treatment to be administered very early in the episode. This increases the importance of knowledge of signs and symptoms of an impending cardiac event and use of emergency communications (calling 911) and medical services (EMS) including transport. The implications for emergency room practice include the
need for rapidly determining whether therapy is indicated. High cost of this form of treatment is a further consideration. This avenue of prevention is clearly available only to those who survive long enough to reach medical care. In a major multicenter trial of rtPA therapy in acute stroke, treatment was shown to be effective in treatment of ischemic or occlusive stroke, if received within 3 hours of symptom onset.142 As for coronary events, rapid recognition of impending stroke and timely emergency response are critical for effective use of this intervention. Further complicating its use in stroke is the contraindication to thrombolytic therapy in the presence of hemorrhagic stroke. Therefore, differentiation between major stroke types, usually depending on noninvasive imaging procedures, becomes an urgent necessity. Current Issues Further Investigation Studies of hemostatic factors, even fibrinogen, are limited by methodologic difficulties and uncertain interpretation. At the same time, the opportunity to deepen understanding of the “thrombotic side” of the atherothrombotic process is important, at least for its potential for prevention in high-risk persons close to the time of clinical events. Whether long-term preventive strategies will also be found practical and whether population interventions may be shown to be beneficial in bringing about favorable change in the delicate balance of hemostatic factors will be important revelations. Application of Current Knowledge With respect to intervention, the argument that antiplatelet therapy is underutilized is a clear message from the Antiplatelet Trialists and appears well supported.133 However, the reservations reported on the basis of the most thorough analysis to date of available evidence on primary prevention with aspirin warrant caution. Recommendations regarding use of this over-the-counter medication require qualification regarding its demonstrated hazard of serious gastrointestinal and other extracranial bleeding, as well as increased risk of hemorrhagic stroke. The present use of thrombolytic therapy is also lacking relative to its potential for improving shortand long-term survival in myocardial infarction. Greater application of the demonstrated benefit of this form of therapy may require substantial reduction in the delay between onset of symptoms and presentation for treatment in the acute phase of myocardial infarction or stroke.
46894_CH15_431_502.qxd
476
2/26/10
12:31 PM
Page 476
CHAPTER 15 OTHER PERSONAL FACTORS
4. EVOLVING AND EMERGING FACTORS Overview Additional factors warrant brief attention because of recent and continuing epidemiologic interest. These include combination pharmacotherapy, hormone replacement therapy (HRT), infection, antiretroviral therapy in HIV/AIDS, pro- and antioxidants, hyperhomocysteinemia, and inflammation and C-reactive protein. Recent developments are emphasized. The topic of “novel” markers of cardiovascular risk and their potential inclusion in risk prediction models for clinical management closes this review of other personal factors. Combination Pharmacotherapy The concept that multiple risk factors could be controlled through low-dose combination pharmacotherapy was presented at least as early as 1998, in an Institute of Medicine report on cardiovascular disease control in developing countries.143 The term “essential vascular package” or “EVP” was used to denote the combination of such agents as aspirin, beta-blockers, angiotensin-converting enzyme inhibitors, and statins in a single once-a-day formulation. The preparation would need to use generic drugs to be available at low cost, universally, for all persons with existing cardiovascular disease. Trials would be needed to evaluate its acceptability and effectiveness. Subsequently, Wald and Law proposed a patented formulation of this kind (the “Polypill”) with the recommendation that it be provided to “everyone aged 55 and older and everyone with existing cardiovascular disease” as a “strategy to reduce cardiovascular disease by more than 80%.”144, p 1419 Changes in diet and lifestyle were dismissed as not practicable in the short term; no reference was made to tobacco cessation or prevention. Following this report, extensive comment was published by the British Medical Journal and through ProCor’s Global Dialogue at http://www.procor.org/. The concept was reviewed in a workshop convened by the Centers for Disease Control and Prevention and published in 2005.145 Use of such a preparation in both primary and secondary prevention was considered. Randomized trials were suggested to address concerns about side effects and poor adherence as well as cost-effectiveness. Potential applicability for low-income groups in the United States and in the developing world was discussed. The Indian Polycap Study (TIPS) has now reported on a 12-week factorial trial in persons aged 45 to 80 years and having one cardiovascular risk factor who
were randomly allocated to one of nine groups.146 Each group received one or a combination of several agents, which were combined in a single preparation (“Polycap”) for one of the groups. Favorable effects were found regarding reductions in LDL cholesterol, diastolic blood pressure, serum homocysteine, and platelet function. There appeared to be no interaction among drug components to diminish their individual efficacy. Toleration as judged by continuation of treatment for the required 12 weeks—84%—was considered no different for the Polycap formulation than the others. The authors concluded:146, p 10 “Large studies assessing prevention strategies with adequate numbers of individuals from each of the major regions of the world are needed.” Hormone Replacement Therapy (HRT) Sex differences are a constant feature of the epidemiology of coronary heart disease mortality across many populations, with age-adjusted rates being consistently higher for men (or lower for women). In a high-rate country such as the United States, the absolute differences are large whereas in a low-rate country such as Japan, they are small. Several known effects of estrogen suggest that women—until some years after menopause—are protected from coronary heart disease via these mechanisms. A model for these relations between sex, rates, and differential mechanisms of occurrence of CHD, in different environments, was proposed by Khaw and Barrett-Connor as shown in Figure 15-16.147 Estrogen, in women, being unopposed by androgens, serves to increase HDLcholesterol concentration and leads to a lesser rate of CHD than in men. Meta-analysis of 11 case-control and 10 cohort studies of HRT indicate the main results of observational studies suggesting protection of women against coronary heart disease incidence and mortality with current HRT (Table 15-13).148 Risks were increased for thromboembolic stroke and thromboembolism experienced as deep vein thrombosis or pulmonary embolism. Against this background, the Women’s Health Initiative (WHI) trial of combined estrogen plus progestin among more than 16,000 healthy postmenopausal women aged 50–79 years was expected to show benefit in reduction of nonfatal myocardial infarction and CHD death. Results of the trial were contrary: Overall health risks exceeded benefits, and it was concluded that this regimen “should not be initiated or continued for primary prevention of CHD.”149, p 321 Beyond this immediate impact on treatment recommendations, the results of the WHI stimulated scrutiny of prior findings in observational studies to de-
46894_CH15_431_502.qxd
2/26/10
12:31 PM
Page 477
4. EVOLVING AND EMERGING FACTORS 477
1. Low incidence population (e.g. Japan)
Men Low fat diet
Women Low fat diet
¯ LDL-C
¯ LDL-C ¯ HDL-C ¯ ¯ CHD rate
¯ Oestrogen
¯ HDL-C ¯ ¯ CHD rate
¯ Oestrogen
2. High incidence population (e.g. USA)
Men High fat diet
- LDL-C = HDL-C
Women High fat diet
- Oestrogen
- LDL-C - HDL-C
- Oestrogen
Blocked by androgens - - CHD rate
- CHD rate
LDL-C, LDL-cholesterol; HDL-C, HDL cholesterol.
Figure 15-16 A Proposed Model for Sex Differences in CHD. Source: Reprinted with permission from Coronary Heart Disease Epidemiology: From Aetiology to Public Health. M Marmot, P Elliott eds. K-T Khaw, E B-C, Sex Differences, Hormones and Coronary Heart Disease, Chapter 19, p 281. © Michael Marmot, Paul Elliott, and Contributors. Oxford University Press, New York.
termine why these were not replicated in the trial, as in Table 15-14.148 Benefits were less clear in WHI than in those studies, because reduced risks were associated with wide confidence intervals and were offset by harms related to coronary heart disease events and thromboembolic events at levels that were unequivocal. Reevaluation of observational data suggested that healthier women self-selected for use of HRT. Discussion continued as to the conditions under which HRT was used in WHI in contrast to the observational studies, including the ages when women were first exposed, timing of their first exposure in relation to CHD occurrence, and other issues.148,150–154 It was also suggested that the WHI might itself suffer from biases that could partially explain the discrepancy with observational studies.155 Whether women in their 50s treated with HRT might show coronary benefits was further studied by examination of coronary artery calcium in that age stratum of the WHI, and a differential in score favoring HRT users was suggested.153 However, because similar assessment was not done in older WHI participants, this shed no light on the issue of differential benefit of HRT by age. This experience has, nonetheless, un-
derscored the importance of controlled trials for evaluating pharmacological interventions. As of 2006, the US Preventive Services Task Force recommended against routine use of combined estrogen and progestin for prevention of chronic conditions in postmenopausal women.156 Infection Infection has long been considered a potential explanation of the origin of atherosclerotic lesions, as noted briefly in Chapter 3, and interest in this area continues. A hypothetical model suggests mechanisms by which infection could influence the atherosclerotic process, leading to lesion rupture and thrombosis (Figure 15-17).157 Agents suggested most often or most recently have included especially Chlamydia pneumoniae, Helicobacter pylori, and cytomegaloviruses (CMV) (Table 15-15).158 It is noteworthy that laboratory evidence regarding this theoretical causal pathway has not generally been supported in epidemiologic studies. Cross-sectional surveys provide some consistent evidence, but cohort studies lend little support and trials virtually none. One meta-analysis of recent trials of antibiotic therapy against Chlamydia pneumoniae in
46894_CH15_431_502.qxd
478
2/26/10
12:31 PM
Page 478
CHAPTER 15 OTHER PERSONAL FACTORS
Table 15-13
Hormone Replacement Therapy (HRT) and Cardiovascular Disease and Thromboembolism Studies Relative Risk (95% Outcome by HRT Use Confidence Interval)* Type of Study† Quality‡ Coronary heart disease Incidence Current19,30,32–35,37,47,49 0.80 (0.68–0.95) RCT, cohort Fair-good Current adjusted for socioeconomic status29,34,35,47 0.97 (0.82–1.16) Ever28,29,34,43 0.91 (0.67–1.33) Cohort, case control Fair-good Past19,30,32,34–36,47 0.89 (0.75–1.05) Cohort, case control Fair-good All groups19,28–30,32,37,43,47,49 0.88 (0.64–1.21) RCT, cohort Fair-good Mortality Current38,44,46,47 0.62 (0.40–0.90) Cohort, case control Fair-good Ever45 0.81 (0.37–1.60) Cohort Fair-good Past38,44,46,47 0.76 (0.53–1.02) Cohort, case control Fair-good All groups37,44,47 0.74 (0.36–1.45) Cohort Fair-good Cardiovascular disease§ Incidence Current47 1.27 (0.80–2.00) Cohort Fair Ever31,43 1.35 (0.92–2.00) Cohort, case control Good Past47 1.26 (0.79–2.08) Cohort Fair All groups31,43,47 1.28 (0.86–2.00) Cohort Fair-good Mortality Current38,40,44,47 0.64 (0.44–0.93) Cohort, case control Fair-good Ever41,43,45,48 0.81 (0.58–1.13) Cohort Fair-good Past44,47 0.79 (0.52–1.09) Cohort Fair-good All groups38,40,41,43,45,47,48 0.75 (0.42–1.23) Cohort Fair-good Stroke Incidence, ever Overall stroke19,31,43,47,50,54 1.12 (1.01–1.23) Cohort Fair-good Thromboembolic19,43,50,53 1.20(1.01–1.40) Cohort Fair-good Subarachnoid19,53,55 0.80 (0.57–1.04) Cohort Fair-good Intracerebral50,53,54,56 0.71 (0.25–1.29) Cohort Fair-good Mortality, ever19,40,41,44,46,47,51,52,57 0.81 (0.71–0.92) Cohort Fair-good ThromboembolismΠ Overall, current 4,58,68 2.14 (1.64–2.81) RCT, case control, cohort Poor-good During first year only4,61,63–66 3.49 (2.33–5.59) RCT, case control Poor-good After the first year4,61,63,66 1.91 (1.18–3.52) RCT, case control Poor-good *Based on meta-analyses conducted by the authors or individual studies as indicated. † RCT indicates randomized controlled trial. ‡ Defined in Harris et al.13 § Includes multiple cardiovascular outcomes such as coronary heart disease, stroke, sudden cardiac death, and congestive heart failure. Π Includes deep vein thrombosis, pulmonary embolism, or both. Source: Reprinted with permission from Journal of the American Medical Association, Vol 288, HD Nelson, LL Humphrey, P Nygren, SM Teutsch, JD Allan, p 874, © 2002, American Medical Association.
patients with coronary heart disease failed to show benefit in cardiovascular outcomes.159 Others have suggested that the more rigorous studies show least evidence to link infection and atherosclerosis.160 From this perspective, strengthening the quality of research requires more specific, sensitive, and standardized reagents and assays, as well as taking into account interactions with other risk factors. It has been suggested that concurrent infection with multiple pathogens may be the factor that stimulates sufficient inflammatory and procoagulant conditions in the in-
dividual to cause atherosclerosis.161 Another hypothesis is that high prevalence of CMV infection explains the association of low socioeconomic status with coronary risk.162 The special case of influenza virus has been advocated by Madjid and others, who have pointed to increased cardiovascular deaths during influenza epidemics and suggested that these infections trigger acute coronary events.163 The authors developed a study in St. Petersburg, Russia, where a high rate of autopsy permitted consistent postmortem diagnosis
NA 1.29 (1.02–1.63) 1.41 (0.86–2.31) 2.11 (1.26–3.55) NA NA 1.26 (1.00–1.59) NA NA
0.66 (0.53–0.82) 0.91 (0.67–1.33) 1.12 (1.01–1.23) 2.14 (1.64–2.81) 3.49 (2.33–5.59) 1.0 to 1.14 1.23 to 1.35 1.8 (1.6–2.0) 2.5 (2.0–2.9)
6 4† 1.4 ... ... 8 ... ...
...
17† 0 1† 1.5 3 0 to 2.5 7 to 11 25 53.5
4 ... 27 3
3 34 32 2
Aged 55–64 Years Review WHI
0 3 1.5 3 0 to 6 10 to 15 25 53.5
34
9 37.5 57 4
9 9 1.4 ... ... 11 ... ...
...
13 ... 49 7
Aged 65–74 Years Review WHI
0 6† 1.5 3 0 to 7 11 to 17 25 53.5
68†
33 45 91 7
11.5 19† 1.4 ... ... 12 ... ...
...
47 ... 78 12.5
Aged 75–84 Years Review WHI
Source: Reprinted with permission from Journal of the American Medical Association, Vol 288, HD Nelson, LL Humphrey, P Nygren, SM Teutsch, JD Allan, p 877, © 2002, American Medical Association.
*WHI indicates Women’s Health Initiative; NA, not applicable; and ellipses, data not computed. Nominal CIs are indicated for main outcomes of the trial (breast cancer and coronary heart disease); adjusted CIs, for secondary outcomes. † Estimates are based on extrapolations.
0.66 (0.33–1.33) NA 0.66 (0.32–1.34) 0.63 (0.32–1.24)
Hazard Ratio (95% CI) from WHI*
Events Prevented or Caused per Year, No.
12:31 PM
0.76 (0.56–1.01) 0.44 (0.23–0.84) 0.60 (0.36–0.99) 0.80 (0.74–0.86)
Relative Risk (95% Confidence Interval [CI]) from Review and Meta-Analysis
Hormone Replacement Therapy Use in 10,000 Women: Benefits and Harms per Year
2/26/10
Benefits (prevention) Hip fractures Wrist fractures Vertebral fractures Cases of colon cancer Uncertain benefits Cases of dementia prevented Harms (caused) Coronary heart disease events Strokes Thromboembolic events Thromboembolic events during first year Breast cancer cases ( 5 years’ use) Breast cancer cases ( 5 years’ use) Cholecystitis cases ( 5 years’ use) Cholecystitis cases ( 5 years’ use)
Table 15-14
46894_CH15_431_502.qxd Page 479
4. EVOLVING AND EMERGING FACTORS 479
46894_CH15_431_502.qxd
480
2/26/10
12:31 PM
Page 480
CHAPTER 15 OTHER PERSONAL FACTORS
Risk factors affecting outcome of infection
Risk factors affected by infection
Age, Sex Heredity Infecting strain and dose Transmission route
Acute respiratory infection
Smoking
Chronic infection in lungs
Chronic bronchitis
Systemic spread inside monocytes Atherosclerotic lesions in vascular wall Age Smoking Sex Heredity Physical activity Diet, iron
Activation of persistent infection with oxidation Production of cytokines, CAMs, Hsp60 etc, Progression of lesion
Reinfection Other infections Stress Unknown factors
Exacerbation of inflammation Production of proteases, phospholipases
Raised CRP, fibrinogen, cytokines Changes in lipid metabolism Obesity Insulin resistance Hypertension Fatigue Depression
Lesion rupture Thrombosis
Figure 15-17 Hypothetical Model on the Participation of Infection in the Development of Atherosclerosis and Coronary Heart Disease, and the Interaction of Infections with Established and Novel Risk Factors of Coronary Heart Disease. Source: Reprinted with permission from The Lancet Infectious Diseases, Vol 2, M Leinonen, P Saikku, p 15, 2002.
Table 15-15
Summary of Degree of Evidence for the Potential Atherogenic Role of Different Pathogens Laboratory/Animal Evidence Evidence from Studies in Humans Other Pathology Laboratory Animal Cross-Sectional Cohort Randomized Studies Evidence Models Epidemiologic Epidemiologic Trial
BACTERIA C. pneumoniae H. pylori Periodontal bacteria C. burnetii Pyogenic bacteria VIRUSES CMV Other herpes Enterovirus Hepatitis A
/ / /
/
/ / / /
/
/
/ / /
Source: Reprinted with permission from Seminars in Vascular Medicine, Vol 2, No 4, GDO Lowe, J Danesh, Guest Eds, Classical and Emerging Risk Factors for Vascular Disease. J Nieto, p 407, © 2002 by Thieme Medical Publishers, Inc.
46894_CH15_431_502.qxd
2/26/10
12:31 PM
Page 481
4. EVOLVING AND EMERGING FACTORS 481
of coronary heart disease as a cause of death. Cautions in interpreting these findings were presented by Smeeth and others, who raised the cold weather associated with influenza outbreaks as a potential contributor to the observations and pointed to nonspecificity in the relation given that other infections and other outcomes were similarly associated.164 Notwithstanding these reservations, an American Heart Association/American College of Cardiology (AHA/ACC) Science Advisory on influenza vaccination in persons with cardiovascular disease underscored the magnitude of the problem: More than 36,000 deaths and 225,000 hospitalizations per year were attributed to influenza in the United States, with cardiovascular disease and diabetes predisposing to complications of influenza infection.165 Inactivated vaccine was recommended for use in persons with existing cardiovascular disease, for whom immunization remains well below recommended levels, especially for specific age and ethnic groups. Trivalent inactivated vaccine was the form recommended for persons at risk of medical complications, including those with cardiovascular conditions.166 Anti-Retroviral Therapy in HIV/AIDS The American Heart Association convened a State of the Science Conference in 2007 to discuss an initiative to decrease cardiovascular risk and increase quality of care for patients living with HIV/AIDS.167,168 The success of antiretroviral therapy prolongs life for many patients with HIV/AIDS. Cardiovascular disease is being recognized as an increasingly prominent problem in this population, whether because of associated dyslipidemia, insulin resistance, inflammation, or change in body composition, or a more direct toxic effect of antiretroviral medications. The need for specific screening, risk prediction, and treatment algorithms for this population was emphasized. Reports have implicated two antiretroviral drugs, abacavir and didanosine, as possibly increasing risk of myocardial infarction and coronary heart disease.169 It is suggested that adverse effects of highly active antiretroviral therapy (HAART) include increased triglycerides and LDL-cholesterol and decreased HDL-cholesterol.170 Impaired glucose tolerance also occurs with this regimen. Recommendations for management of cardiovascular risk focus on adherence to the Adult Treatment Panel III (ATPIII) approach of the National Cholesterol Education Program, emphasizing therapeutic lifestyle change and attention to the lipid disorders that accompany treatment for HIV/AIDS. Pro- and Antioxidants On the basis of Steinberg and colleagues’ review of laboratory investigations conducted over a 20-year pe-
riod and a workshop of the National Heart, Lung and Blood Institute in 1991, a metabolic mechanism was proposed that strongly influences the behavior of the low-density lipoprotein (LDL) molecule and potentiates its role in atherogenesis.171,172 This process results in oxidatively modified (oxidized) LDL, which promotes atherogenesis in several ways. These include facilitating migration of monocytes from the circulation into the intima of the arterial wall; inhibiting motility of these cells, thus limiting their outmigration; increasing accumulation of monocytes and conversion of these to macrophages (scavenger cells) in the intima; and increasing the rate of production of “foam cells,” due to both the increased number of macrophages and the greatly increased propensity of oxidized LDL, over that of nonoxidized LDL, to enter these cells. A direct action of oxidized LDL is to damage endothelial cells, with loss of integrity of the endothelial layer. This initiates a succession of adverse consequences, including localized platelet adhesion and promotion of thrombosis. Antioxidants Natural defenses against this process comprise a variety of biochemical competitors for free oxygen radicals and related forms of oxygen. They effectively prevent oxidation of LDL, as well as DNA and other molecules, thus protecting against atherosclerosis, cancer, and other disorders. These competitors are the antioxidants, described in Table 15-16.173 Several of them are familiar vitamins or vitamin precursors. They include substances related to both fat-soluble vitamins A and E and the water-soluble vitamin C. These antioxidants have been the most closely studied epidemiologically. The figure indicates various mechanisms of action by which one or another prevents production of oxidized LDL. As either vitamins or their precursors, these antioxidants are dietary constituents. But they are available and commonly consumed as supplements, often in quantities many times greater than those provided in the usual diet. This distinction between dietary consumption of antioxidants and exposure to them as supplement products is important in interpreting evidence from both observational studies and trials. Population Comparisons and Cohort Studies. Population differences in plasma concentrations of vitamin E were investigated in relation to coronary heart disease mortality in a case-comparison study organized within the framework of the World Health Organization MONICA Project.174 Men aged 40–49 years were selected in 16 European centers of the Project with an approximate fivefold range of mortality. Plasma samples were processed to determine concentrations of vitamins A, C, and E; precursors of
46894_CH15_431_502.qxd
482
2/26/10
12:31 PM
Page 482
CHAPTER 15 OTHER PERSONAL FACTORS
Table 15-16
The Nature of Selected Antioxidants Type and Location of Antioxidants in Blood Tocopherols, -carotene, lycopenes, Ascorbic acid, protein thiols, bilirubin, urate coenzyme Q-10 Fat-soluble Water-soluble LDL particle, cell membrane Blood plasma (extracellular fluid)
Antioxidant Type Location
Proposed Mechanisms for Relationship of Antioxidants to Risk Factors for Heart Disease Vitamin Antioxidant Tocopherols (Vitamin E),
-Carotene (Vitamin A) Ascorbic Acid (Vitamin C) Relationship to risk factors (epidemiological)
Smokers found to have lower carotene levels Smokers found to have more oxidized E in lungs
Inverse relationship to blood pressure Direct relationship to HDL levels Lower levels in smokers
Hypothesized mechanism of action
Inhibits LDL oxidationa Suppresses LDL uptake by macrophagesa Reduces incidence of major coronary eventsa,b Protects against oxidative damage during reperfusiona
Protects and restores parent vitamin E Inhibits LDL oxidation in vitro Suppresses LDL uptake by macrophages Inhibits lipid peroxide formation
Note: HDL, high-density lipoprotein; LDL, low-density lipoprotein. a Vitamin E. b
-carotene. Source: Reprinted with permission from Nutrition Today, Vol 27, pp 30–33, ©1992, Lippincott Williams and Wilkins.
vitamin A (carotene and other carotenoids); and selenium. The strongest predictive relation between plasma antioxidant levels and coronary mortality was found in a model that included vitamins E and A, cholesterol, and diastolic blood pressure. Mortality predicted from this model corresponded closely with observed mortality and thus suggested inverse association between plasma concentration of vitamin E and coronary mortality as these varied across 16 European populations. Cohort studies in large numbers of US health professionals indicated, for both women and men, inverse association between use of vitamin E supplements and risk of coronary events, with significantly reduced relative risks of about 0.60 for both women and men.175,176 In neither group was dietary intake alone significantly related to coronary event rates. Intake from supplements was many times greater than dietary intake, with the upper quintile by diet being 11.1 international units (IU) per day and by supplements 250 IU/day. However, corresponding point estimates of relative risk were 0.79 for the highest quintile group by diet and 0.70 for the highest quintile group by supplemental intake. A subsequent cohort study among more than 34,000 postmenopausal women found that reported dietary intake of vitamin E, but not
of vitamin A, retinol, carotenoids, or vitamin C, was associated with lower coronary mortality with a significant multivariate-adjusted relative risk of 0.38.177 According to other reports, vitamin C, flavonoids (another type of antioxidant substance found principally in tea), and metabolic products of sun-induced vitamin D all add to the menu of antioxidants apparently associated with reduced risk of coronary heart disease in observational studies. Clinical and Preventive Trials. Numerous trials of antioxidant administration to prevent coronary events have been conducted with vitamin E (-tocopherol) and with -carotene, sometimes in combination, with or without vitamin A, C, or cardiovascular drugs. An early summary of evidence from available studies, both observational and experimental, was reported by Jha and colleagues in 1995.178 For -carotene, vitamin E, and vitamin C, the observational study results were mixed but, on balance, favorable for lower coronary heart disease risk. However, the trials showed no benefit of intervention. A more recent systematic review summarized findings from large trials, each with 1000 or more participants, in both primary and secondary prevention (Table 15-17).179 Overall results were at best indicative of no effect and often showed increased risk
Finland
USA
USA USA USA Australia
Australia
USA
France
Alpha Tocopherol Beta Carotene Cancer Prevention Study (ATBC)40,41
Beta Carotene and Retinol Efficacy Trial (CARET)42
Physicians’ Health Study (PHS)43
Women’s Health Study44
Women’s Health Study45
Vitamin A and Cancer Prevention II46
Skin Cancer Prevention Trial47
Physicians’ Health Study II23
SUpplementation en VItamines et Mineraux AntioXydants (SU.VI.Max)48
12,735 men (45–69 years) and women (–60 years)
15,000 healthy male physicians, 55 years
1720 men and women, 27–84 years, with recent nonmelanoma skin cancer
1204 former asbestos workers, men and women, 40–83 years
39,876 healthy women, 45 years
39,876 healthy women, 45 years
22,071 male physician, 40–84 years
8
12
4.3
5
Unknown
2.1
12
4
No effect on CVD mortality
Effects on CVD awaited
Unpublished data show no effect on CVD
50 mg -carotene
50 mg -carotene, 400 IU -tocopherol (alternate days) plus 500 mg vitamin C, multivitamin (daily) 6 mg -carotene, 30 mg -tocopherol, 120 mg vitamin C, 100 g selenium and 20 mg zinc
(continues)
No effect of -carotene on CHD mortality
No effect on incidence of MI or stroke or on CVD mortality
50 mg -carotene (alternate days)
30 mg -carotene or 25 000 IU retinol (no placebo group)
No effect on incidence or mortality from MI or stroke
50 mg -carotene and/or aspirin (alternate days)
Effect on MI, stroke and CVD mortality awaited
26% ↑ in CVD (NS) 17% ↑ in total mortality
30 mg -carotene and 25 000 IU retinol
600 IU -tocopherol and/or 100 mg aspirin (alternate days)
11% ↑ in CHD mortality among
-carotene group 50% ↑ in hemorrhagic stroke mortality among vitamin E group 62% ↑ in intracerebral hemorrhage among -carotene group 14% ↓ in cerebral infarction among vitamin E group
50 mg -tocopherol and/or 20 mg
-carotene
12:31 PM
14,254 heavy smokers, 4060 absestos workers, 45–69 years
6.1
Non-significant ↓ in cerebrovascular mortality
Results
15 mg -carotene, 30 mg -tocopherol and 50 µg selenium
Daily Dose
2/26/10
29 133 male cigarette smokers, 50–69 years
1000 Subjects): Antioxidants and CVD Large Intervention Trials ( Duration of Treatment Study Country Study Population (Years) PRIMARY PREVENTION Linxian Cancer 5.2 China 29,584 poorly nourished men Prevention Study39 and women, 40–69 years
Table 15-17
46894_CH15_431_502.qxd Page 483
4. EVOLVING AND EMERGING FACTORS 483
No effect of vitamin E on MI, stroke or CVD death
No effect of vitamin E on any prespecified CVD endpoint including CVD mortality, MI and stroke. However, the study had inadequate power due to premature interruption of the trial Effect on CVD awaited
No reduction in fatal or non-fatal MI or stroke
400 IU -tocopherol and/or ACE inhibitor
Low-dose aspirin and/or 300 mg -tocopherol
50 mg -carotene (alternate days) or 600 IU -tocopherol (alternate days) or 500 mg vitamin C (daily) 20 mg -carotene, 600 mg -tocopherol and 250 mg vitamin C
3.5
11 324 patients with recent MI (no defined age range)
Italy
Canada
Italy
USA
UK
GISSI Prevenzione Trial52
Heart Outcomes Prevention Evaluation Study (HOPE)53
Primary Prevention Project (PPP)54
Women’s Antioxidant Cardiovascular Study (WACS)55
Heart Protection Study56
5
4
3.6
Source: Reprinted with permission from Public Health Nutrition, Vol 7, SA Stanner, J Hughes, CNM Kelly, J Butriss, pp 409–410, © The Authors 2003.
CVD—cardiovascular disease; GISSI—Gruppo Italiano per no Studio della Sopravvivenza nell’Infarto Miocardio; MI—myocardial infarction; PUFA—polyunsaturated fatty acids; ACE—angiotensin-converting enzyme. CHD—coronary heart disease; NS—not significant. *Secondary prevention is defined as including patients with known or documented vascular disease.
20 536 high-risk men and women, 40–80 years
8000 women with prior CVD event or 3 coronary risk factors, 40 years
4495 men and women with one or more CVD risk factors, mean age 64 years
4–6
No benefit from vitamin E 15% ↓ in risk of death, non-fatal MI and stroke from n – 3 PUF A
300 mg -tocopherol and/or 1 g n – 3 PUFA
1.4
2002 patients with coronary atherosclerosis, mean age 62 years
UK
Cambridge Heart Antioxidant Study (CHAOS)51
12:31 PM
9541 high-risk men and women, 55 years
77% ↓ in non-fatal MI No benefit on CVD mortality
400 or 800 IU -tocopherol
4
1795 male heavy smokers with previous angina pectoris, 50–69 years
Finland
ATBC50
Reduced non-fatal acute ischemia (vitamin E and -carotene) No effect on risk of MI (vitamin E) Increased risk of MI ( -carotene) No effect on symptoms or progression of angina pectoris (vitamin E)
Results
484
2/26/10
50 mg -tocopherol and/or 20 mg
-carotene
1000 Subjects): Antioxidants and CVD—continued Large Intervention Trials ( Duration of Treatment Study Country Study Population (Years) Daily Dose SECONDARY PREVENTION* 50 mg -tocopherol and/or 20 mg ATBC49 5.3 Finland 1862 male heavy smokers with
-carotene previous MI, 50–69 years
Table 15-17
46894_CH15_431_502.qxd Page 484
CHAPTER 15 OTHER PERSONAL FACTORS
46894_CH15_431_502.qxd
2/26/10
12:31 PM
Page 485
4. EVOLVING AND EMERGING FACTORS 485
of cardiovascular outcomes. There were few exceptions. This led the authors to conclude:179, p 407 The suggestion that antioxidant supplements can prevent chronic diseases has not been proved or consistently supported by the findings of published intervention trials. Further evidence regarding the efficacy, safety, and appropriate dosage of antioxidants in relation to chronic disease is needed. The most prudent public health advice remains to increase the consumption of plant foods, as such dietary patterns are associated with reduced risk of chronic disease. Further, posttrial, follow-up of participants in the ATBC Study among male smokers in Finland (see Table 15-17) extended observations by 6 years.180 Risk of posttrial coronary events was slightly, but not significantly, reduced among participants who had received -tocopherol during the trial. Major coronary events, nonfatal myocardial infarction, and fatal coronary heart disease were all increased, all but fatal events significantly, among those who had received -carotene during the trial. Neither form of supplement was recommended for prevention of CHD among male smokers on the basis of these results. The HOPE Trial of -tocopherol in secondary prevention of cardiovascular diseases and cancer (also in Table 15-17) was also extended, but with continued intervention, for a median follow-up of 7 years.181 It was concluded that long-term vitamin E supplementation in patients with vascular disease or diabetes did not prevent major cardiovascular events or cancer and might increase risk of heart failure. Again, as with HRT, favorable findings from observational studies of vitamin intakes or supplements were not borne out by randomized controlled trials. Pro-Oxidants If the action of antioxidants protects against transformation of LDL to an especially virulent promoter of atherosclerosis, antioxidants would be opposed by any factor whose action favors production of oxidized LDL. Certain metals or their compounds may be candidates. In the laboratory, iron and copper may be used to amplify oxidation of LDL, and hemin (an iron-containing compound) renders endothelial cells in culture highly vulnerable to damage by oxidants.182 LDL that has been oxidized by hemin is also reported to be extremely toxic to endothelial cells in culture. An epidemiologic observation adds to the plausibility of such a process in human atherosclerosis, on the basis of a cohort study of 1931 Finnish men at selected ages in their 40s, 50s, or 60s at the beginning of a 5-year follow-up period.183 Serum ferritin concentra-
tion was measured as an index of excess iron storage in the body and ranged from 10 to 2270 mg/L (mean value, 166 mg/L). When the men with ferritin concentrations of 200 mg/L or greater were compared with those with lower values, after adjustment for several correlated risk factors, the relative risk was 2.2 (CI 1.2–4.0). In the group of men with LDLcholesterol concentration of 5 mmol/L or greater, the corresponding relative risk was 4.7 (CI 1.4–16.3). These results were interpreted as evidence for a role of excess dietary iron in increasing the risk of acute myocardial infarction in men and possibly postmenopausal women. This might also explain variation in the relation of LDL cholesterol to risk of coronary heart disease in populations with low levels of stored iron. It was suggested that dietary guidelines should caution against excessive iron intake. Subsequent reports include the 13-year follow-up of 4237 participants aged 40–74 years in the First National Health and Nutrition Examination Survey.184 Serum iron concentration was found inversely related to risk of myocardial infarction in women (relative risk 0.82, CI 0.70–0.95), and there was no association in men. There was no support for an increase in risk in relation to either serum iron or transferrin. Interpretation of the conflicting US and Finnish findings concerns differences in iron components measured (iron and transferrin versus ferritin), variability in the measures used in the US study, and less frequent occurrence of excess iron stores in the US population. Hyperhomocysteinemia Biochemical, animal experimental, and clinical observations from the 1930s to the 1970s identified several metabolic products of methionine (a sulfurcontaining essential amino acid), linked them with a rare human disease that included atherosclerosis with thrombotic complications at very early ages, and established their potential role in atherogenesis.185 The central focus is on three closely related compounds: homocysteine, a single-chain four-carbon molecule with one amino and one sulfhydryl group; homocystine, with a disulfide bond joining two such molecules; and cystine-homocysteine disulfide, in which homocysteine and a three-carbon sulfhydryl-containing amine are joined. The total plasma concentration of the homocysteine comprising the three compounds together is conventionally denoted by homocyst(e)ine, tHcy, H(e), or simply homocysteine, the present usage. Aspects of metabolism, laboratory determination, and nutritional and genetic influences were reviewed recently.186 A schematic representation of homocysteine metabolism shows a close link with B vitamins. Either a pyridoxal phosphate-dependent
46894_CH15_431_502.qxd
486
2/26/10
12:31 PM
Page 486
CHAPTER 15 OTHER PERSONAL FACTORS
enzymatic reaction or a transfer of methyl groups from vitamin B12 or folic acid is required to convert homocysteine to its immediate metabolic products. This suggests strongly that adverse levels of homocysteine may reflect relative deficiency of one or more components of the B vitamins, a potential point of intervention should elevated homocysteine concentration be considered causally related to cardiovascular disease. In addition, numerous factors increase blood homocysteine levels, among them older age, male sex, smoking, coffee and alcohol consumption, diabetes mellitus, and several medications.
Study Type
Events, No.
Observational Studies A meta-analysis by the Homocysteine Studies Collaboration reported in 2002 assembled individuallevel data from 12 prospective and 18 retrospective studies that included more than 5000 ischemic heart disease events and more than 1000 strokes.187 The central question addressed was how great a reduction in odds ratios for these events might be expected from a 25 percent lower level of homocysteine than was observed in these studies. Figure 15-18 presents the results for ischemic heart disease from each study and groups of studies by type of design. The direction of
OR (95% CI)
Prospective 82 Stenouwer et al15 Bots et al18 84 96 Arnesen et al11 Alfthan et al10 125 Omenn et al (unpublished 166 data. September 2002) 167 Ubbink et al14 13 209 Evans et al Whincup et al19 210 215 Wald et al17 Folsom et al16 238 Verhoef et al29 376 Subtotal 1968 Heterogeneity 102 21 (P .02)
0.85 (0.641.14) 0.61 (0.430.86) 0.65 (0.500.84) 0.92 (0.651.31) 0.76 (0.561.03) 0.89 (0.731.09) 1.19 (0.921.54) 0.92 (0.741.15) 0.66 (0.530.81) 0.84 (0.720.99) 0.84 (0.681.03) 0.83 (0.770.89)
Retrospective (Population Controls) Schwartz et al30 Pancharuniti et al21 Joubran et al31 Verhoef et al29 Verhoef et al25 Genest et al20 Hopkins et al23 von Eckardstein et al22 Silberberg et al28 Graham et al27 Malinow et al26 Chambers et al32 Subtotal
50 78 109 122 126 155 168 183 260 337 381 527 2496
0.39 (0.270.57) 0.65 (0.411.02) 0.76 (0.581.00) 0.63 (0.420.96) 0.63 (0.460.87) 0.51 (0.400.64) 0.46 (0.330.64) 0.48 (0.330.68) 0.77 (0.600.98) 0.79 (0.680.92) 0.59 (0.510.69) 0.81 (0.710.92) 0.67 (0.620.71)
Heterogeneity 112 39 (P .001) Retrospective (Other Controls) Dalery et al34 Lotin et al36 Robinson et al35 Subtotal
135 177 297 609
Heterogeneity 22 13 (P .001)
0.77 (0.630.93) 0.99 (0.771.28) 0.51 (0.400.66) 0.73 (0.640.83)
0.2
0.4
0.6
0.8 1.0 1.2 1.4
Odds Ratio (95% CI) Data were adjusted for study, sex, and age at enrollment and were corrected for regression dilution. The size of the square is inversely proportional to the variance of the log odds ratio (OR). The horizontal lines represent the 95% confidence intervals (Cls). The combined ORs in the subtotals for each study design and their 95% Cls are indicated by the diamonds.
Figure 15-18 Odds Ratios of Ischemic Heart Disease for a 25% Lower Usual Homocysteine Level in Individual Studies. Source: Reprinted with permission from Journal of the American Medical Association, Vol 288, No 16, The Homocysteine Studies Collaboration, p 2018, © 2002, American Medical Association.
46894_CH15_431_502.qxd
2/26/10
12:31 PM
Page 487
4. EVOLVING AND EMERGING FACTORS 487
results was generally quite consistent among all three types of studies shown, although the estimated reduction in odds was notably less in prospective than in retrospective studies. Larger studies tended to show more modest decreases in the odds ratios. For stroke (not shown), the contributing studies were fewer in number, smaller in size, much wider in confidence limits, and somewhat less consistent in direction or magnitude of change in odds ratio for a 25% reduction in homocysteine levels. Separately for ischemic heart disease and stroke, overall odds ratio reductions for 25% lower homocysteine levels were summarized as in Table 15-18. Here the influence of other factors is shown by successive levels of adjustment across the table. Some, but only slight, attenuation of the estimated benefit of lower homocysteine levels was observed as age, sex, smoking, systolic blood pressure, and total cholesterol level were taken into account. There remained after full adjustment an estimated 11% reduction in ischemic heart disease and 19% reduction in stroke with reduced homocysteine levels. The authors concluded:187, p 2015 This meta-analysis of observational studies suggests that elevated homocysteine is at most a modest independent predictor of IHD [ischemic heart disease] and stroke risk in healthy populations. Studies of the impact on disease risk of genetic variants that affect blood homocysteine concentrations will help determine whether homocysteine is causally related to vascular disease, as may large randomized trials of the effects on IHD and stroke of vitamin supplementation to lower blood homocysteine concentrations.
Table 15-18
IHD Stroke
Through a different approach to analysis, an earlier review had similarly found that, based on prevailing homocysteine levels in the population, about 10% of the population risk of coronary heart disease was attributable to this factor.188 On this assumption, the public health impact of increased folic acid intake was estimated. Among three alternatives of increased dietary intake—natural food sources, supplementation by tablets, and fortification of grains to increase folate content—that study concluded that the latter would be the most effective and could prevent from 13,500 to 50,000 coronary deaths annually in the United States. Regarding potential insights from study of genetic variation, an enzyme that facilitates the transfer of a methyl group from a folate derivative to homocysteine, converting it to methionine, is methylenetetrahydrofolate reductase (MTHFR).189 A common mutation in this enzyme renders it unstable and may lead to elevated homocysteine concentration, possibly requiring even greater folate supplementation than has generally been proposed. A review was conducted of 40 case-control studies of the MTHFR 677C (arrow) T polymorphism, reported from Europe, North America, and elsewhere, predominantly during the late 1990s. The TT genotype was found to be associated with 16% greater odds of CHD than the CC genotype. However, results differed importantly between Europe (14% increased odds) and North America (13% decreased odds). This difference was interpreted as indicating interaction between genotype and population distributions of folate, with the TT-related increase being present only when folate levels are low. The authors concluded that their findings supported the hypothesis of a causal relationship between impaired folate
Odds Ratios for Ischemic Heart Disease (IHD) and for Stroke Associated with 25% Lower Usual Homocysteine Levels in Prospective Studies Adjusted Odds Ratio (95% Confidence Interval) Age, Sex, Age, Sex, Smoking, Smoking, and Systolic Blood Age, Sex, and Systolic Blood Pressure, and Total Events, No.* Age and Sex Smoking Pressure Cholesterol Level 1855 0.83 (0.77–0.90) 0.85 (0.78–0.91) 0.89 (0.82–0.96) 0.89 (0.83–0.96)
21 24
21 20
21 10
21 9 435 0.77 (0.66–0.90) 0.78 (0.67–0.91) 0.81 (0.69–0.96) 0.81 (0.69–0.95)
21 11
21 10
21 6
21 6
*Among people with all available data used for adjustment for known cardiovascular risk factors. Source: Reprinted with permission from Journal of the American Medical Association, Vol 288, No 16, The Homocysteine Studies collaboration, p 2019, © 2002, American Medical Association.
46894_CH15_431_502.qxd
488
2/26/10
12:31 PM
Page 488
CHAPTER 15 OTHER PERSONAL FACTORS
metabolism, resulting in high homocysteine levels and increased risk of coronary heart disease. Clinical Trials Two major trials of homocysteine reduction through administration of folic acid or B vitamin components, versus placebo, to prevent cardiovascular outcomes in persons with existing cardiovascular disease were reported in 2006—the Heart Outcomes Prevention Evaluation (HOPE) 2 and the Norwegian Vitamin (NORVIT) Trial.190,191 Differences in design and study populations notwithstanding, the findings of the two trials were similar. Homocysteine levels were reduced with intervention, but the risk of major cardiovascular events was not reduced. NORVIT found, further, an increased relative risk in the group receiving the combination of folic acid and vitamins B12 and B6. Smith and Shah, in an editorial following preliminary presentation of NORVIT results, noted that some populations may yet be shown to benefit from folate or B vitamin supplementation or benefit may be found in primary though not in secondary prevention.192 They called, in addition, for closer scrutiny to understand why the observational studies “reached the wrong conclusions.”192, p 1680 The same concern applies to the genetic studies of the type cited previously regarding MTHFR, which, if presumed free of any bias because of “Mendelian randomization” at conception, should have led to the right conclusions. Inflammation and C-Reactive Protein Hansson, in a 2005 review, presented a coherent picture of inflammation, atherosclerosis, and coronary artery disease with exceptional graphic clarity.193 Topics addressed included the nature of atherosclerotic lesions, evolution of the rupture-prone atherosclerotic plaque, and acute coronary syndromes. Inflammation was characterized as an expression of balance between inflammatory and antiinflammatory activity at the level of the arterial wall, interacting with metabolic factors to determine the progression of atherosclerosis. Multiple inflammatory factors are elevated in the blood in association with activated plaques in acute coronary syndromes and have prognostic value in predicting outcomes. These markers are also elevated, though only moderately, as part of the “smoldering inflammation” that characterizes silent plaques. Although several inflammatory markers are thus associated with coronary artery disease, Hansson considered it unlikely that any of them, including C-reactive protein, is causally related. Instead, they were seen as reflections of the inflammatory process in the artery and perhaps other tissues such as adipose tissue.
A workshop had been convened in 2002 by the Centers for Disease Control and Prevention and AHA to consider application of knowledge of inflammatory markers to clinical and public health practice.194 Scientific evidence regarding several factors, including high-sensitivity C-reactive protein (hs-CRP) and others was reviewed; clinical chemistry and assays were considered; areas for research were identified; recommendations for testing for markers were developed; and public health implications of these associations were explored. Among the markers addressed, hs-CRP was alone in meeting criteria for practical clinical laboratory testing with commercially available assays that could be standardized and have adequate precision. The workshop summary therefore focused on hs-CRP. This marker was found to have moderately strong association with clinical cardiovascular disease but some inconsistency across studies and unknown applicability among specific age, sex, and race-ethnic groups. Testing for hs-CRP was not recommended but was thought to have potential utility in upgrading risk assessment for persons otherwise found to have intermediate CVD risk or for adding prognostic information in predicting CVD outcomes. It was not considered appropriate to rely on hs-CRP in making treatment decisions. A subsequent review from the perspective of laboratory determination of emerging biomarkers for primary prevention of cardiovascular disease similarly found hs-CRP alone to meet criteria for acceptance in risk assessment for primary prevention.195 Ridker and others, on the basis of analysis of the Women’s Health Study, considered both LDLcholesterol and C-reactive protein as biologic markers of cardiovascular risk and compared their independent predictive value for first cardiovascular events.196 The curves for event-free survival by quintile levels of each of these measures are shown in Figure 15-19. The range of differences in 8 years is small, from 100 to about 96 percent, and except for the first and second quintiles by LDL-C that are indistinct, in contrast to those for CRP, there is very little difference in survival according to baseline rank of the two measures. The authors found that C-reactive protein did add to the predictive value of the Framingham risk score for future coronary events and concluded that it was a stronger predictor of outcomes than LDL-C. Others found elevated CRP to be largely attributable to other risk factors and therefore not useful as a screening tool.197 However, additional interest follows from the finding of CRP levels in children and adolescents that correspond to the high-risk category adopted for adults, 3.0 mg/L.198 A workshop convened by the National Heart, Lung and Blood Insti-
46894_CH15_431_502.qxd
2/26/10
12:31 PM
Page 489
4. EVOLVING AND EMERGING FACTORS 489
C-Reactive Protein
LDL Cholesterol 1.00
1.00
1st Quintile
Probability of Event-free Survival
0.99
2nd Quintile
0.99 1st Quintile 2nd Quintile
3rd Quintile
0.98
0.98 3rd Quintile
4th Quintile 4th Quintile 0.97
0.97
5th Quintile
5th Quintile 0.96
0.96
0.00
0.00 0
2
4 6 Years of Follow-up
8
0
2
4 6 Years of Follow-up
8
Figure 15-19 Event-Free Survival According to Baseline Quintiles of C-Reactive Protein and LDL Cholesterol. The range of values for C-reactive protein was as follows: first quintile, 0.49 mg per liter; second quintile, 0.49 to 1.08 mg per liter; third quintile, 1.08 to 2.09 mg per liter; fourth quintile, 2.09 to 4.19 mg per liter; fifth quintile, 4.19 mg per liter. For LDL cholesterol, the values were as follows: first quintile, 97.6 mg per deciliter; second quintile, 97.6 to 115.4 mg per deciliter; third quintile, 115.4 to 132.2 mg per deciliter; fourth quintile, 132.2 to 153.9 mg per deciliter; fifth quintile, 153.9 mg per deciliter. To convert values for LDL cholesterol to millimoles per liter, multiply by 0.02586. Note the expended scale on the ordinate.Source: Reprinted with permission from The New England Journal of Medicine, Vol 347, No 20, PM Ridker, N Rifai, L Rose, JE Buring, NR Cook, p 1559, © 2002 Massachusetts Medical Society.
tute in 2006 concluded that it remained controversial whether CRP plays a pathophysiologic role in atherosclerosis.199 Further research was called for regarding potential utility of CRP in study of the etiology of CVD and in risk categorization of patients as a guide to preventive therapy. It was further recommended that circulating inflammatory mediators in general, and not CRP alone, should be studied. A large multicenter trial with nearly 18,000 participants in 26 countries was undertaken to evaluate the effect of a statin in preventing vascular events in apparently healthy men and women with LDLcholesterol less than 130 mg/dl but having CRP levels of 2.0 mg/L or higher—the Justification for Use of Statins in Prevention: an Intervention Trial Evaluating Rosuvastatin (JUPITER).200 The trial was terminated early, after a median follow-up of 1.9 years, because of the benefit of treatment over placebo. LDL cholesterol was reduced by 50% and CRP by 37%. Major vascular events were reduced from 1.36 per
100 person-years in the placebo group to 0.77 with treatment, corresponding to a hazard ratio of 0.56 (95% CI, 0.46–0.69). All-cause mortality was also significantly reduced. Physician-diagnosed diabetes mellitus was increased to a small but significant degree, warranting further evaluation. Public health implications of the JUPITER results are substantial, as data from NHANES 1999–2004 indicate that, under the eligibility criteria for the trial, approximately 80% of US men aged 50 years or older and women aged 60 years or older would be eligible for rosuvastatin therapy for primary prevention of cardiovascular outcomes.201 “Novel” Markers and Risk Prediction In 1961, the expression “risk factor” was introduced by the Framingham Heart Study investigators in reference to three predictors of coronary events—serum cholesterol concentration, blood pressure, and electrocardiographic evidence of left ventricular hypertrophy.202 In 1981, Hopkins and Williams identified
46894_CH15_431_502.qxd
490
2/26/10
12:31 PM
Page 490
CHAPTER 15 OTHER PERSONAL FACTORS
246 factors associated with occurrence of coronary heart disease, as noted previously.203 In the immediate past 10 years, the theme of innovation in discovery of such individual-level characteristics has been prominent in many reports addressing, for example, “conditional risk factors,”204,205 “potential new cardiovascular risk factors,”206 “emerging risk factors,”207 “novel risk factors,”208 “novel serologic markers of cardiovascular risk,”209 “multiple biomarkers,”210 and “novel risk markers.”211 The candidate characteristics proposed in these several reports are compiled in Table 15-19. A 2009 AHA Scientific Statement, Criteria for Evaluation of Novel Markers of Cardiovascular Risk, responded to the recognized need to assess the utility of novel markers in view of current concepts of risk evaluation.212 Because of the increased availability of novel markers, a systematic approach was seen as warranted in order to make appropriate use of new information to improve patient management. The Statement therefore focused on criteria and methods for critical appraisal of risk assessment methods, including several underlying requirements:212, p 2408
Table 15-19
An adequate evaluation of a novel risk marker requires a sound research design, a representative at-risk population, and an adequate number of outcome events. Studies of a novel marker should report the degree to which it adds to the prognostic information provided by standard risk markers. No single statistical measure provides all the information needed to assess a novel marker, so measures of both discrimination and accuracy should be reported. The clinical value of a marker should be assessed by its effect on patient management and outcomes. In general, a novel risk marker should be evaluated in several phases, including initial proof of concept, prospective validation in independent populations, documentation of incremental information when added to standard risk markers, assessment of effects on patient management and outcomes, and ultimately, cost-effectiveness. Further details were provided in the Statement regarding analytic methods and desirable elements of re-
Recently Proposed Characteristics Associated with Cardiovascular Disease Neurohormonal markers Inflammatory markers Candidate gene polymorphisms B-type natriuretic peptide Cellular adhesion molecules Angiotensin-converting enzyme N-terminal pro-atrial natriuretic C-reactive protein/hs-CRP genotype peptide Inflammatory cytokines ApoE genotype Plasma renin Interleukins 6 and 18 PAI-1 genotype Serum aldosterone Leukocyte count Selectins P and E Hemostasis/thrombosis markers Serum amyloid A Platelet-related factors D-dimer Soluble C40 ligand Aspirin resistance Factors V, VII, and VIII Tumor necrosis factor- Platelet activity Fibrinogen Vascular and cellular adhesion Platelet aggregation Fibrinopeptide A molecules Platelet size and volume Plasminogen activator inhibitor 1 (PAI-1) Lipid-related factors Prothrombin fragment 12 Other factors Soluble thrombomodulin Antibodies against oxidized LDL Asymmetric dimethylarginine Tissue-plasminogen activator Antioxidant deficiency Homocysteine Von Willebrand factor antigen Apolipoproteins A1 and B Insulin resistance Elevated serum triglycerides Left ventricular hypertrophy High-density lipoprotein subtypes Measures of oxidative stress Infectious agents Lipoprotein(a) Microalbuminuria C. pneumoniae Lipoprotein-associated Myeloperoxidase Cytomegalovirus phospholipase A2 Nitrotyrosine H. pylori Oxidized LDL Phytosterols/sitosterol Herpes simplex virus Remnant lipoproteins Pregnancy-associated plasma Small dense low-density phosphatase lipoprotein (LDL) Urinary albumin-to-creatinine ratio Vitamin B6
46894_CH15_431_502.qxd
2/26/10
12:31 PM
Page 491
REFERENCES 491
ports regarding novel risk markers. It can be anticipated that new markers, relevant principally to pathogenesis of atherosclerosis and its cardiovascular consequences at a mechanistic level, will continue to be proposed. Evaluating their contribution to understanding and to clinical practice, as well as their potential public health significance, will remain an active research focus. REFERENCES 1. Alcohol Consumption 1. Committee on Diet and Health, Food and Nutrition Board, Commission on Life Sciences, National Research Council. Diet and Health. Washington, DC: National Academy of Sciences; 1989. 2. Kuller LH. Alcohol and cardiovascular disease. In: Pearson TA, Criqui MH, Luepker RV, Oberman A, Winston M, eds. Primer in Preventive Cardiology. Dallas, TX: American Heart Association; 1994:227–233. 3. Doll R, Peto R, Hall E, Wheatley K, et al. Mortality in relation to consumption of alcohol: 13 years’ observations on male British doctors. Br Med J. 1994; 309:911–918. 4. Liu S, Serdula MK, Byers T, Willamson DF, et al. Reliability of alcohol intake as recalled from 10 years in the past. Am J Epidemiol. 1996;143: 177–186. 5. Emberson JR, Shaper AG, Wannamethee SG, Morris RW, Whincup PH. Alcohol in middle age and risk of cardiovascular disease and mortality: accounting for intake variation over time. Am J Epidemiol. 2005;161:856–863. 6. Wellman J, Heidrich J, Berger K, Döring A, Heuschmann PU, Keil U. Changes in alcohol intake and risk of coronary heart disease and all-cause mortality in the MONICA/KORAAugsburg cohort 1987-97. Eur J Cardiovasc Prevention Rehab. 2004;11:48–55. 7. Cahalan D. Quantifying alcohol consumption: patterns and problems. Circulation. 1981;64 (suppl III):III-7–III-13. 8. Criqui MH. Alcohol and the heart: implications of present epidemiologic knowledge. Contemp Drug Prob. Spring 1994:125–142.
9. Ruidavets J-B, Bataille V, Dallongeville J, et al. Alcohol intake and diet in France, the prominent role of lifestyle. Eur Heart J. 2004;25:1153–1162. 10. Hill JA. In vino veritas: alcohol and heart disease. Am J Med Sci. 2005;329:124–135. 11. Kagan A, Yano K, Rhoads GG, McGee DL. Alcohol and cardiovascular disease: the Hawaiian experience. Circulation. 1981; 64(suppl III):III-27–III-31. 12. Hulley SB, Gordon S. Alcohol and high-density lipoprotein cholesterol. Circulation. 1981;64 (suppl III):III-57–III-67. 13. Wallace RB, Lynch CF, Pomrehn PR, Criqui MH, et al. Alcohol and hypertension: epidemiologic and experimental considerations. Circulation. 1981;64(suppl III):III-41–III-47. 14. Dyer AR, Stamler J, Paul O, et al. Alcohol, cardiovascular risk factors and mortality: the Chicago experience. Circulation. 1981;64 (suppl III):III-20–III-27. 15. Klatsky AL, Friedman GD, Siegelaub AB. Alcohol use and cardiovascular disease: the Kaiser-Permanente experience. Circulation. 1981;64(suppl III):III-32–III-41. 16. Fumeron F, Betoulle D, Luc G, et al. Alcohol intake modulates the effect of a polymorphism of the cholesterol ester transfer protein gene on plasma high density lipoprotein and the risk of myocardial infarction. J Clin Invest. 1995;96: 1664–1671. 17. Ridker PM, Vaughan DE, Stampfer MJ, et al. Association of moderate alcohol consumption and plasma concentration of endogenous tissue-type plasminogen activator. JAMA. 1994; 272:929–933. 18. Folsom AR, Wu KK, Davis CE, et al. Population correlates of plasma fibrinogen and factor VII, putative cardiovascular risk factors. Atherosclerosis. 1991;91:191–205. 19. Folsom AR, Qamhieh HT, Flack JM, Hilner JE, et al., for the investigators of the Coronary Artery Risk Development in Young Adults (CARDIA) Study. Am J Epidemiol. 1993; 138:1023–1026.
46894_CH15_431_502.qxd
492
2/26/10
12:31 PM
Page 492
CHAPTER 15 OTHER PERSONAL FACTORS
20. Mukamal KJ, Jensen MK, Grønbæk M, et al. Drinking frequency, mediating biomarkers, and risk of myocardial infarction in women and men. Circulation. 2005;112:1406–1413. 21. Freiberg MS, Samet JH. Alcohol and coronary heart disease. The answer awaits a randomized controlled trial. Circulation. 2005;112: 1379–1381. 22. Nielsen NR, Schnohr P, Jensen G, Grønbæk M. Is the relation between type of alcohol and mortality influenced by socio-economic status? J Internal Med. 2004;255:280–288. 23. US Department of Health and Human Services. Health, United States, 2008 with Special Feature on the Health of Young Adults. Washington DC: US Department of Health and Human Services. Centers for Disease Control and Prevention. National Center for Health Statistics; 2008. 24. Kerr WC, Greenfield TK, Tujague J. Estimates of the mean alcohol concentration of the spirits, wine, and beer sold in the United States and per capita consumption: 1950 to 2002. Alcoholism: Clin and Exp Res. 2006;30: 1583–1591. 25. Rose G. Ancel Keys lecture. Circulation. 1991; 84:1405–1409. 26. Klipstein-Grobusch K, Slimani N, Krogh V, et al. Trends in self-reported past alcoholic beverage consumption and ethanol intake from 1950 to 1995 observed in eight European countries participating in the European Investigation into Cancer and Nutrition (EPIC). Public Health Nutr. 2002;5:1297–1310. 27. Criqui MH, Ringel BL. Does diet or alcohol explain the French paradox? Lancet. 1994; 344:1719–1723.
cohol consumed in coronary heart disease in men. N Engl J Med. 2003;348:109–118. 30. Grønbæk M, Johansen D, Becker U, et al. Changes in alcohol intake and mortality. A longitudinal population-based study. Epidemiology. 2004;15:222–228. 31. Albert CM, Manson JE, Cook NR, Ajani UA, Gaziano JM, Hennekens CH. Moderate alcohol consumption and the risk of sudden cardiac death among US male physicians. Circulation. 1999;100:944–950. 32. Fuchs FD, Chambless LE, Folsom AR, et al. Association between alcoholic beverage consumption and incidence of coronary heart disease in whites and blacks. Am J Epidemiol. 2004;160:466–474. 33. Janszky I, Mukamal KJ, Orth-Gomér K, et al. Alcohol consumption and coronary atherosclerosis progression—the Stockholm Female Coronary Risk Angiographic Study. Atherosclerosis. 2004;176:311–319. 34. Klatsky AL, Chartier D, Udaltsova N, et al. Alcohol drinking and risk of hospitalization for heart failure with and without associated coronary artery disease. Am J Cardiol. 2005; 96:346–351. 35. Mehlig K, Skoog I, Guo X, et al. Alcoholic beverages and irks of dementia: 34-year follow-up of the Prospective Population Study of Women in Göteborg. Am J Epidemiol. 2008;167: 684–691. 36. Rehm J, Gmel G, Sempos CT, Trevisan M. Alcohol-related morbidity and mortality. Alcohol Res Health. 2002;27:39–51. 37. Marmot M, Brunner E. Alcohol and cardiovascular disease: the status of the U-shaped curve. Br Med J. 1991;303:565–568.
28. Marques-Vidal P, Montaye M, Arveiler D, et al. Alcohol consumption and cardiovascular disease: differential effects in France and Northern Ireland. The PRIME Study. Eur J Cardiovasc Prev Rehab. 2004;11:336–343.
38. Mukamal KJ, Ascherio A, Mittleman MA, et al. Alcohol and risk of ischemic stroke in men: The role of drinking patterns and usual beverage. Ann Intern Med. 2005;142:11–19.
29. Mukamal KJ, Conigrave KM, Mittleman MA, et al. Roles of drinking pattern and type of al-
39. Jernigan DH, Monteiro M, Room R, Saxena S. Towards a global alcohol policy: alcohol, pub-
46894_CH15_431_502.qxd
2/26/10
12:31 PM
Page 493
REFERENCES 493
lic health and the role of WHO. Bull World Health Organ. 2000;78:491–499. 40. Cochrane J, Chen H, Conigrave KM, Hao W. Alcohol use in China. Alcohol Alcohol. 2003; 38:537–542. 41. Das SK, Balakrishnan V, Vasudevan DM. Alcohol: its health and social impact in India. Natl Med J India. 2006;19:94–99.
49. Gans JE, Shook KL, eds. Policy Compendium on Tobacco, Alcohol, and Other Harmful Substances Affecting Adolescents: Alcohol and Other Harmful Substances. Chicago: American Medical Association; 1994. 50. Sensible Drinking: The Report of an InterDepartmental Working Group. London: Department of Health; 1995.
42. The Lancet. Alcohol and harm reduction in Russia. Lancet. 2009;373:2171.
51. Brand DA, Saisana M, Rynn LA, Pennoni F, Lowenfels AB. Comparative analysis of alcohol control policies in 30 countries. PLoS Med. 2007;4:0752–0759.
43. Rehm J, Chisholm D, Room R, Lopez AD. Alcohol. In: Jamison DT, Breman JG, Measham AR, et al., eds. Disease Control Priorities in Developing Countries. 2nd ed. Washington, DC: International Bank for Reconstruction and Development/The World Bank; 2006:887–906.
52. World Health Organization. Global Status Report on Alcohol 2004. Department of Mental Health and Substance Abuse. Geneva: World Health Organization, 2004.
44. Rehm J, Mathers C, Popova S, Thavorncharoensap M, Teerawattananon Y, Patra J. Global burden of disease and injury and economic cost attributable to alcohol use and alcohol-use disorders. Lancet. 2009;373: 2223–2233. 45. Room R, Babor T, Rehm J. Alcohol and public health. Lancet. 2005;365:519–530. 46. US Department of Health and Human Services. Agency for Healthcare Research and Quality. US Preventive Services Task Force. The Guide to Clinical Preventive Services 2006. Recommendations of the U.S. Preventive Services Task Force. AHRQ Publication No. 06-0588. Washington, DC: Agency for Healthcare Research and Quality; 2006. http://www .ahrq.gov/clinic/uspstf/uspstbac.htm. Accessed October 14, 2007.
53. Heather N, ed. WHO Collaborative Project on Identification and Management of AlcoholRelated Problems in Primary Health Care. Report on Phase IV. Development of CountryWide Strategies for Implementing Early Identification and Brief Intervention in Primary Health Care. Department of Mental Health and Substance Abuse. Geneva: World Health Organization; 2006. 54. World Health Organization. WHO Expert Committee on Problems Related to Alcohol Consumption. Second Report. Technical Report Series 944. Geneva: World Health Organization; 2007. 55. Marmot M. A not-so-sensible drinks policy. Lancet. 1995;346:1643–1644. 56. Rose G. Sick individuals and sick populations. Int J Epidemiol. 1985;14:32–38.
47. Guide to Community Preventive Services. Preventing excessive alcohol use. www .thecommunityguide.org/alcohol.html. Last updated February 10, 2009.
57. Casswell S, Thamarangsi T. Reducing harm from alcohol: call to action. Lancet 2009;373:2247–2257.
48. Addiction Research Foundation of Ontario and the Canadian Centre on Substance Abuse. Appendix 1—moderate drinking and health. Can Med Assoc J. 1994;151:13–16.
2. Adverse Psychosocial Factors 58. Hemingway H, Marmot M. Evidence based cardiology: psychosocial factors in the aetiology and prognosis of coronary heart disease.
46894_CH15_431_502.qxd
494
2/26/10
12:31 PM
Page 494
CHAPTER 15 OTHER PERSONAL FACTORS
Systematic review of prospective cohort studies. BMJ. May 29 1999;318(7196):1460–1467. 59. Theorell T. The psycho-social environment, stress, and coronary heart disease. In: Marmot M, Elliott P, eds. Coronary Heart Disease Epidemiology: From Aetiology to Public Health. Oxford, England: Oxford University Press; 1992:256–273. 60. Rozanski A, Blumenthal JA, Davidson KW, Saab PG, Kubzansky L. The epidemiology, pathophysiology, and management of psychosocial risk factors in cardiac practice: the emerging field of behavioral cardiology. J Am Coll Cardiol. Mar 1 2005;45(5):637–651. 61. Matthews KA. Psychological perspectives on the development of coronary heart disease. Am Psychol. Nov 2005;60(8):783–796. 62. Cassel J. Psychosocial process and “stress”: theoretical formulation. Int J Health Serv. 1974;4:471–482. 63. Matsumoto YS. Social stress and coronary heart disease in Japan. Milbank Memorial Fund Quar Bull. 1979;48:9–36. 64. Everson-Rose SA, Lewis TT. Psychosocial factors and cardiovascular diseases. Annu Rev Public Health. 2005;26:469–500. 65. US Department of Health and Human Services. Health, United States, 2008 with Special Feature on the Health of Young Adults. Washington DC: US Department of Health and Human Services. Centers for Disease Control and Prevention. National Center for Health Statistics; 2008. 66. House JS. Occupational stress and coronary heart disease: a review and theoretical integration. J Health Soc Behav. 1974;15:12–26. 67. Jenkins CD. Appraisal and implications for theoretical development. Milbank Memorial Fund Quar Bull. 1967;45:141–150. 68. Bhattarchayya MR, Steptoe A. Emotional triggers of acute coronary syndromes: strength of evidence, biological processes, and clinical implications. Prog Cardiovasc Dis. 2007;49: 353–365.
69. Krantz DS, McCeney MK. Effects of psychological and social factors on organic disease: a critical assessment of research on coronary heart disease. Annu Rev Psychol. 2002;53: 341–369. 70. Bairey Merz CN, Dwyer J, Nordstrom CK, Walton KG, Salerno JW, Schneider RH. Psychosocial stress and cardiovascular disease: pathophysiological links. Behav Med. Winter 2002;27(4):141–147. 71. Levine DM. Behavioral and psychosocial factors, processes, and strategies. In: Pearson TA, Criqui MH, Luepker RV, Oberman A, Winston M, eds. Primer in Preventive Cardiology. Dallas, TX: American Heart Association; 1994:217–226. 72. Friedman M, Rosenman RH. Association of specific overt behavior pattern with blood and cardiovascular findings. JAMA. 1959;169: 1286–1296. 73. Rosenman RH, Brand RJ, Sholtz RI, Friedman M. Multivariate prediction of coronary heart disease during 8.5 year follow-up in the Western Collaborative Group Study. Am J Cardiol. 1976;37:903–910. 74. Friedman M, Thoresen CE, Powell LH, et al. Alteration of Type A behavior and reduction in cardiac recurrences in postmyocardial infarction patients. Am Heart J. 1984;108: 237–248. 75. Shekelle RB, Hulley SB, Neaton JD, et al., Multiple Risk Factor Intervention Trial Research Group. The MRFIT behavior pattern study, II: type A behavior and incidence of coronary heart disease. Am J Epidemiol. 1985;122:559–570. 76. Ragland DR, Brand RJ. Type A behavior and mortality from coronary heart disease. N Engl J Med. 1988;318:65–69. 77. Barefoot JC, Dahlstrom G, Williams RB Jr. Hostility, CHD incidence, and total mortality: a 25-year follow-up study of 255 physicians. Psychosom Med. 1983;45:59–63. 78. Mittleman MA, Maclure M, Sherwood JB, et al. Triggering of acute myocardial infarction
46894_CH15_431_502.qxd
2/26/10
12:31 PM
Page 495
REFERENCES 495
onset by episodes of anger. Circulation. 1995; 92:1720–1725. 79. Jenkins CD. Recent evidence supporting psychologic and social risk factors for coronary disease. (First of two parts.) N Engl J Med. 1976;294:987–994. 80. Jenkins CD. Recent evidence supporting psychological and social risk factors for coronary disease. (Second of two parts.) N Engl J Med. 1976;294:1033–1038. 81. Dembroski TM, Weiss SM, Shields JL, Haynes SG, Feinleib M, eds. Coronary-Prone Behavior. New York: Springer-Verlag; 1978. 82. Yan LL, Liu K, Matthews KA, Daviglus ML, Ferguson TF, Kiefe CI. Psychosocial factors and risk of hypertension. The Coronary Artery Risk Development in Young Adults (CARDIA) Study. JAMA. 2003;290: 2138–2148. 83. Grunbaum J, Vernon SW, Clasen CM. The association between anger and hostility and risk factors for coronary heart disease in children and adolescents: a review. Ann Behav Med. 1997; Spring;19(2):179–189. 84. Aronowitz RA. Making Sense of Illness: Science, Society, and Disease. Cambridge, UK: Cambridge University Press; 1998. 85. Alfredsson L, Karasek R, Theorell T. Myocardial infarction risk and psychosocial work environment: an analysis of the male Swedish working force. Soc Sci Med. 1982;16:463–467. 86. Karasek RA, Theorell T, Schwartz JE, et al. Job characteristics in relation to the prevalence of myocardial infarction in the US Health Examination Survey (HES) and the Health and Nutrition Examination Survey (HANES). Am J Public Health. 1988;78: 910–918. 87. Johnson JV, Hall EM, Theorell T. Combined effects of job strain and social isolation on cardiovascular disease morbidity and mortality in a random sample of the Swedish male working population. Scand J Work Environ Health. 1989;15:271–279. 88. Rosvall M, Ostergren PO, Hedblad B, Isacsson SO, Janzon L, Berglund G. Work-related psy-
chosocial factors and carotid atherosclerosis. Int J Epidemiol. Dec 2002;31(6):1169–1178. 89. Kivimäki M, Hintsanen M, KeltikangasJärvinen L, et al. Early risk factors, job strain, and atherosclerosis among men in their 30s: The Cardiovascular Risk in Young Finns Study. Am J Public Health. 2007;97:450–452. 90. Frasure-Smith N, Lesperance F. Reflections on depression as a cardiac risk factor. Psychosom Med. May–Jun 2005;67(suppl 1):S19–S25. 91. Lett HS, Blumenthal JA, Babyak MA, et al. Depression as a risk factor for coronary artery disease: evidence, mechanisms, and treatment. Psychosom Med. May–Jun 2004;66(3):305–315. 92. Rugulies R. Depression as a predictor for coronary heart disease. A review and meta-analysis. Am J Prev Med. Jul 2002;23(1): 51–61. 93. Stansfeld S, Rasul F. Mental illness and coronary heart disease. In: Marmot M, Elliott P, editors. Coronary Heart Disease Epidemiology: From Aetiology to Public Health. 2nd ed. Oxford (England): Oxford University Press; 2005:414–431. 94. Murray CJL, Lopez AD. Alternative visions of the future: Projecting mortality and disability, 1990–2020. In: Murray CJL, Lopez AD, eds. The Global Burden of Disease: A Comprehensive Assessment of Mortality and Disability from Diseases, Injuries, and Risk Factors in 1990 and Projected to 2020. Boston, MA: The Harvard School of Public Health; 1996: 325–395. 95. Mathers CD, Lopez AD, Murray CJL. The burden of disease and mortality by condition: data, methods and results for 2001. In: AD Lopez et al., eds. Global Burden of Disease and Risk Factors. The International Bank for Reconstruction and Development/The World Bank, Washington, DC; 2006:45–240. 96. Hyman S, Chisholm D, Kessler R, Patel V, Whiteford H. Mental disorders. In: Jamison DT, Breman JG, Measham AR, et al., eds. Disease Control Priorities in Developing Countries. 2nd ed. Washington, DC: International Bank for Reconstruction and Development/The World Bank; 2006:605–625.
46894_CH15_431_502.qxd
496
2/26/10
12:31 PM
Page 496
CHAPTER 15 OTHER PERSONAL FACTORS
97. Berkman LF. Assessing the physical health effects of social networks and social support. Ann Rev Public Health. 1984;5:413–432.
Care and Outcomes Research. Circulation. 2008;118:1768–1775.
98. House JS, Landis KR, Umberson D. Social relationships and health. Science. 1988;241: 540–545.
106. Rees K, Bennett P, West R, Davey SG, Ebrahim S. Psychological interventions for coronary heart disease. Cochrane Database Syst Rev. 2004(2):CD002902.
99. Reed D, McGee D, Yano K. Psychosocial processes and general susceptibility to chronic disease. Am J Epidemiol. 1984;119:356–370.
107. Cohen S, Syme SL, eds. Social Support and Health. Orlando, FL: Academic Press, Inc; 1985.
100. Farmer IP, Meyer PS, Ramsey DJ, et al. Higher levels of social support predict greater survival following acute myocardial infarction: the Corpus Christi Heart Project. Behav Med. 1996;22:59–66.
108. Eriksen W. The role of social support in the pathogenesis of coronary heart disease: a literature review. Fam Prac. 1994;11:201–209.
3. Hemostatic Factors 101. Lett HS, Blumenthal JA, Babyak MA, Strauman TJ, Robins C, Sherwood A. Social support and coronary heart disease: epidemiologic evidence and implications for treatment. Psychosom Med. Nov–Dec 2005; 67(6):869–878. 102. Rosengren A, Hawken S, Ounpuu S, et al. Association of psychosocial risk factors with risk of acute myocardial infarction in 11119 cases and 13648 controls from 52 countries (the INTERHEART study): case-control study. Lancet. Sep 11–17 2004;364(9438): 953–962. 103. Sheps DS, Frasure-Smith N, Freedland KE, Carney RM. The INTERHEART Study: intersection between behavioral and general medicine. Psychosom Med. 66;2004: 797–798. 104. Graves KD, Miller PM. Behavioral medicine in the prevention and treatment of cardiovascular disease. Behav Modif. Jan 2003;27(1): 3–25. 105. Lichtman JH, Bigger Jr JT, Blumenthal JA, et al. Depression and coronary heart disease. Recommendations for screening, referral, and treatment. A Science Advisory from the American Heart Association Prevention Committee of the Council on Cardiovascular Nursing, Council on Clinical Cardiology, Council on Epidemiology and Prevention, and Interdisciplinary Council on Quality of
109. Bithell TC. Platelets, hemostasis, and coagulation. In: Lee RG, Bithell TC, Foerster J, Athens JW, eds. Wintrobe’s Clinical Hematology. 9th ed. Philadelphia, PA: Lea & Febiger; 1993;1:509. 110. Bithell TC. Blood coagulation. In: Lee RG, Bithell TC, Foerster J, Athens JW, eds. Wintrobe’s Clinical Hematology. 9th ed. Philadelphia, PA: Lea & Febiger; 1993; 1:566–615. 111. Bithell TC. The diagnostic approach to the bleeding disorders. In: Lee RG, Bithell TC, Foerster J, Athens JW, eds. Wintrobe’s Clinical Hematology. 9th ed. Philadelphia, PA: Lea & Febiger; 1993;2:1301–1324. 112. Chambless LE, McMahon R, Wu K, Folsom A, et al. Short-term intraindividual variability in hemostasis factors: the ARIC Study. Ann Epidemiol. 1992;2:723–733. 113. Bithell TC. Thrombosis and antithrombotic therapy. In: Lee RG, Bithell TC, Foerster J, Athens JW, eds. Wintrobe’s Clinical Hematology. 9th ed. Philadelphia, PA: Lea & Febiger; 1993;2:1515–1551. 114. Catena C, Novello M, Lapenna R, et al. New risk factors for atherosclerosis in hypertension: focus on the prothrombotic state and lipoprotein(a). J Hypertens. Sep 2005;23(9): 1617–1631.
46894_CH15_431_502.qxd
2/26/10
12:31 PM
Page 497
REFERENCES 497
115. Pearson TA, LaCava J, Weil HF. Epidemiology of thrombotic-hemostatic factors and their associations with cardiovascular disease. Am J Clin Nutr. May 1997;65(5 Suppl): 1674S–1682S. 116. Koenig W. Fibrin(ogen) in cardiovascular disease: an update. Thromb Haemost. Apr 2003;89(4):601–609. 117. Shah PK. Thrombogenic risk factors for atherothrombosis. Rev Cardiovasc Med. Winter 2006;7(1):10–16. 118. Folsom AR. Epidemiology of fibrinogen. Eur Heart J. 1995;16(suppl A):21–24. 119. Tousoulis D, Davies G, Ambrose J, Tentolouris C, Stefanadis C, Toutouzas P. Effects of lipids on thrombotic mechanisms in atherosclerosis. Int J Cardiol. Dec 2002;86(2-3):239–247. 120. Lefevre M, Kris-Etherton PM, Zhao G, Tracy RP. Dietary fatty acids, hemostasis, and cardiovascular disease risk. J Am Diet Assoc. 2004;104:410–419. 121. Wilson TW, Kaplan GA, Kauhanen J, et al. Association between plasma fibrinogen concentration and five socioeconomic indices in the Kuopio Ischemic Heart Disease Risk Factor Study. Am J Epidemiol. 1993;137: 292–300. 122. Keavney B, Danesh J, Parish S, et al. Fibrinogen and coronary heart disease: test of causality by “Mendelian randomization.” Int J Epidemiol. 2006;35:935–943. 123. Meade TW, North WRS. Population-based distributions of haemostatic variables. Br Med Bull. 1977;33:283–288. 124. Iso H, Folsom AR, Wu KK, et al. Hemostatic variables in Japanese and Caucasian men. Am J Epidemiol. 1989;130:925–934. 125. Meilahn EN, Kuller HL, Matthews KA, Kiss JE. Hemostatic factors according to menopausal status and use of hormone replacement therapy. Ann Epidemiol. 1992;2: 445–455.
126. Nabulsi AA, Folsom AR, White A, et al., for the Atherosclerosis Risk in Communities Study Investigators. Association of hormonereplacement therapy with various cardiovascular risk factors in postmenopausal women. N Engl J Med. 1993;328:1069–1075. 127. Fernandez-Ortiz A, Fuster V. The role of thrombosis in acute coronary heart disease. In: Willich SN, Muller JE, eds. Triggering of Acute Coronary Syndromes. Dordrecht (the Netherlands): Kluwer Academic Publishers; 1996:201–218. 128. Meade TW, Brozovic M, Chakrabarti RR, et al. Haemostatic function and ischaemic heart disease: principal results of the Northwick Park Heart Study. Lancet. 1986;2:533–537. 129. Folsom AR, Wu KK, Rosamond WD, Sharrett AR, et al. Prospective study of hemostatic factors and incidence of coronary heart disease: the Atherosclerosis Risk in Communities (ARIC) Study. Circulation. 1997;96: 1102–1108. 130. Ernst E, Resch KL. Fibrinogen as a cardiovascular risk factor: a meta-analysis and review of the literature. Ann Intern Med. 1993;118: 956–963. 131. Danesh J, Collins R, Appleby P, Peto R. Association of fibrinogen, C-reactive protein, albumin, or leukocyte count with coronary heart disease: meta-analyses of prospective studies. JAMA. May 13, 1998;279(18): 1477–1482. 132. Ridker PM, Vaughan DE, Stampfer MJ, et al. Endogenous tissue-type plasminogen activator and risk of myocardial infarction. Lancet. 1993;341:1165–1168. 133. Antiplatelet Trialists’ Collaboration. Collaborative overview of randomised trials of antiplatelet therapy I: prevention of death, myocardial infarction, and stroke by prolonged antiplatelet therapy in various categories of patients. Br Med J. 1994;308: 81–106. 134. Goff DC, Nichaman MZ, Ramsey DJ, et al. A population-based assessment of the use and
46894_CH15_431_502.qxd
498
2/26/10
12:31 PM
Page 498
CHAPTER 15 OTHER PERSONAL FACTORS
effectiveness of thrombolytic therapy: the Corpus Christi Heart Project. Ann Epidemiol. 1995;5:171–178. 135. Antithrombotic Trialists’ Collaboration. Collaborative meta-analysis of randomised trials of antiplatelet therapy for prevention of death, myocardial infarction, and stroke in high risk patients. BMJ. 2002;324:71–86. 136. US Department of Health and Human Services. Health, United States, 2008 with Special Feature on the Health of Young Adults. Washington DC: US Department of Health and Human Services. Centers for Disease Control and Prevention. National Center for Health Statistics. 2008. 137. Bartolucci AA, Howard G. Meta-analysis of data from the six primary prevention trials of cardiovascular events using Aspirin. Am J Cardiol. 2006;98:746–750. 138. Berger JS, Roncaglioni MC, Avanzini F, Pangrazzi I, Tognoni G, Brown DL. Aspirin for the primary prevention of cardiovascular events in women. A sex-specific meta-analysis of randomized controlled trials. JAMA. 2006;295:306–313. 139. Antithrombotic Trialists’ (ATT) Collaboration. Aspirin in the primary and secondary prevention of vascular disease: collaborative meta-analysis of individual participant data from randomised trials. Lancet. 2009;373: 1849–1860. 140. Bode C, ed. Antiplatelet therapy: current issues and new developments. Eur Heart J. 2007;9(suppl D):D1–D27. 141. Herlitz J. Fibrinolytic treatment in acute myocardial infarction: benefits and risks. In: Workshop: Thrombolytic Therapy in Myocardial Infarction. Oslo (Norway): Medical Products Agency, The Norwegian Medicines Control Authority; 1994:49–59. 142. The National Institute of Neurological Disorders and Stroke rt-PA Stroke Study Group. Tissue plasminogen activator for acute ischemic stroke. N Engl J Med. 1995;333:1581–1587.
4. Evolving and Emerging Factors 143. Howson CP, Reddy KS, Ryan TJ, Bale JR, eds. Control of Cardiovascular Diseases in Developing Countries. Research, Development, and Institutional Strengthening. Washington, DC: Institute of Medicine, National Academy of Sciences, National Academy Press; 1998. 144. Wald NJ, Law MR. A strategy to reduce cardiovascular disease by more than 80%. BMJ. 2003;326:1419–1424. 145. Combination Pharmacotherapy and Public Health Research Working Group. Combination pharmacotherapy for cardiovascular disease. Ann Intern Med. 2005;143:593–599. 146. The Indian Polycap Study (TIPS). Effects of a polypill (Polycap) on risk factors in middleaged individuals without cardiovascular disease (TIPS): a phase II, double-blind, randomised trial. Lancet. Published online March 30, 2009. doi:10.1016/S0140-6736 (09)60611-5. 147. Khaw K-T, Barrett-Connor E. Sex differences, hormones, and coronary heart disease. In: Marmot M, Elliott P, eds. Coronary Heart Disease Epidemiology: From Aetiology to Public Health. Oxford (England): Oxford University Press; 1992:274–286. 148. Nelson HD, Humphrey LL, Nygren P, Teutsch SM, Allan JD. Postmenopausal hormone replacement therapy. Scientific Review. JAMA. 2002;288:872–881. 149. Writing Group for the Women’s Health Initiative Investigators. Risks and benefits of estrogen plus progestin in healthy postmenopausal women. Principal results from the Women’s Health Initiative randomized controlled trial. JAMA. 2002;288:321–333. 150. Prentice RL, Langer R, Stefanick ML, et al., for the Women’s Health Initiative Investigators. Combined postmenopausal hormone therapy and cardiovascular disease: toward resolving the discrepancy between observational studies and the Women’s Health Initiative clinical trial. Am J Epidemiol. 2005;162:404–414.
46894_CH15_431_502.qxd
2/26/10
12:31 PM
Page 499
REFERENCES 499
151. Willett WC, Manson JE, Grodstein F, Stampfer MJ, Colditz GA. Re: “Combined postmenopausal hormone therapy and cardiovascular disease: toward resolving the discrepancy between observational studies and the Women’s Health Initiative clinical trial.” Am J Epidemiol. 2006;163:1067–1068. 152. Prentice RL, Stefanick ML, Howard BV, Barad D, Kuller L. Five authors reply. Am J Epidemiol. 2006;163:1068–1069. 153. Manson JE, Bassuk SS. Invited commentary: hormone therapy and risk of coronary heart disease––why renew the focus on the early years of menopause? Am J Epidemiol. 2007;166:511–517.
161. Vercellotti GM. Overview of infections and cardiovascular diseases. J Allergy Clin Immunol. 2001;108:S117–S120. 162. Simanek AM, Dowd JB, Aiello AE. Persistent pathogens linking socioeconomic position and cardiovascular disease in the US. Int J Epidemiol. 2009;38:775–787. 163. Madjid M, Miller CC, Zarubaev VV, et al. Influenza epidemics and acute respiratory disease activity are associated with a surge in autopsy-confirmed coronary heart disease death: results from 8 years of autopsies in 34,892 subjects. Eur Heart J. 2007;28: 1205–1210.
154. Barrett-Connor E. Hormones and heart disease in women: the timing hypothesis. Am J Epidemiol. 2007;166:506–510.
164. Smeeth L, Casa JP, Hingorani AD. The role of infection in cardiovascular disease: more support but many questions remain. Eur Heart J. 2007;28:1178–1179.
155. Garbe E, Suissa S. Issues to debate on the Women’s Health Initiative (WHI) study. Hormone replacement therapy and acute coronary outcomes: methodological issues between randomized and observational studies. Human Reprod. 2004;19:8–13.
165. Davis MM, Taubert K, Benin AL, et al. Influenza vaccination as secondary prevention for cardiovascular disease. A Science Advisory from the American Heart Association/ American College of Cardiology. Circulation. 2006;114:1549–1553.
156. US Department of Health and Human Services. Health, United States, 2008 with Special Feature on the Health of Young Adults. Washington DC: US Department of Health and Human Services. Centers for Disease Control and Prevention. National Center for Health Statistics; 2008.
166. Centers for Disease Control and Prevention. Prevention and control of influenza. Recommendations of the Advisory Committee on Immunization Practices (ACIP), 2007. MMWR. 2007;56 (No. RR-6): 1–54.
157. Leinonen M, Saikku P. Evidence for infectious agents in cardiovascular disease and atherosclerosis. Lancet Inf Dis. 2002;2:11–17. 158. Nieto FJ. Infective agents and cardiovascular disease. Sem Vasc Med. 2002;2:401–415. 159. Andrews R, Berger JS, Brown DL. Effects of antibiotic therapy on outcomes of patients with coronary artery disease. A meta-analysis of randomized controlled trials. JAMA. 2005;293:2641–2647. 160. O’Connor S, Taylor C, Campbell LA, Epstein S, Libby P. Potential infectious etiologies of atherosclerosis: a multifactorial perspective. Emerg Inf Dis. 2001;7:780-788.
167. Grinspoon SK, Grunfeld C, Kotler DP, et al. Initiative to decrease cardiovascular risk and increase quality of care for patients living with HIV/AIDS. Executive summary. Circulation. 2008;118:198–210. 168. Grunfeld C, Kotler DP, Arnett DK, et al. Contribution of metabolic and anthropometric abnormalities to cardiovascular disease risk factors. Circulation. 2008;118:e20–e28. 169. Currier JS, Lundgren JD, Carr A, et al. Epidemiological evidence for cardiovascular disease in HIV-infected patients and relationship to highly active antiretroviral therapy. Circulation. 2008;118:e29–e35. 170. Zou W, Berglund L. HIV and highly active antiretroviral therapy: dyslipidemia, metabolic
46894_CH15_431_502.qxd
500
2/26/10
12:31 PM
Page 500
CHAPTER 15 OTHER PERSONAL FACTORS
aberrations, and cardiovascular risk. Prev Cardiol. 2007;Spring:96–103. 171. Steinberg D, Parthasarathy S, Carew TE, et al. Beyond cholesterol: modifications of low-density lipoprotein that increase its atherogenicity. N Engl J Med. 1989;320: 915–924. 172. Steinberg D, National Heart, Lung and Blood Institute Workshop Participants. Antioxidants in the prevention of human atherosclerosis. Summary of the proceedings of a National Heart, Lung, and Blood Institute Workshop: September 5–6, 1991, Bethesda, Maryland. Circulation. 1992;85: 2338–2344. 173. Kritchevsky D. Antioxidant vitamins in the prevention of cardiovascular disease. Nutr Today. 1992;27:30–33. 174. Gey FK, Puska P, Jordan P, Moser UK. Inverse correlation between plasma vitamin E and mortality from ischemic heart disease in cross-cultural epidemiology. Am J Clin Nutr. 1991;53(suppl):326S–334S. 175. Stampfer MJ, Hennekens CH, Manson JE, Colditz GA, et al. Vitamin E consumption and the risk of coronary disease in women. N Engl J Med. 1993;328:1444–1449. 176. Rimm EB, Stampfer MJ, Ascherio A, et al. Vitamin E consumption and the risk of coronary heart disease in men. N Engl J Med. 1993;328:1450–1456. 177. Kushi LH, Folsom AR, Prineas RJ, et al. Dietary antioxidant vitamins and death from coronary heart disease in postmenopausal women. N Engl J Med. 1996;334:1156–1162. 178. Jha P, Flather M, Lonn E, et al. The antioxidant vitamins and cardiovascular disease: a critical review of epidemiologic and clinical trial data. Ann Intern Med. 1995;123: 860–872. 179. Stanner SA, Hughes J, Kelly CN, Buttriss J. A review of the epidemiological evidence for the “antioxidant hypothesis.” Public Health Nutr. May 2004;7(3):407–422.
180. Törnwall ME, Virtamo J, Korhonen PA, et al. Effect of -tocopherol and -carotene supplementation on coronary heart disease during the 6-year post-trial follow-up in the ATBC study. Eur Heart J. 2004;25:1171–1178. 181. The HOPE and HOPE-TOO Trial Investigators. Effects of long-term vitamin E supplementation on cardiovascular events and cancer. A randomized controlled trial. JAMA. 2005;293:1338–1347. 182. Balla G, Jacob HS, Eaton JW, et al. Hemin: a possible physiological mediator of low density lipoprotein oxidation and endothelial injury. Arterioscler Thromb. 1991;11: 1700–1711. 183. Salonen JT, Nyyssönen K, Korpela H, et al. High stored iron levels are associated with excess risk of myocardial infarction in Eastern Finnish men. Circulation. 1992;86:803–811. 184. Liao Y, Cooper RS, McGee DL. Iron status and coronary heart disease: negative findings from the NHANES I Epidemiologic FollowUp Study. Am J Epidemiol. 1994;139: 704–712. 185. Malinow MR. Homocyst(e)ine and arterial occlusive diseases. J Intern Med. 1994;236: 603–617. 186. Furie KL, Kelly PJ. Homocyst(e)ine and stroke. Sem Neurol. 2006;26:24–32. 187. The Homocysteine Studies Collaboration. Homocysteine and risk of ischemic heart disease and stroke. A meta-analysis. JAMA. 2002;288:2015–2022. 188. Boushey CJ, Beresford SAA, Omenn GS, Motulsky AG. A quantitative assessment of plasma homocysteine as a risk factor for vascular disease. JAMA. 1995;274:1049–1057. 189. Jacques PF, Bostom AG, Williams RR, et al. Relation between folate status, a common mutation in methylenetetrahydrofolate reductase, and plasma homocysteine concentrations. Circulation. 1996;93:7–9.
46894_CH15_431_502.qxd
2/26/10
12:31 PM
Page 501
REFERENCES 501
190. The Heart Outcomes Prevention Evaluation (HOPE) 2 Investigators. Homocysteine lowering with folic acid and B vitamins in vascular disease. N Engl J Med. 2006;354: 1567–1577. 191. Bønaa KH, Njølstad I, Ueland PM, et al. Homocysteine lowering and cardiovascular events after acute myocardial infarction. N Engl J Med. 2006;354:1578–1588. 192. Smith GD, Shah E. Folate supplementation and cardiovascular disease. Lancet. 2005; 366:1679–1681. 193. Hansson GK. Inflammation, atherosclerosis, and coronary artery disease. N Engl J Med. 2005;352:1685–1695. 194. Pearson TA, Mensah GA, Alexander RW, et al. Markers of inflammation and cardiovascular disease. Application to clinical and public health practice. A statement for healthcare professionals from the Centers for Disease Control and Prevention and the American Heart Association. Circulation. 2003;107: 499–511. 195. Myers GL, Christenson RHM, Cushman M, et al. National Academy of Clinical Biochemistry Laboratory Medicine Practice Guidelines: emerging biomarkers for primary prevention of cardiovascular disease. Clin Chem. 2009;55:378–384. 196. Ridker PM, Rifai N, Rose L, Buring JE, Cook NR. Comparison of C-reactive protein and low-density lipoprotein cholesterol levels in the prediction of first cardiovascular events. N Engl J Med. 2002;347:1557–1565. 197. Miller M, Zhan M, Havas S. High attributable risk of elevated C-reactive protein level to conventional coronary heart disease risk factors: the Third National Health and Nutrition Examination Survey. Arch Intern Med. Oct 10 2005;165(18):2063–2068. 198. Lambert M, Delvin EE, Paradis G, O’Loughlin J, Hanley JA, Levy E. C-reactive protein and features of the metabolic syndrome in a population-based sample of children and adolescents. Clin Chem. 2004;50:1762–1768.
199. National Heart, Lung and Blood Institute. NHLBI Workshop Report. C-reactive protein: basic and clinical research needs. July 10–11, 2006. http://www.nhlbi.nih.gov/meetings/ workshops/crp/report.htm. Accessed June 26, 2007. 200. Ridker PM, Danielson E, Fonseca FAH, et al., for the JUPITER Study Group. Rosuvastatin to prevent vascular events in men and women with elevated C-reactive protein. N Engl J Med. 2008;359:2195–2207. 201. Spatz E, Canavan ME, Desai MM. From here to JUPITER. Identifying new patients for statin therapy using data from the 1999–2004 National Health and Nutrition Examination Survey. Circ Cardiovasc Qual Outcomes. 2009;2:41–48. 202. Kannel WB, Dawber TR, Kagan A, et al. Factors of risk in the development of coronary heart disease—six-year follow-up experience: the Framingham Study. Ann Intern Med. 1961;55:33–50. 203. Hopkins PN, Williams RR. A survey of 246 suggested coronary risk factors. Atherosclerosis. 1981;40:1–52. 204. Grundy SM, Pasternak R, Greenland P, Smith Jr S, Fuster V. Assessment of cardiovascular risk by use of multiple-risk-factor assessment equations. A statement for healthcare professionals from the American Heart Association and the American College of Cardiology. Circulation. 1999;100:1481–1492. 205. Kullo IJ, Ballantyne CM. Conditional risk factors for atherosclerosis. Mayo Clin Proc. Feb 2005;80(2):219–230. 206. Harjai KJ. Potential new cardiovascular risk factors: left ventricular hypertrophy, homocysteine, lipoprotein(a), triglycerides, oxidative stress, and fibrinogen. Ann Intern Med. Sep 7 1999;131(5):376–386. 207. Hackam DG, Anand SS. Emerging risk factors for atherosclerotic vascular disease: a critical review of the evidence. JAMA. Aug 20 2003;290(7):932–940.
46894_CH15_431_502.qxd
502
2/26/10
12:31 PM
Page 502
CHAPTER 15 OTHER PERSONAL FACTORS
208. Wilson PW. Assessing coronary heart disease risk with traditional and novel risk factors. Clin Cardiol. Jun 2004;27(6 suppl 3): III7–III11. 209. Oliveira GH. Novel serologic markers of cardiovascular risk. Curr Atheroscler Rep. Mar 2005;7(2):148–154. 210. Wang TJ, Gona P, Larson MG, et al. Multiple biomarkers for the prediction of first major cardiovascular events and death. N Engl J Med. 2006;355:2631–2639.
211. Folson AR, Chambless LE, Ballantyne CM, et al. An assessment of incremental coronary risk prediction using C-reactive protein and other novel risk markers. The Atherosclerosis Risk in Communities Study. Arch Intern Med. 2006;166:1368–1373. 212. Hlatky MA, Greenland P, Arnett DK, et al. Criteria for evaluation of novel markers of cardiovascular risk. A scientific statement from the American Heart Association. Circulation. 2009 May 5;119(17): 2408–2416.
46894_CH16_503_532.qxd
2/26/10
1:00 PM
Page 503
C H A P T E R
16 Social and Physical Environment The concept of social determinants of health has received new prominence under a WHO Commission and is stimulating a new level of attention to health inequities within and across populations. What these several concepts mean, in relation to the personal factors in cardiovascular disease discussed throughout Chapters 7–15, presented previously, is a question to be addressed in terms of theory, practice, and research. These are the current issues that close this chapter and Part III.
SUMMARY The social environment and physical environment are distinct from all the determinants considered in the preceding chapters, which are personal characteristics. Aspects of the social environment include social status (e.g., occupational class, income and education, disadvantage, and social status) and social conditions and their change over time (e.g., changes in social circumstances over the life course and societal changes over time). The physical environment viewed narrowly presents specific exposures (e.g., second-hand smoke or particulate air pollution), but more broadly includes, for example, characteristics of community design (e.g., food availability or opportunities for physical activity). Social conditions, as properties of societies or populations rather than of individuals, have been a subject of theory and research in cardiovascular epidemiology for several decades. Work in this area was stimulated especially by several writers around the early 1960s. Culture change or cultural mobility, occupational status and social class, and measures of education, income, and income distribution within societies have all been studied. Especially in the past two decades, attention has included aspects of fetal and early postnatal development and their relation to risks of cardiovascular diseases in adulthood, with the implication that social conditions affecting this critical period may have lifelong health effects. Other examples of social conditions or change in social circumstances include long-term shifts in demography and culture, such as the epidemiologic transition, Westernization, or globalization, all of which have been considered as influences on population health and specifically on cardiovascular disease.
INTRODUCTION “Environment” can be understood in many ways, each with its own relevance to individual and population health. Perhaps the broadest sense is found in the commonplace distinction between “genes and (or versus) environment,” suggesting that “environment” is everything that is not human DNA. By contrast, the narrowest sense of “environment” may be in reference to a specific physical agent, such as second-hand tobacco smoke, exposure to which is associated with increased risk of disease. In the context of cardiovascular epidemiology and prevention, it is useful to consider “environment” as embracing all of the conditions under which people live that are essentially external to the individual. Dividing environment in this sense into its social and physical components, within which specific factors can be identified, is helpful for purposes of discussion. Of course this distinction is not absolute, given that each of these components may be influenced in part by the other. “Social environment” is a term intended to denote influences on cardiovascular health beyond personal, or individual-level, characteristics, referring
503
46894_CH16_503_532.qxd
504
2/26/10
1:00 PM
Page 504
CHAPTER 16 SOCIAL AND PHYSICAL ENVIRONMENT
distinctly to attributes of groups or populations. This distinction is in some instances imperfect, as illustrated by the demographic composition of a population by age, sex, and race/ethnicity or the population distribution of years of school completed—each being derived by aggregating data on one or more individual attributes to construct a population measure. A different type of social indicator is the income gradient within a population, such as the proportion of the population at each of several levels of poverty. This is more readily recognizable as a measure of the community or population that has no equivalent at the individual level, although this, too, is based on the household size and income of individuals. Social organizations, from neighborhood level to the global community, more clearly transcend individual-level characteristics, having such properties as size, composition, purpose, and influence that are uniquely their own. Further attributes of a community or society, rather than of its members as individuals, are policies, regulations, or laws that apply to a population or to particular groups within it. Several aspects of the social environment have been noted in preceding chapters, in which they contribute to the presence of individual-level determinants of cardiovascular risk. One such example is the hierarchy of multiple factors that influence food choice and thereby largely determine dietary patterns, as illustrated in Figure 8-6. The social environment is also the “social” in “psychosocial,” discussed in the preceding chapter. There, attention was focused on psychological states or responses of individuals to social factors, or stressors, in causing disease. In the present chapter, social conditions themselves are the focus, although in both contexts the mechanisms that link them with disease states in individuals are at least implicitly at issue. Currently in the forefront of discussion of the social environment is the concept of “social determinants of health.” In this connection, the World Health Organization (WHO) created the Commission on Social Determinants of Health (CSDH) in 2005, with the following goals:1, p 3 • to support health policy change in countries by assembling and promoting effective, evidence-based models and practices that address the social determinants of health; • to support countries in placing health equity as a shared goal to which many government departments and sectors of society contribute; • to help build a sustainable global movement for action on health equity and social determinants, linking governments, international organizations, research institutions, civil society and communities.
The stated premise of the Commission’s work indicates the breadth of their concept of social determinants of health:1, p 1 “The conditions in which people live and work can help to create or destroy their health––lack of income, inappropriate housing, unsafe workplaces, and lack of access to health systems are some of the social determinants of health leading to inequalities within and between countries.” The distinction between social and physical environment is blurred here, with housing and the work environment subsumed under social determinants. Sir Professor Michael Marmot, Chair of the Commission, stated:1, p 3 “At the core of the Commission’s work is the belief that a society that has organized its social conditions so that its population has better health is a better society. Health is a measure of the degree to which the society delivers a good life to its citizens.” This viewpoint is remarkably consistent with the mission of public health, as defined 20 years earlier by the Institute of Medicine Committee for the Study of the Future of Public Health:2, p 7 “The committee defines the mission of public health as fulfilling society’s interest in assuring conditions in which people can be healthy.” Society is interested in and accountable for the conditions in which people live, evaluated in terms of population health, and the mission of public health is to assure these conditions. This proposition indicates an inherent link between public health and the social environment that is directly relevant to epidemiology and prevention of cardiovascular diseases. “Physical environment,” in this context, comprises factors across a spectrum from specific agents of disease (e.g., dietary trans fat), to specific characteristics of neighborhoods (e.g., accessibility to outdoor spaces for physical activity), to general attributes of place (e.g., geographic location). Exposures to these factors may be universal within a community or population or may be differentially distributed, for example, in accordance with socioeconomic status or social rank. Again, the distinction between social and physical environment as determinants of health is not absolute. There are, however, attributes specific to the physical environment that contribute to cardiovascular risk, such as second-hand smoke, discussed in some detail in Chapter 14, “Smoking and Other Tobacco Use,” and particulate air pollution, reviewed as follows. Less readily classified are some cited “environmental” components of gene-environment interactions (e.g., alcohol intake and the effect of cholesteryl ester transfer protein, CETP, on HDLcholesterol concentration—see Chapter 7, “Genes and Environment”). This perspective on the social and physical environment can be illustrated through discussion of the following topics as they relate to cardiovascular dis-
46894_CH16_503_532.qxd
2/26/10
1:00 PM
Page 505
SOCIAL STATUS 505
eases: social status, including socioeconomic status; changes in social conditions, whether for individuals over the life course or in society at large over time; environmental hazards, such as particulate air pollution; and neighborhood characteristics as the source of environmental barriers and opportunities for promoting cardiovascular health.
SOCIAL STATUS Among a small group of epidemiologists with early interest in health aspects of the social environment was Hinkle, who wrote in 1961:3, p 290 One cannot doubt the need for studies of health and the social milieu. Questions of the relations between illness and social class, economic conditions, migration, social mobility, status change, acculturation, and similar social phenomena are pressing, and the methods of the social scientist readily lend themselves to their investigation. Nevertheless, it may be predicted that the answers that are obtained will be complex and that it will be much easier to see the application of these to the sample and to the circumstances under study than to extrapolate them to any large segment of mankind, or to any general class of social phenomena. Hinkle’s statement added impetus to epidemiologic study of social conditions. But it also expressed reservations about whether any general inferences could be reached from this research, which might be limited to only isolated observations in particular populations under momentary circumstances. Occupational Class Early Studies Occupation has been addressed in preceding chapters in relation to physical activity, or lack thereof, and to psychosocial stress resulting from “job strain.” Here occupation is considered rather as a marker of social status or social class. The occupational situation in this respect was the primary focus of work by Reeder and by Hinkle and colleagues, but from different perspectives.4,5 Reeder reviewed studies in which the status aspect of occupation or socioeconomic level was emphasized. He found no coherent theoretical framework guiding these studies, but their findings were generally consistent in that higher occupational or socioeconomic status was associated with increased frequency of coronary heart disease.
Hinkle and colleagues studied health records based on the experience of 270,000 men in the Bell System Operating Companies in the early 1960s and investigated the relation between educational background (with or without college degree), occupational level (from executives to skilled workers, on a sevenpoint scale), and incidence of coronary heart disease. They found prior education to be more strongly related to risk than job classification. Higher educational level was interpreted as reflecting more favorable opportunities and conditions of life through the school years, rather than education itself, as the operative influence on risk. The observation that factors determining risk in later adult years were better explained by differences in level of education than by occupational level led Hinkle and colleagues to conclude that “some aspects of the origin of coronary heart disease must be sought for in childhood or adolescence, if not earlier.”5, p 244 Social Class in England and Wales Occupation has been studied extensively in England and Wales, where the Registrar General’s “social class” is based on this characteristic. Classes range from professionals and certain others (Class I) to unskilled occupations (Class V). A 1987 review by Marmot and colleagues of social class and health emphasized the importance of social forces, such as those represented by social class, as they operate through variation in lifestyles and specific exposures to produce differences in health.6 Their thesis was that the basis for associations between social class and health status was unlikely to be understood completely; therefore, intervention to modify particular risk factors identified with a given class would be correspondingly incomplete, overlooking factors inherent in social class that may be no less important in causing disease. These considerations may be particularly relevant to understanding the relation between social class and mortality from heart disease in England and Wales, which changed historically between 1951 and 1971. The category of “nonvalvular heart disease” was defined to be consistent in meaning across several revisions of the ICD code over this period. Over these 20 years, rates increased sharply for the lowest social classes, IV and V, and surpassed the modest continuing increase for highest classes, I and II. This cause of death was more frequent among the higher social classes in 1951, but the relative frequencies between classes were reversed within one decade, and the trend continued over the next one. Notably, these changes occurred only among men. For women, the rates were higher for the lower classes from the beginning of this period, with no change in relative frequencies.
46894_CH16_503_532.qxd
506
2/26/10
1:00 PM
Page 506
CHAPTER 16 SOCIAL AND PHYSICAL ENVIRONMENT
This analysis of cardiovascular mortality by social class thus revealed significant changes in the natural history of coronary heart disease—or in social class itself, as a determinant of lifestyles and exposures— differences in this relation between men and women. This change in the relationship between social class and coronary mortality had been examined in greater detail in an earlier report.7 Some intermediate factors between social class and coronary mortality— specifically, dietary intakes of fat, refined sugar and fiber, and smoking behavior—were investigated within the limitations of data available at the national level over the 40-year period of interest, 1931 to 1971. The shift to relative dominance of the lower social classes (IV and V) over the higher classes (I and II) in coronary mortality in the more recent years could not be explained by differential changes between the classes. Fat intake changed similarly in the upper and lower classes with no net difference in change over time. There were relative changes in intake of both fiber and refined sugars, and both were correlated with changing mortality. But because they were strongly correlated with each other (inversely), no separate effects could be evaluated. A relative decrease in smoking in the higher classes did explain a part of the relative difference in coronary mortality trends over this period. It was observed that a pattern of decreasing coronary mortality, experienced earlier in the United States and Australia, had begun in Great Britain with greater benefit for the upper classes. As suggested earlier by Hinkle, by Cassel, and by others, social conditions may relate to disease occurrence in a complex manner, depending on circumstances that may not be completely identified or understood. Occupational Class in Europe Kunst and others, as part of the EU Working Group on Socioeconomic Inequalities in Health, compared cause-specific mortality by occupational class among men aged 40–59 years at death in 11 Western European countries, from Finland in the north to Italy in the south.8 The data were obtained from longitudinal studies in some countries and cross-sectional studies in others and reflect experience of the 1980s. Occupational classes defined so as to be comparable across countries were reduced to three: nonmanual, including self-employment; manual; and farmers and farm laborers. For several causes of death, the rate ratio for manual versus nonmanual classes is shown, by country, in Table 16-1. For ischemic heart disease, mortality ratios significantly greater than 1 were observed for most, but not all countries, ranging from 1.23 to 1.47. The ratio was elevated but not
significantly so for France and was inverse for Switzerland, Spain, and Portugal. Thus, the manual class excess is not universal. By contrast, for cerebrovascular disease the mortality ratios were significantly increased in every country, without exception. Generally, the greatest excesses were for respiratory disease, but the high frequency of death coupled with its high mortality ratios resulted in ischemic heart disease contributing most to the overall difference in mortality between manual and nonmanual classes in the northern, but not the southern, countries. Income and Education In a 1995 conference on socioeconomic status and cardiovascular health and disease, several presentations addressed background data (principally for the United States), possible mechanisms linking socioeconomic status and cardiovascular disease, and experience in educational and preventive programs among different socioeconomic groups.9 An extensive chartbook of data on socioeconomic indicators and cardiovascular disease for the United States was compiled for the conference provided the following illustrations. Income and Education In the United States, a major data source for analysis of demographic factors and mortality is the National Longitudinal Mortality Study, which links data from household surveys of the Bureau of the Census with National Death Index information and death certificates. Data for some 600,000 people aged 25–64 years were investigated for the conference.10 Figures 16-1 and 16-2 indicate, for men and women from 1979 through 1989, the relation between cardiovascular mortality, plotted on a logarithmic scale, and income or education. For the analysis by income shown in Figure 16-1, the lowest stratum of $0–4000 was the reference category, and relative mortality for that group was set at 1.0. All analyses were adjusted for age and race (on the left in each panel of the figures) and, in addition, for socioeconomic status (SES) (on the right). For mortality by income, the socioeconomic adjustment included education, marital status, employment status, and household size; for mortality by education, income was substituted for education among the adjusting variables. Analyses by income (Figure 16-1) showed consistent trends of decreasing relative cardiovascular mortality with increasing categories of income for both men and women. After adjustment, these gradients were attenuated but remained substantial, with decrements of 40% for men and 50% for women
Source: Reprinted with permission from British Medical Journal, Vol 316, AE Kunst, F Groenhof, JP Mackenbach, and the EU Working Group on Socioeconomic Inequalities in Health, p 1638.
*p 0.05 for difference from 1.00. † Combined with other diseases. ‡ No distinction could be made between specific neoplasms or specific cardiovascular diseases.
Other Disease 1.50* 1.95* 1.49* 1.48* 1.49* 1.67* 1.89* 1.69* 1.23 1.42* 1.54*
1:00 PM
Mortality Rate Ratio Comparing Manual Classes to Non-Manual Classes for Specific Causes of Death in Men Aged 45–59 Other Lung Other Ischemlc Cerebrovascular Cardiovascular Respiratory Gastrointestinal Country Cancer Cancers Heart Disease Disease Disease Disease Causes Finland 2.20* 1.14* 1.47* 1.55* 1.52* 2.37* 1.37* Sweden 1.46* 1.11* 1.36* 1.31* 1.42* 1.91* 1.58* Norway 1.62* 1.15* 1.35* 1.21* 1.31* 1.68* 1.42* Denmark 1.51* 1.09* 1.28* 1.28* 1.28* 2.30* 1.65* England and Wales 1.54* 1.07 1.50* 1.74* 1.46* 2.13* † Ireland 1.95* 1.17* 1.23* 1.57* 1.40 2.00* 1.08 France 1.65* 1.75* 1.14 1.61* 1.54* 2.63* 2.20* Switzerland 1.73* 1.29* 0.96 1.43* 1.26 2.31* 1.62* Italy‡ ‡ ‡ ‡ ‡ ‡ 1.63* 1.78* Spain 1.38* 1.31* 0.98 1.18* 1.68* 1.89* 1.43* Portugal 1.07 1.15* 0.76* 1.44* 1.14 2.13* 1.59*
2/26/10
Table 16-1
46894_CH16_503_532.qxd Page 507
SOCIAL STATUS 507
46894_CH16_503_532.qxd
1:00 PM
Page 508
CHAPTER 16 SOCIAL AND PHYSICAL ENVIRONMENT
Income, in 1,000 Dollars
50+
25–49
15–19
20–24
0–4
50+
25–49
15–19
20–24
5–9
10–14
0–4
50+
25–49
15–19
0.1
5–9
0.1
10–14
0.2
0–4
0.2
5–9
0.4
10–14
0.4
0.8 0.6
0–4
0.6
1.0
50+
0.8
Adjusted for age, race, and SES
25–49
1.0
Adjusted for age and race
15–19
Relative Mortality
2.0
20–24
Adjusted for age, race, and SES
Women 25–64 Years
5–9
Adjusted for age and race
2.0
Relative Mortality
B
Men 25–64 Years
10–14
A
20–24
508
2/26/10
Income, in 1,000 Dollars
Figure 16-1 Relative Cardiovascular Mortality by Income for (A) Men and (B) Women. Source: Reprinted from PD Sorlie, NJ Johnson, and E Blacklund, Report of the Conference on Socioeconomic Status and Cardiovascular Health and Disease, p 24, 1995, National Institutes of Health.
17+
16
12
13–15
9–11
8
16
17+
12
13–15
9–11
8
0–7
17+
16
13–15
0.1
12
0.1
8
0.2
9–11
0.2
0–7
0.4
12
0.4
0.8 0.6
9–11
0.6
Adjusted for age, race, and SES
1.0
8
Relative Mortality
0.8
Education, Highest Grade Completed
Adjusted for age and race
2.0
1.0
0–7
Relative Mortality
Adjusted for age, race, and SES
Women 25–64 Years
0–7
Adjusted for age and race
2.0
B
17+
Men 25–64 Years
A
16
Income Inequality Further insight into the strength of socioeconomic status as a predictor of total mortality is given in
Figure 16-3, which shows the slope index of inequality in 1980, contrasted with 1960, for White men aged 25–64 years in six strata of increasing educational attainment. Total mortality, on a logarithmic scale, averaged 5.1/1000 in 1980, a substantially lower rate than 8.0/1000 in 1960. However, the differential across categories of educational attainment increased over this interval. Thus the gap in mortality between the lowest and highest educational strata increased from 3.9/1000 in 1960 to 4.1/1000 in 1980, and the ratio of this difference to the overall average mortality increased from 49 to 80%. Expressed in various terms, the highest educational stratum gained most in reduced mortality, the lowest stratum gained least, or the gap widened. Graphically, the slope of differ-
13–15
from lowest to highest income. Analysis by years of education (Figure 16-2) was based on 12 years, or completion of high school, as the reference category. In general, fewer years of school completed meant higher mortality and more years of school lesser mortality from cardiovascular diseases. Peak mortality was generally not in the lowest category. Similar patterns by race/ethnicity (African American, Hispanic White, and non-Hispanic White) showed consistent trends of lesser mortality with higher educational status in every group of both men and women.
Education, Highest Grade Completed
Figure 16-2 Relative Cardiovascular Mortality by Education for (A) Men and (B) Women. Source: Reprinted from PD Sorlie, NJ Johnson, and E Blacklund, Report of the Conference on Socioeconomic Status and Cardiovascular Health and Disease, p 24, 1995, National Institutes of Health.
46894_CH16_503_532.qxd
2/26/10
1:00 PM
Page 509
SOCIAL STATUS 509
Annual Death Rate/1000
100
1960 difference = 3.9 av death rate = 8.0 diff/average = .49
1980s difference = 4.1 av death rate = 5.1 diff/average = .80
10 3.9 4.1
1 0%
100%
0%
100%
Education (Percentile) Figure 16-3 Estimation of Slope Index of Inequality for White Men Aged 25–64 Years. The slope index of inequality represents the decrease in the death rate from the lowest education to the highest education, with education scaled as a percentile. The bars indicate the death rates for each level of education. The width of each bar represents the percent of population at each education category. The line shows the slope of a regression of the death rates on education scaled as cumulative percentiles of the actual education levels. “Difference” is the difference in the death rate from the lowest to the highest education level. “Average death rate” is the death rate for the group as a whole. Source: SH Preston and IT Elo, Journal of Aging and Health, Vol 7, p 486, Copyright © 1995 by Sage Publications, Inc.
ential mortality by education increased despite, or in consequence of, the overall decrease in mortality. The Robin Hood Index Differentials in income have been widening in the United States. For investigating the impact on mortality, the “Robin Hood index” was calculated for each US state for 1990.11 This index is the proportion of income that would need to be redistributed from higher to lower income strata to eliminate disproportionate income distribution. Analyses of causespecific mortality by state for the same year showed that the wider the income disparity as measured by the Robin Hood index, the higher the heart disease (p 0.004) and cerebrovascular disease (p 0.058) mortality, as well as total mortality, infant mortality, deaths from malignant neoplasms, and homicide. This was found even after adjustment for absolute differences in income level. It was projected that proportionate redistribution of incomes in the United States to resemble those in the United Kingdom would correspond to a decrease in the Robin Hood index from 30% to 25%
and a corresponding 25% reduction in coronary heart disease mortality. It should be noted that this result would not have been expected earlier in the course of the US epidemic of coronary heart disease, when rates were highest in the higher socioeconomic strata. The dependence of these associations on other factors was also evident from the changing social class distribution of coronary heart disease in the United Kingdom, discussed previously.6 Disadvantage and Social Status Social Disadvantage Apart from occupation, other measures have been devised to represent variation in social well-being, with special interest in the least favorable categories. Anand and others, for example, developed a “social disadvantage index” through analysis of several social factors as statistical predictors of prevalent cardiovascular disease in a multiethnic sample from four Canadian communities.12 Factors that emerged to form the index were two strata of low income, unemployment, and being unmarried. Scores for the
46894_CH16_503_532.qxd
510
2/26/10
1:00 PM
Page 510
CHAPTER 16 SOCIAL AND PHYSICAL ENVIRONMENT
component variables, weighted on the basis of the observed regression coefficients, were: lowest income, 2; next lowest income, 1; unemployment 2; and unmarried, 1. Least disadvantage scored 0 and the greatest disadvantage scored 5. Figure 16-4 presents the relation of the score to a composite measure of cardiovascular disease, based on history and examination findings in a cross-sectional analysis. Probability of cardiovascular disease (“risk” in the figure) increased continuously with the social disadvantage score, as would be expected from the method of its derivation. The score was greatest for older people, women, and non-White population groups. The gradient of risk differed among sex– ethnic groups, such that at any given score the risk was substantially greater for aboriginal and South Asian men and women than for other groups, especially Chinese men and women, for whom it was lowest. The index may therefore be useful as a tool for assessing social characteristics in relation to cardiovascular disease in other populations. “Unnatural Causes” and “The Eight Americas” The idea of disparities in health associated with relative deprivation, and of inequity in health especially afflicting the lowest strata of society, underlies a film 0.05
documentary depicting these issues in the United States at the start of the new century: Unnatural Causes: Is inequality making us sick?13 The four-hour film series aims “to enlarge our public discourse about health” by achieving several objectives. Collectively, these are to increase public awareness of the role of social conditions and extent to which they determine poor health for a large segment of our society, as well as to point to potential solutions at the community level. Murray and others presented data on causespecific mortality for eight geographic/demographic strata of the US population (“The Eight Americas”) showing gaps of more than 15 years for men and 12 years for women from lowest to highest strata of life expectancy at birth.14 Cardiovascular deaths were a large component of the differences in life expectancy both among the eight Americas and between the United States and other countries. Residual Effects Kaplan and Keil presented an extensive review of relationships between socioeconomic factors and cardiovascular diseases.15 They addressed methodologic aspects, including assessment of education, income, occupation, employment status, social class, and living
Risk of Cardiovascular Disease and Social Disadvantage
Risk of Cardiovascular Disease
0.045 0.04 0.035 0.03 0.025 0.02 0.015 0.01 0.005 0 0
1
2 3 Social Disadvantage Score
4
Aboriginal Male
South Asian Male
Aboriginal Female
South Asian Female
European Male
Chinese Male
Chinese Female
European Female
5
Note: Figure depicts the differences in the predicted probability of CVD by sex-ethnic groups using the multivariate logistic regression model. In the overall dataset, Aboriginal people and South Asians have a higher probability of CVD compared with Europeans, whereas Chinese have a lower probability. In addition, women have a lower probability of CVD than men, although considering both sex and ethnic differences, Aboriginal and South Asian women have a higher predicted probability of CVD compared with European and Chinese men. Figure 16-4 Relationship of Social Disadvantage and Risk of Cardiovascular Disease in Men and Women. Source: Reprinted with permission from International Journal of Epidemiology, Vol 35, SS Anand, F Rasak, AD Davis, et al., p 1244, © The Author 2006.
46894_CH16_503_532.qxd
2/26/10
1:00 PM
Page 511
SOCIAL STATUS 511
conditions, as well as several area-based measurements. Special problems of assessing socioeconomic status over the life span and opportunities for study of income inequalities were also discussed. Numerous studies of socioeconomic status and coronary heart disease were reviewed, from work by Lilienfeld in the 1950s to the report of the Scottish Heart Study in 1992. Occupational studies and studies of psychosocial factors as mediators between social class and coronary heart disease were included. The decline of coronary mortality in the United States was found to be associated with socioeconomic status for males but not for females, with more highly educated males experiencing more rapid decline in coronary mortality. Widening disparities in coronary mortality between higher and lower socioeconomic strata, even while absolute rates fell in all strata, were observed in both the United States and Great Britain. They concluded that socioeconomic status has been found to be associated with coronary heart disease in many studies over a 40-year period; patterns of association between socioeconomic status and coronary heart disease have changed, among men, over this period; declines in coronary mortality have not affected all socioeconomic groups equally; and several cardiovascular risk factors are inversely associated with socioeconomic status, such that higher risk is found in lower socioeconomic categories, yet some residual association between socioeconomic status and coronary heart disease remains when these factors are taken into account, suggesting a possible independent contribution of socioeconomic status to risk. “Fundamental Causes” Link and Phelan in 1995 introduced the term “fundamental causes” to denote the concept that certain social conditions determine the occurrence of disease and are not merely markers of true causes or pointers toward risk factors that operate more proximally in relation to disease.16 Social conditions were claimed to be fundamental causes on grounds that they lack dependence on specific, biologically plausible mechanisms and persist in producing disease even if one or another particular associated mechanism is successfully countered by intervention. For example, the relation of poverty with excess morbidity and mortality persists despite the removal of successive hazards to which the poor are known to be differentially exposed. A fundamental social cause of mortality would influence multiple disease outcomes, affect those outcomes through multiple risk factors, and persist through time regardless of change in the particular mechanisms of disease operating at any particular time. Further, the “essential feature of fundamental social causes is that they involve access to resources
that can be used to avoid risks or to minimize the consequences of disease once it occurs.”16, p 87 On reflection, this characterization appears to define not a fundamental cause of mortality, but a cause of survival or of differential survival given the lack of uniform access to the assets described. In this sense, a fundamental cause is a positive influence, although its distribution is unequal across society. But the authors’ discussion of the importance of fundamental causes for improving health seems to revert to a negative concept: the argument was advanced that intervention against specific risk factors is inherently less promising than intervention that addresses fundamental causes, that is, the social conditions under which the occurrence of diseases aggregates. Corollaries of this argument were that broad social policies may have far-reaching effects on health through their influence on these fundamental causes and that medical sociologists and social epidemiologists should become capable of preparing a “health impact statement” as part of the development of social and economic policy. The fundamental meaning of “fundamental causes” remained unclear. Phelan and others elaborated further on their theory of fundamental causes and reported the outcome of an empirical test.17 The focus of the theory was again expressed as factors that protect health and ensure persistence of socioeconomic disparities “because socioeconomic status embodies an array of resources, such as money, knowledge, prestige, power, and beneficial social connections, that protect health no matter what mechanisms are relevant at any given time.”17, p 265 The authors predicted that higher educational attainment or family income would be associated with a survival advantage for causes of death that are preventable, but not for other causes, given that social advantage would not be availing for diseases lacking effective preventive measures. Observations from the National Longitudinal Mortality Study were consistent with the predicted outcome and so supported the theory. The “Status Syndrome” A different conception of social status and its relation to health is captured in Sir Michael Marmot’s term, the “status syndrome.”18 Concisely stated, his thesis is that “social standing affects our health and longevity.” “For people above a threshold of material well-being,” he reasoned, referring to most populations free of absolute deprivation, “another kind of well-being is central. Autonomy—how much control you have over your life—and the opportunities you have for full social engagement and participation are crucial for health, well-being, and longevity. It is inequality in these that plays a big
46894_CH16_503_532.qxd
512
2/26/10
1:00 PM
Page 512
CHAPTER 16 SOCIAL AND PHYSICAL ENVIRONMENT
part in producing the social gradient in health. Degrees of control and participation underlie the status syndrome.”18, p 2 As with other aspects of social environment discussed here, two levels of meaning can be distinguished—the individual, whose position or status within society entails a “syndrome” of personal ills, and society, whose graded structure engenders these status differences and a metaphorical “syndrome” of societal ills. The concept of a gradient in health associated with degrees of these qualities of control, engagement, and participation is one distinguishing aspect of the thesis; it does not focus exclusively on one extreme of the social distribution, whether in terms of education, income, occupation, or other characteristics. By its avoidance of the frequent dichotomy between disfavored and favored strata of society, it emphasizes that differences in health are evident all across the spectrum, such that nearly everyone has room to gain in health by improvement in social organization. It is also true that gradients can change in amplitude and at least for specific causes such as heart disease mortality reverse, as occurred among male British civil servants, alluded to previously.7 This point leads to consideration of changes in social conditions, with the likelihood that the distribution of health and disease in a population will in some way reflect the change.
CHANGES IN SOCIAL CONDITIONS Concepts of Social Change Two distinct concepts of social change are reflected in studies of the social environment and cardiovascular disease. In brief, one concerns experience of social factors of every kind by individuals as they go through life—the “life course” approach to understanding the influence of social factors. Another concerns changes in social conditions themselves, over time, such that populations have different experiences over a period of years or decades, or between generations, or over long historical epochs. Fuller discussion of these concepts will illustrate how they have shaped thinking and been investigated in epidemiologic studies. Examples to illustrate these concepts will draw from studies of the maternal and fetal origins of adult cardiovascular disease and the experience of several populations undergoing culture change. Social Conditions over the Life Course Pollitt and others reviewed 49 observational studies regarding subclinical cardiovascular disease, incident events, or traditional risk factors including data
from ages other than adulthood.19 Methodologic aspects of this research were summarized as shown in Table 16-2. Several life course study designs were identified: Early life socioeconomic conditions related to later-life CVD outcomes; a similar approach taking risk factors as outcomes; following the “social trajectory” of individuals changing socioeconomic levels within their lifetimes and CVD outcomes; and a “cumulative SES” approach based on an index of overall negative social influences preceding a CVD outcome. These several approaches suggest different perspectives on social conditions throughout life that are reflected in the different kinds of study questions addressed through each approach. As will be seen, these concepts have longstanding precedents in the epidemiologic literature. Pollitt’s findings were summarized as showing a modest impact of low lifetime socioeconomic status on cardiovascular risk, including risk factors and morbidity and mortality; little support for social mobility as influencing cardiovascular outcomes; and consistent support for an adverse impact of cumulative negative social experiences and conditions on lifetime cardiovascular risk. Culture Change Many years previously, Cassel and colleagues considered that epidemiologic study of health phenomena depended for its utility on generating and testing specific hypotheses to understand the significance of social and cultural processes for health.20 This view indicated confidence in the potential for deriving valid general inferences from studies of multiple populations or groups. Contemporary sociology and anthropology offered many concepts for potential investigation but did not establish the relevance of any of them to health. Cassel and Tyroler chose to focus on culture change, specifically on changes in way of life from rural agricultural to industrial settings.21,22 They adopted an “open” model of chronic disease causation, which meant that social factors might have very diverse health effects and that any particular health effect might occur in consequence of a variety of social factors. Specific associations would be highly conditional on immediate circumstances. Health status should therefore be studied, in their view, by reference to several dimensions of health: specific diseases; growth, nutritional status, and selected physiological functions; psychological attributes; and social adjustment. They proposed and later tested these hypotheses: (1) Among workers in industrialized areas of the Appalachian Mountain region, recent migrants from rural areas would exhibit
46894_CH16_503_532.qxd
2/26/10
1:00 PM
Page 513
CHANGES IN SOCIAL CONDITIONS 513
Table 16-2
Life Course Study Designs: Hypotheses Tested and Typical Study Questions Posed Typical Chronological Analysis Set-Up Life Course Study Design Used in Study Design Typical Study Questions Early SES → Outcome Is there a significant independent effect Early-life SES variable(s) used to predict of early-life (childhood) SES on the CVD (e.g., CHD, stroke). Adjusted for adult risk of CVD after adjusting for later-life events, behaviors, risk factor later-life SES and risk factors? levels to determine “direct” effect of early life SES. Early SES → Risk Factor
Early-life SES variable(s) used to predict adult CVD risk factors levels.
How does early-life SES affect later-life levels of behavioral and physiological CVD risk factors?
Social Trajectory
Inter- or intragenerational movement from one SES level to another (i.e., Low SES to High SES) used to predict adult CVD risk factors levels or CVD outcomes.
How does social mobility from one point to another during the life course affect the risk of CVD?
Cumulative SES
A summary variable indicating number of negative SES events/environments over the life course used to predict adult CVD risk factor levels or CVD outcomes.
How does the accumulated number of negative SES-related exposures across the life course influence the risk of CVD?
CHD Coronary heart disease. Pollitt et al. BMC Public Health 2005 5:7 doi: 10.1186/1471-2458-5-7 Source: Reprinted with permission from BMC Public Health, Vol 5, RA Pollitt, KM Rose, JS Kaufman, Table 1, © Pollitt et al.
poorer health status than either those who had not migrated or those who had migrated a generation earlier; and (2) those recent migrants with the least family solidarity or who were experiencing upward social mobility would manifest the greatest adverse effects of culture change. Their results from this line of investigation, in relation to both general health and coronary heart disease, were generally consistent with these expectations.21,22 Cultural Mobility Syme and colleagues considered culture change to have four aspects: “generational, career, residential, and situational mobility.”23, p 178 They investigated several social factors in relation to coronary heart disease in a case-comparison study in the early 1960s. Factors such as urban residence, white-collar occupation, and geographic and occupational mobility were present in newly occurring cases of coronary heart disease significantly more frequently than in comparison subjects. These associations were independent of diet, relative body weight, blood pressure, smoking, and parental longevity. Migration The studies of Cassel and Tyroler were part of a broader interest in migration and its effects on cardiovascular disease, especially hypertension. Scotch and others studied blood pressure levels in rural and
(migrant) urban Zulu tribes in South Africa.24,25 Results indicated lower starting levels of blood pressure in early adulthood than in many populations, but steeper slopes of increasing blood pressure with age for urban Zulu than for those who remained in the rural settings—for both men and women. In general, migrants to a new social setting have been observed to develop health characteristics resembling those of the “host” population, as will be seen in examples as follows. Societal Change over Time “Disturbances of Human Culture” As often quoted by Stamler, Virchow expressed the view a century and a half ago that human culture is a decisive determinant of epidemic diseases:26, p 35 Epidemics of a character unknown so far appear, and often disappear without traces when a new culture period has started. . . . The history of artificial epidemics is therefore the history of disturbances of human culture. Their changes announce to us in gigantic signs the turning points of culture into new directions. This view directs attention to broad features of social conditions and their changes over time. Many reflections of this perspective are found in the epidemiology of the past 50 years as it has addressed coronary heart disease and hypertension in particular.
46894_CH16_503_532.qxd
514
2/26/10
1:00 PM
Page 514
CHAPTER 16 SOCIAL AND PHYSICAL ENVIRONMENT
Epidemiologic Transition and Westernization The theory of epidemiologic transition, introduced by Omran in 1971 (see Chapter 1, “Cardiovascular Diseases: A Global Public Health Challenge”), occupies a position intermediate between the prehistory of modern societies and contemporary social change.27 Other sources consider the paleoanthropology of human society and the relation between evolutionary development of prehistoric patterns of nutrition, physical exertion, and energy balance. Omran began with observations on secular trends in proportionate mortality in the United States from the beginning of the 20th century. He then generalized to a pattern of sequential phases of demographic and epidemiologic development, described by Pearson and others in terms specific to atherosclerotic and hypertensive diseases (see Chapter 1 and Table 1-1).28 An optimistic view of the epidemiologic transition was not that the playing out of the underlying forces was inevitable but that they were caused by human—societal—action and could therefore in principle be modified or controlled. Trowell and Burkitt, in Western Diseases: Their Emergence and Prevention, justified the designation of the “man-made” diseases as “Western” in the sense that they are typical of affluent Western technologies, are less common in other countries and espe-
cially in lower socioeconomic groups, and are linked to the greater longevity and consequently increased opportunity for their expression in Western society.29 They identified several conditions as Western diseases by these criteria and indicated their increased frequency in the course of Westernization in many population groups, as shown in Table 16-3. Prominent among them are coronary heart disease (CHD), hypertension, obesity, and diabetes. Globalization and Economic Development The premise that social change at every level has implications for population health, and that globalization and economic development are strong determinants of global health, seems unquestioned. References throughout the preceding chapters to the Global Burden of Disease Study and the Disease Control Priorities Project, among others, reflect current perspectives.30–33 The projected burden of cardiovascular and other chronic diseases has large implications for social and economic development in low- and middle-income countries, and for organization and financing of health care in the United States, as examples of growing awareness and concern about societal change. Translation of this awareness and concern into political action for population health will be addressed in Part IV.
Table 16-3
Increased Incidence of Medical Diseases During Westernization Group Hypertension Obesity Diabetesa Gallstonesb Hunter-gatherers Eskimos Australian Aborigines North American Indians Agriculturalists in: West Nile, Uganda d Zimbabwe O South Africa (Bantu) Papua New Guinea d Pacific Islands Sub-Saharan Africa Polynesia Migrants Maoris South African Indians Israelis Far East Japan Taiwan Hawaiian groups O
Renal Stones
CHDc
d d
d
Note: Increase, ; doubtful increase, ; no change, O; no report, blank space. a Type 2 non-insulin-dependent variety, formerly called maturity-onset type. b Cholesterol-rich variety. c Coronary heart disease and angina. d Very rare in rural areas; no data concerning increased incidence. Source: Reprinted with permission from H Trowell, D Burkitt, Joint Enquiry into Western Disease, in Western Diseases: Their Emergence and Prevention, HC Trowell and D Burkitt, eds. © 1981, Edward, Arnold/Hodder & Stoughton Educational.
46894_CH16_503_532.qxd
2/26/10
1:00 PM
Page 515
CHANGES IN SOCIAL CONDITIONS 515
From Fetal Development to Adult Cardiovascular Disease Forsdahl on Norway Whether social conditions early in life are associated with later-appearing atherosclerotic and hypertensive diseases has been investigated for more than three decades. The specific hypothesis that poor living conditions in childhood and adolescence portend increased mortality from atherosclerotic heart disease in adulthood (in contrast to more-general theories of social conditions and health) appears to have originated in the writing of Forsdahl, whose first-cited Norwegianlanguage papers on the topic appeared in the early 1970s.34 He noted the general principle of an inverse relation between standard of living and mortality, which contrasted with the contemporary situation among adults in Norway, where there was a lack of variation in current living conditions to correspond with considerable variation in mortality by county. He reasoned that current differences in mortality might reflect former differences in living conditions earlier in life, when these conditions did vary greatly by county. His specific hypothesis was that “poverty during adolescence is positively correlated with the risk of dying from arteriosclerotic heart disease.”34, p 91 Forsdahl used as a measure of historical social conditions infant mortality by county in Norway for the years 1896–1925. Mortality from arteriosclerotic heart disease in 1964–1967 (as coded under the International Classification of Diseases, Eighth Revision) was analyzed for men and women in the age group 40–69, whose years of childhood and youth corresponded to the period of infant mortality data. The results shown in Figure 16-5a and b illustrate his findings. For both males and females, death rates for arteriosclerotic heart disease and previous infant mortality by county were strongly and significantly associated, whether measured by product-moment (r) or Spearman (rs) correlation coefficients. These and other results led Forsdahl to speculate that the association was not explained by early poverty alone; it depended on later affluence, which perhaps had adverse effects in consequence of an early nutritional deficiency, such as a reduced tolerance to some types of dietary fats. He concluded, “it seems justified to consider a poor standard of living in early years followed by prosperity as a potential risk factor.”34, p 95 The Southampton Group In a prodigious program of research beginning in the mid-1980s, Barker and colleagues at Southampton, England, have pursued the concept of early origins of coronary heart disease and have stimulated related
studies and several reviews of work in this area by others. Recently summarizing the results at an earlier stage, Barker wrote:35, p 162 A new model for the causation of coronary heart disease is emerging [citation to Barker DJP. Mothers, Babies and Health in Later Life. London, England: British Medical Journal Publications, 1994]. Under the old model an inappropriate lifestyle, including cigarette smoking and lack of exercise, leads to accelerated destruction of the body in middle and late life, including the more rapid development of insulin resistance. Under the new model, coronary heart disease results not primarily from external forces but from the body’s internal environment, homeostatic settings of enzyme activity, cell receptors, and hormone feedback, which are established in response to undernutrition in utero and lead eventually to premature death. The background of this conclusion includes numerous reports by this group of investigators and others, many of which were collected in Mothers, Babies and Health in Later Life or in Fetal and Infant Origins of Adult Disease.36,37 Infant Death Rates and Cardiovascular Mortality Among the earliest studies by the Southampton group was a historical-ecological investigation of mortality in infancy and adulthood, similar to that of Forsdahl. Adult mortality at ages 35–74 years in 1968–1978 was analyzed in relation to the corresponding infant mortality in 1921–1925 in each of the 212 local authority areas of England and Wales. The results for ischemic heart disease (ICD, Eighth Revision, codes 410-414) and other selected causes of death (bronchitis, stomach cancer, and rheumatic heart disease) indicated strong correlations (r 0.72–0.82).38 Stroke and lung cancer, though they had weaker correlations, were included in the analysis because they might serve as surrogates for hypertension and smoking history. Correlations were presented of mortality from ischemic heart disease with these other causes of death and with infant mortality in 1921–1925. Ischemic heart disease was more strongly related to stroke and infant mortality than to other causes for both men and women, except for women in urban areas. This overall result suggested that the relation of infant to adult mortality was to some degree specific for cardiovascular causes. Weight at Birth and Age 1 Year A further stage of investigation was to link individual data from birth and early postnatal records with mor-
46894_CH16_503_532.qxd
516
2/26/10
1:00 PM
Page 516
CHAPTER 16 SOCIAL AND PHYSICAL ENVIRONMENT
600
Males
Deaths from Arteriosclerotic Heart Disease/100,000 per Year
550 500 450 400 r = +0.86 P<0.001 rs = +0.79 P<0.001
350 300 250
A
200 40
60
80
100
120
140
Infant Mortality 1896–1925
200
Females
Deaths from Arteriosclerotic Heart Disease/100,000 per Year
175
150
125
100 r = +0.74 rs = +0.61
P<0.001 P<0.01
75
B
50 40
60
80
100
120
140
Infant Mortality 1896–1925
Figure 16-5 Correlation Between Mortality from Arteriosclerotic Heart Disease, 1964–1967, in Men (Upper Panel) and Women (Lower Panel) Aged 40–69 Years and Infant Mortality Rates, 1896–1925. Source: Reprinted with permission from Forsdahl, British Journal of Preventive & Social Medicine, Vol 31, p 92, © 1977, BMJ Publishing Group.
tality in adult years in a historical cohort approach. This was done through identification of 5654 men born in Hertfordshire, England, in the period 1911–1930, whose vital status was ascertained at ages 20–74 years from 1951 through 1987.39 Early records for each man included weight at birth and age 1 year, as well as feeding practices (breast, bottle, or both) and other information. For men who had been breast fed,
who constituted all but 7.6% of the cohort, the relation of ischemic heart disease mortality to weight at birth and 1 year is shown in Table 16-4. Standardized mortality ratios (SMRs) represent the relation of the death rate for each category to that for the national population of the same age and calendar period. The lowest ratios occurred in the categories of above-average weights both at birth and at age 1 year.
46894_CH16_503_532.qxd
2/26/10
1:00 PM
Page 517
CHANGES IN SOCIAL CONDITIONS 517
Table 16-4
Standardized Mortality Ratios for Ischemic Heart Disease According to Birth Weight and Weight at One Year in Men Who Had Been Breast Fed (Numbers of Deaths in Parentheses) Weight at Birth (lb) Below Average Above Weight (lb) at 1 Year Average ( 7) (7.5–8.5) Average ( 9) Total Below average ( 21) 100 (80) 100 (77) 58 (17) 93 (174) Average (22–23) 86 (34) 87 (67) 80 (29) 85 (130) Above average ( 24) 53 (14) 65 (42) 59 (32) 60 (88) Total 88 (128) 85 (186) 65 (78) 81 (392) Source: Reprinted with permission from DJP Barker et al., Weight in Infancy and Death from Ischaemic Heart Disease, in Fetal and Infant Origins of Adult Disease, DJP Barker, ed, p 145, © 1992, BMJ Publishing Group.
A similar analysis was subsequently reported for women and showed a similar pattern for weight at birth but not at age 1 year. For a sample of women, as for men, who were still residing in Hertfordshire, risk factors were assessed by direct physical and laboratory examination.40 For risk factors, as for mortality, associations in women were present for birth weight but not for weight at age 1 year. However, like men, women who were smaller at birth, and especially those who became obese as adults, exhibited higher blood pressure and triglyceride concentrations and lower concentrations of high-density lipoprotein cholesterol—prominent features of insulin resistance.
Further observations on the consequences of early growth patterns have come from several studies beyond the United Kingdom, for example, a birth cohort analysis in Helsinki.41 Among 8760 people born between the mid-1930s and 1940s, 444 hospital admissions or deaths from coronary heart disease had occurred at follow-up through 1998. Details of birth weight and BMI at 2 years of age for cases and controls permitted estimation of hazard ratios for coronary heart disease as shown in Table 16-5a and b. With the highest one-third of birth weight and of BMI at age 2 years as reference, every other group had an increased hazard ratio, greatest
Table 16-5a
Hazard Ratios for Coronary Heart Disease According to Birth Weight and BMI at Two Years of Age for Boys and Girls Combined* Hazard Ratio (95% CI) BMI at Two Years of Age 16 17 16–17 Birth Weight (kg) Adjusted for Sex, Adult Occupational Status, and Household Income 3.0 1.9 (1.3–2.8)/1.9 (1.3–2.9) 1.9 (1.2–3.0)/1.9 (1.2–3.1) 1.3 (0.7–2.2)/1.1 (0.6–2.1) 3.0–3.5 1.5 (1.0–2.1)/1.3 (0.9–2.0) 1.6 (1.1–2.2)/1.4 (1.0–2.1) 1.2 (0.8–1.8)/1.2 (0.8–1.8) 3.5 1.7 (1.2–2.5)/1.3 (0.9–2.1) 1.5 (1.1–2.2)/1.4 (0.9–2.0) 1.0/1.0† *Values for birth weight and body-mass index were divided into three groups of equal size. CI denotes confidence interval. †
This group served as the reference group.
Source: Reprinted with permission from The New England Journal of Medicine, Vol 353, DJP Barker, C Osmond, TJ Forsén, E Kajantie, JG Ericksson, p 1806, Copyright © 2005 Massachusetts Medical Society.
Table 16-5b
Hazard Ratios for Coronary Heart Disease According to BMI at 2 and 11 Years of Age for Boys and Girls Combined* Hazard Ratio (95% Cl) BMI at 11 Years of Age 16 17.5 16–17.5 BMI at Two Years of Age Adjusted for Sex, Adult Occupational Status, and Household Income 16 1.6 (0.8–3.3)/1.8 (0.8–4.2) 2.4 (1.2–4.9)/2.5 (1.1–6.0) 3.0 (1.4–6.3)/3.1 (1.3–7.8) 16–17 1.4 (0.7–3.1)/1.7 (0.7–4.1) 1.6 (0.8–3.3)/1.8 (0.8–4.4) 1.9 (0.9–3.9)/2.0 (0.8–4.7) 17 1.0/1.0† 1.3 (0.6–2.7)/1.5 (0.6–3.7) 1.1 (0.5–2.3)/1.2 (0.5–3.0) *Values for body-mass index were divided into three groups of equal size. CI denotes confidence interval. † This group served as the reference group. Source: Reprinted with permission from The New England Journal of Medicine, Vol 353, DJP Barker, C Osmond, TJ Forsén, E Kajantie, JG Ericksson, p 1806, Copyright © 2005 Massachusetts Medical Society.
46894_CH16_503_532.qxd
518
2/26/10
1:00 PM
Page 518
CHAPTER 16 SOCIAL AND PHYSICAL ENVIRONMENT
for those in the lowest category of both measures. Adjustment for adult occupational status and household income did not attenuate the hazard ratio, indicating that adult status did not materially modify the effect of these characteristics at birth and in infancy. Additional data on BMI at age 11 years showed that greater-than-average increase from age 2 to 11 years, to reach the population average, was also associated with coronary events. This finding suggests that the pattern of growth in body mass from birth to adolescence is an important predictor of adult cardiovascular disease, including (from other data presented) insulin resistance. A Mechanism: “Programming” A plethora of relationships has been identified in this area of research, implicating multiple risk factors or disease outcomes, at various ages, in associations with different aspects of fetal and early infant development. These, in turn, have been measured by weight at birth or age 1 year, placental weight, abdominal girth at birth, and other indices. The fundamental mechanism that Barker proposed to account for these relationships is deficiency of nutrient or oxygen supply to the fetus or infant, to which the body responds with particular physiologic or metabolic adaptations. When these adaptations occur at critical points in early development they constitute “programming,” which determines the subsequent expression of genetic controls over growth, physiology, and metabolism throughout later life, a phenomenon recognized in animal experimental research.35 The link between programming and social conditions derives from the expectation that circumstances of maternal poverty are likely to underlie an adverse fetal or infant environment. Clearly, if major determinants of adult cardiovascular diseases operate in this period of life, intervention strategies for modification of behavior and social environment in adulthood may be too little, and too late, relative to the potential for true “early” intervention. Counterargument The theory that risks of coronary heart disease are determined during fetal development and infancy has received much comment and has met with some skepticism. For example, two reports have reviewed related work, one on 10 ecological studies and the other on 15 longitudinal and 4 case-comparison studies.42,43 Their findings led Elford and colleagues to conclude in 1992 that the hypothesis was insufficiently clear; that biological mechanisms specific to the hypothesis remained to be formulated; and that tests of the hypothesis in relation to geographic and temporal vari-
ations in coronary heart disease occurrence were required for rigorous testing of the hypothesis. In summary, they found the evidence insufficient to support a claim of causality for the reported associations. A subsequent editorial acknowledged the merit of the theory but shared these reservations: “The Southampton group has provided an intriguing but very general hypothesis, often ingeniously pursued, that has served to provoke the somewhat complacent world of cardiac epidemiology. As a hypothesis with substantial implications for public policy it deserves rigorous testing.”44, p 412 The charge of excessive claims for what has been dubbed by some “the Barker hypothesis” (by others than Barker himself) points to issues in evaluation of competing theories of disease causation that are addressed in Chapter 17, “What Causes Cardiovascular Diseases?” Barker’s “new model for the causation of coronary heart disease” has nonetheless received wide interest and could stimulate further action to improve maternal and child health through greater attention to living conditions.35, p 162 Migration and Cardiovascular Disease US Foreign Born/Native Born with Foreign Parents It was noted previously that migrants tend to adopt the disease patterns of the new host population. Taking this concept a step further, Kelleher and others investigated whether the epidemic curve of coronary heart disease mortality in the United States might be explained by immigration.45 They studied the time course of immigration to the United States between 1850 and 1930 and of coronary mortality from 1900 to 1980. Immigrants were those who were foreign born or native born with foreign parentage (FBNBFP). As shown for women in Figure 16-6A, with a 38year lag to optimize the fit, a close correspondence was found between the percentage of the US population that was FBNBFP and the age-adjusted time course of coronary mortality. Figure 16-6B indicates increasing mortality with increasing percentage of the population of women 38 years earlier who were FBNBFP. The authors concluded:45, p 465 . . . there is an impact of immigration on the pattern of the epidemic, mediated through a combination of factors, such as accumulated life-course susceptibility, deprived socio-economic conditions upon arrival, and the enthusiastic uptake of behaviours related to the classic risk factors of smoking, high saturated fat and salt diet. Our analysis provides a more contextualized understanding of the scale and timing of the epidemic of CHD in the U.S.
46894_CH16_503_532.qxd
2/26/10
1:00 PM
Page 519
35
500
30
400
25
300
20
White Female Age-Adjusted Heart Disease Rate
38 Years Lag of % FBNBFP
CHANGES IN SOCIAL CONDITIONS 519
200 1920
1940
1960 Year
38 Years Lag of % FBNBFP
1980
2000
White Female Age-Adjusted Heart Disease Rate
(a)
Figure 16-6A Thirty-Eight Years Lag of Percentage of US Combined Foreign Born and Native Born with Foreign Parentage, 1870–1970, and Female Age-Adjusted Heart Disease Rate (HDR) Among Whites, 1914–1998. Source: Reprinted with permission from Social Science & Medicine, Vol 63, CC Kelleher, JW Lynch, L Daly, et al., p 476, © 2006 Elsevier Ltd.
White Female Age-Adjusted Heart Disease Rate/Fitted Values
Japan, Hawaii, and California The effect of migration on coronary heart disease in Japanese men residing in Hawaii or California, relative to the situation of those remaining in Japan, and the role of culture change in this process were studied in detail in the Ni-Hon-San Study, especially in a cross-sectional analysis of the three study groups.46 Analysis of differences in prevalence of coronary heart
disease in baseline surveys in the three areas indicated only partial explanation by differences in distributions of dietary pattern, blood cholesterol concentration, blood pressure, smoking, or other specific risk factors. Indices of culture change from traditional Japanese orientations to the conditions of Hawaii and California were also studied. These measured both the degree of acculturation versus adher-
500
400
300
200 20
25 30 38 Years Lag of % FBNBFP White Female Age-Adjusted Heart Disease Rate
35
Fitted Values
(b)
Figure 16-6B Linear Regression Fit for Female Age-Adjusted HDR with 38 Years Lag of Combined US Percentage Foreign Born and Native Born with Foreign Parent. Source: Reprinted with permission from Social Science & Medicine, Vol 63, CC Kelleher, JW Lynch, L Daly, et al., p 476, © 2006 Elsevier Ltd.
46894_CH16_503_532.qxd
1:00 PM
Page 520
CHAPTER 16 SOCIAL AND PHYSICAL ENVIRONMENT
Kenya Social conditions in rural and urban areas differ in innumerable ways, and study of migration from one to the other can contribute importantly to understanding determinants of health. The Kenyan Luo Migrant Study illustrates this point and its main results were described previously (see Chapter 12, “High Blood Pressure”). Among migrants from a rural tribal area to urban Nairobi, in comparison with rural nonmigrants, increased systolic and diastolic blood pressure, body weight, pulse rate, and urinary Na/K ratio were observed within weeks of migration (Figure 16-7).47,48 In an earlier report, Poulter pointed to the importance of such an epidemiologic study:49, pp 67–68 Assuming the basic premise that the causes of high blood pressure changes upon migration are those responsible for the pathogenesis of essential hypertension, then this study has provided the data which forms the basis of a new hypothesis for the evolution of essential hypertension . . . and is a unique model upon which it is possible to base primary intervention studies. Such studies, preferably carried out in countries where hypertension is increasing rapidly, may in turn bring us as close as we are ever likely to get towards understanding the aetiology and pathogenesis of essential hypertension. Population Experience of Social Change Among the more common changes in social conditions taking place in recent decades, and expected to continue, are changes in economic circumstances, culture, and lifestyle attendant on political change, economic development, migration, and other factors, including military conflict. Studies of some of these processes and their relation to cardiovascular disease
SBP (mmHg)
130 125 120 115
DPB (mmHg)
70 65 60 55 50 Pulse Weight (Beats/min) (kg)
ence to traditional patterns of behavior and the extent of access to and reliance upon other members of the Japanese community. Further analyses incorporating these measures indicated that those who changed least and had the strongest support for adherence to aspects of traditional culture had the lowest prevalence of coronary heart disease. Thus, for the men age 45–54 years, the highest prevalence of coronary heart disease was 7.3% in those who were most acculturated and had least reliance on the Japanese community. This was about 2.5 times the lowest prevalence, which was in those who were more traditional and reliant on the Japanese community. Corresponding frequencies for the age group younger than 45 years ranged from 5.1 to 1.0%, a fivefold difference.
60 55 70 65 60 5
Na/K
520
2/26/10
4 3 2 0
3
6
12 Months
18
24
Figure 16-7 Age-Adjusted Mean Variables in Rural Control (---) and Urban Migrant (—) Men. SBP, Systolic Blood Pressure; DBP Diastolic Blood Pressure. Source: Reprinted with permission from Ethnic Factors in Health and Disease, JK Cruikshank, DG Beevers, eds, N Poulter, pp 61–68, © Butterworth & Co. (Publishers) Ltd, 1989.
and other health conditions have drawn on the concepts reviewed previously and the works of Hinkle, Cassel, Tyroler, Syme, and others. This area of work is illustrated by several examples. Socioeconomic Status and Social Change in the United States Change in cardiovascular and all-cause mortality among US men and women aged 25–64 years over 30 years, from 1969 to 1998, was examined by Singh and Siahpush in relation to a county-level composite index of areal socioeconomic status.50 Counties were stratified by quintile groups of the index and allcause death rates and relative risks of cardiovascular death were calculated for each socioeconomic group and each time period, separately by sex (Figure 16-8). The figure shows marked declines in all-cause death rates for all groups and both sexes from 1969 to
46894_CH16_503_532.qxd
2/26/10
1:00 PM
Page 521
1997
1995
1993
1991
1989
1987
1985
40 1983
1997
1995
1993
1991
1989
1987
1985
1983
1981
1979
1977
60
Men (CVD Mortality)
1.3 1.2 1.1
1.5 1.4 1.3 1.2 1.1
1969–70
1997–98
1993–94
1989–90
1985–86
1981–82
0.9 1977–78
0.9 1973–74
1.0
1969–70
1.0
1997–98
1.4
1.6
1993–94
1.5
1.7
1989–90
1.6
1.8
1985–86
1.7
1981–82
1.8
1st Quintile (Low SES) 2nd Quintile 3rd Quintile 4th Quintile 5th Quintile (High SES)
1.9 Age-Adjusted Relative Risk of Mortality
1.9
Women (CVD Mortality)
2.0
1977–78
1st Quintile (Low SES) 2nd Quintile 3rd Quintile 4th Quintile 5th Quintile (High SES)
1973–74
2.0
Age-Adjusted Relative Risk of Mortality
1975
1973
1971
100
80
1981
150
100
1979
200
120
1977
250
140
1975
300
160
1973
350
1st Quintile (Low SES) 2nd Quintile 3rd Quintile 4th Quintile 5th Quintile (High SES)
180
1971
400
Women
200
1969
1st Quintile (Low SES) 2nd Quintile 3rd Quintile 4th Quintile 5th Quintile (High SES)
Age-Adjusted Death Rate per 100,000 Population 1970 US Population Used as Standard
Men
450
1969
Age-Adjusted Death Rate per 100,000 Population 1970 US Population Used as Standard
CHANGES IN SOCIAL CONDITIONS 521
Figure 16-8 Age-Adjusted Death Rates (Upper Panels) and Relative Risks (Lower Panels) of Cardiovascular Disease (CVD) Mortality Among US Men (Left Panels) and Women (Right Panels) Aged 25–64 Years by the 1990 Area Socio-Economic Status (SES) Index, 1969–1998. Source: Adapted with permission from International Journal of Epidemiology, Vol 31, GP Singh, M Siahpush, pp 606–607, © International Epidemiologic Society 2002.
1998. The middle category declined from about 400 to 150/100,000 deaths among men and from about 150 to 75/100,000 deaths for women, and relative positions of the groups were fairly constant over this interval. For cardiovascular mortality, however, with high SES set as the reference at 1.0, every other stratum of SES experienced an increasing relative risk over these three decades. From the beginning of the period, the relative risk was increasing for the lowest SES quin-
tile group of counties; the rate of change was amplified, and joined by the other three groups, from about 1983–1984. These patterns of change were similar for men and women. As interpreted by the authors, the pattern resulted from greater decreases in mortality for the highest-SES group than for the other groups. This difference was attributed, tentatively, to “increasing temporal differences in the material and social living conditions between areas.”50, p 600 Notably, the differential gain for the highest SES group
46894_CH16_503_532.qxd
522
2/26/10
1:00 PM
Page 522
CHAPTER 16 SOCIAL AND PHYSICAL ENVIRONMENT
was substantially greater for CVD than for all-cause mortality. This observation recalls, as an alternative interpretation, the understanding of “fundamental causes” as enabling the most favored stratum of society to benefit from prevention of preventable causes of death. Palau Palau, located in the Western Caroline Islands, was a post-World War II protectorate of the United States and is now an independent republic. In the late 1960s and early 1970s a study compared inhabitants of three distinct areas: a traditional area of Palau not greatly changed from prewar village life (Ngerchelong), another area undergoing major social and environmental change as the administrative center for this District of the US Trust Territory (Koror), and an area intermediate in exposure to such change (Peleliu).51 Interview questionnaires and health examinations documented differences in culture, behaviors, and risk factors for cardiovascular diseases. Risk factors and prevalence of electrocardiographic abnormalities were, overall, least favorable in Koror, the most modern of the three settings, and most favorable in Ngerchelong, the most traditional area (Table 16-6). The age patterns of change in systolic blood pressure also favored the most traditional area, which exhibited similar slopes of increase in blood pressure with age, but lower levels at all ages for both males and females. Other examples of the impact of change from traditional to Western culture include the local island population of Nauru, whose epidemic obesity and diabetes of recent decades are clearly linked to this change (see Chapter 13, “Diabetes and the Metabolic Syndrome”). Rural India In a rural population of Rajhastan, India, a survey of 3182 men and women villagers older than age 20
years permitted comparison of risk factors and prevalence of coronary heart disease by level of education.52 Educational levels were graded as nil (group 1), 1–5 years (group 2), 6–10 years (group 3), and more than 10 years (group 4). Proportions of men and women who had elevated blood pressure, cholesterol concentration (based on a subsample of 300 participants), and body mass index; who smoked; and who had a sedentary lifestyle were determined. Except for sedentary living, all factors tended to be most adverse for the group with no education, for both men and women. High proportions of persons in all groups were reported as having a sedentary lifestyle. Coronary heart disease prevalence by specified criteria in each group was greatest among those without education, for both men and women. This report is contrary to the preconception that higher socioeconomic status will necessarily be associated with the early phases of epidemic coronary heart disease in developing countries. This observation suggests the importance of further studies in more diverse community settings and the need to understand the situation in a particular locality more clearly before planning targeted intervention toward one or another presumed high-risk stratum of the population.
PARTICULATE AIR POLLUTION Focus on PM2.5 Evidence of increased risk of coronary heart disease resulting from both short- and long-term exposure to particulate air pollution stimulated development of an American Heart Association Scientific Statement on this issue in 2004.53 The main focus is particulate matter (PM) in the range of 2.5 m median aerodynamic diameter or less (PM2.5), which is small enough to reach small airways and alveoli. Figure 16-9 illus-
Table 16-6
Age-Adjusted Mean Values for Selected Characteristics, by Sex and Area, Palau, 1968/1970 Males Females Characteristic Koror Peleliu Ngerchelong Koror Peleliu Ngerchelong Blood pressure, systolic (mm Hg) 136 135 121 133 137 124 Blood pressure, diastolic (mm Hg) 88 83 81 84 82 77 ECG abnormalitya 43.0 46.4 37.2 36.4 29.1 29.3 Cholesterol (mg/100 ml) 171 163 147 177 175 165 Triglycerides (mEq/l) 4.7 3.6 3.1 4.0 3.5 3.0 Glucose (mg/100 ml) 134 111 112 121 107 118 Uric acid (mg/100 ml) 6.8 6.2 6.7 5.2 5.5 4.8 a
Percentage of persons with electrocardiographic examinations who manifested one or more items classifiable under the Minnesota Code, except for subclasses 9–4 and 9–8. Source: Reprinted with permission from American Journal of Epidemiology, Vol 98, p 167, © The Johns Hopkins University School of Hygiene and Public Health.
46894_CH16_503_532.qxd
2/26/10
1:00 PM
Page 523
PARTICULATE AIR POLLUTION 523
Molecules
Virus
0.05 mm
0.01mm
Bacteria
0.5 mm
0.1mm
RBC
Cell
Pollen
5 mm
Hair
50 mm
10 mm
1mm
Pin
Limit of Vision
100 mm
PM10 Thoracic Particles
PM1022.5 Coarse Fraction
PM2.5 Fine Particles
UFP(PM0.1) Ultrafine Particles
Figure 16-9 Particulate Matter Air Pollution Size Distribution. Source: Reprinted with permission from Circulation, Vol 109, RD Brook, B Franklin, W Cascio, et al., p 2657, © 2004, American Heart Association, Inc.
trates the spectrum of particle sizes as to their relative frequencies and correspondence to familiar biologic objects. Many sources of these particles are commonplace in various locations, and their concentration in ambient air is monitored in the United States especially through the National Mortality and Morbidity Air Pollution Study (NMAPS), which collects data from 90 of the largest US cities. PM2.5 is distributed widely on a regional basis but can also be found in more localized areas of concentration because of smelters or other point sources or in “street canyons” in large cities. Mechanisms Both acute and chronic effects of PM are related to coronary events, both arrhythmias and acute myocardial infarction. Figure 16-10 outlines the mechanisms underlying these effects. Pulmonary reflexes lead to autonomic nervous system stimulation with effects on cardiac electrophysiology. Pulmonary inflammation leads to the familiar consequences of systemic inflammation, which may also occur directly in reaction to exposure to PM. Progression of atherosclerosis, promotion of plaque rupture, and thrombosis all follow from this general systemic effect. The AHA report addressed other sources of toxic exposure, including second-hand smoke (see Chapter 14, “Smoking and Other Tobacco Use”), summarizing studies of both long-term and short-term effects. The two major studies in the United States are the Harvard Six Cities Study and the American Cancer Society ACS Cancer Prevention II Study (CPS-II). The former established an independent relation of air pollu-
tion exposure specifically to cardiovascular mortality, and the latter estimated increases in risk of ischemic heart disease death and arrhythmias of 18% and 13%, respectively, for every 10 g/m3 increase in long-term exposure. Despite inherent difficulties in exposure measurement and data interpretation, it has been concluded that the observed relations are real and, if anything, underestimated. Evidence for Cardiovascular Effects In reviewing studies available as of 2004, Pope concluded that several lines of evidence from epidemiologic studies supported the conclusion that fine-particle air pollution is an important risk factor for cardiovascular morbidity and mortality:54, p 490 1. Increases in both respiratory and cardiovascular disease deaths during and immediately following pollution episodes. 2. Associations between daily changes in particulate pollution and cardiovascular and respiratory deaths and hospitalizations. 3. Increased risk of adult cardiopulmonary disease mortality and increased risk of postneonatal infant mortality associated with spatial differences in ambient fine particulate pollution concentrations. In the setting of the Women’s Health Initiative (WHI) observational study, nearly 66,000 postmenopausal women without previous cardiovascular disease and residing in one of 36 metropolitan areas in the United States were followed for an average of
46894_CH16_503_532.qxd
524
2/26/10
1:00 PM
Page 524
CHAPTER 16 SOCIAL AND PHYSICAL ENVIRONMENT
Ambient PM Pulmonary Reflexes
?
Autonomic Nervous System Automaticity Conduction Repolarization
Heart Rate Rhythm
?
Pulmonary Inflammation
Systemic Inflammation
Oxidant Stress
Endothelial Dysfunction
Leukocyte & Platelet Activation
Atherosclerosis Progression and Plaque Instability
Acute Phase Response & Coagulation Factors
Plaque Rupture
Thrombosis
Arrhythmia
Myocardial Infarction
Figure 16-10 Possible Biological Mechanisms Linking PM with Cardiovascular Disease. Source: Reprinted with permission from Circulation, Vol 109, RD Brook, B Franklin, W Cascio, et al., p 2663, © 2004, American Heart Association, Inc.
6 years.55 Monitoring data from the station nearest the reported place of residence were used to link exposure with follow-up status, which included 1806 fatal or nonfatal cardiovascular events. Results are illustrated in Figure 16-11. Within the confidence bands for each analysis (overall, between-city, and within-city effects), which widen at the extremes of exposure, the solid line indicates a generally linear increase in relative risk of cardiovascular death with increasing exposure. For specific categories of events, the increase per 10 g/m3 increase in long-term exposure was 24% for any event, 76% for cardiovascular death, and 35% for cerebrovascular events. As Dockery, director of the Harvard Six Cities Study, commented in an editorial accompanying the WHI report, the extensive data on demographic, behavioral, and risk factors in the WHI study, further analysis may provide insight into factors increasing susceptibility to effects of ambient air pollution.56 In the Multi-Ethnic Study of Atherosclerosis (MESA), linkage of 20-year monitoring data with findings of subclinical atherosclerosis showed a weak positive association with carotid artery intimal-medial thickness, but no consistent relation to coronary calcium or ankle-brachial index.57 The CPS-II, begun in 1982, reported in 2009 on findings linking PM2.5 air pollution with mortality through December 31, 2000, among the CPS-II cohort of 1.2 million persons.58 In the nationwide analysis, for every 10 g/m3 increase in exposure, the hazard ratio for ischemic heart disease death increased by 24%, the strongest relation for any specific cause of
death. This report reflected newer methodologic approaches to exposure estimation and analysis and extended conclusions beyond those of a previous report. It remains to be determined whether critical time windows over the years-long exposure history indicate periods of increased vulnerability to PM2.5 air pollution, and continuing advances in exposure estimation and modeling techniques remains to be evaluated.
NEIGHBORHOOD CHARACTERISTICS Measurement Characteristics of neighborhoods have been investigated with interest in their contribution to health behaviors and health status, some of their findings being included in Chapters 8 and 9, “Dietary Imbalance” and “Physical Inactivity,” presented previously. Variation among studies in availability or use of indicators of neighborhood assets, liabilities, or other properties presents difficulty in comparing or synthesizing observations. An aid to studies in this area is a compilation of data sources for US communities across a wide range of dimensions of both social and physical environments.59 An example of such studies, in which methods were standardized, was a comparison of towns in the Czech Republic and Germany regarding the relation of socioeconomic indicators of neighborhoods to prevalence of obesity, hypertension, smoking, and physical inactivity.60 Unemployment and overcrowding, based on local administrative data, were used for
46894_CH16_503_532.qxd
2/26/10
1:00 PM
Page 525
NEIGHBORHOOD CHARACTERISTICS 525
Between-City Effect 12 11 10 9 8 7 6 5 4 3 2 1 0
Within-City Effect
Relative Risk of Death from Cardiovascular Disease
Relative Risk of Death from Cardiovascular Disease
Relative Risk of Death from Cardiovascular Disease
Overall Effect 12 11 10 9 8 7 6 5 4 3 2 1 0
12 11 10 9 8 7 6 5 4 3 2 1 0
0 3 6 9 12 15 18 21 24 27 30
0 3 6 9 12 15 18 21 24 27 30
0 3 6 9 12 15 18 21 24 27 30
0 3 6 9 12 15 18 21 24 27 30 PM2.5 (g/m3)
0 3 6 9 12 15 18 21 24 27 30 PM2.5 (g/m3)
0 3 6 9 12 15 18 21 24 27 30 PM2.5 (g/m3)
(A)
(B)
(C)
Figure 16-11 Level of Exposure to Fine Particulate Matter and the Risk of Death from Cardiovascular Causes in Women. The graphs demonstrate the observed relationship between the risk of death from cardiovascular disease and the level of particulate matter of less than 2.5 µm in aerodynamic diameter (PM2.5), including both definite and possible deaths from coronary heart disease or cerebrovascular disease. Panel A shows the overall relationship between the PM2.5 level and death, Panel B the effects between metropolitan areas, and Panel C the effects within metropolitan areas, with an indicator variable used to adjust for each city. These results suggest a generally linear relationship between exposure and risk, though the 95% confidence intervals (shaded areas) are wide at the extreme of exposure. Risk is depicted in comparison with a reference value of 11 µg per cubic meter. The histogram in each panel illustrates the density of exposure distribution for air pollution. All estimates are adjusted for age, race, or ethnic group, educational level, household income, smoking status, systolic blood pressure, body-mass index, and presence or absence of a history of diabetes, hypertension, or hypercholesterolemia. Source: Reprinted with permission from The New England Journal of Medicine, Vol 356, KA Miller, DS Siscovick, L Sheppard, et al., p 453, © 2007 Massachusetts Medical Society.
classification of towns, and baseline screening for new cohort studies provided the outcome data. Neighborhood status was most strongly associated with smoking, after adjusting for individual-level education, and low physical activity was also associated with neighborhood characteristics in both countries. Links of Environment with Cardiovascular Disease Neighborhood characteristics related to diet and physical activity have been of particular interest, within a broader range of social, behavioral, and health concerns. Diez Roux proposed a schematic representation of links across these dimensions of residential neighborhoods, as shown in Figure 16-12. Aspects of physical and social environments were seen as influencing physical activity, stress and psychosocial factors, diet, smoking, and sleep disturbance and stress. These factors in turn, whether arising from the physical or social environment or both, would bear on proximate biological risk factors (blood pressure, body mass index, diabetes, blood lipids, stress response, and others) and ultimately clinical cardiovascular disease. Notably, she identified air pollution as leading to clinical disease
through mechanisms of inflammation, endothelial function, heart rate variability, and arrhythmia. Identifying challenges to future research, she emphasized defining the appropriate geographic scale for characterizing environmental features, choice of residential, work, or school environments, and measurement of relevant features. Developments in geographic information systems (GIS) were noted as promising for future studies. The observational nature of studies and lack of feasibility of experimental designs add to the importance of evaluating changes in built environments when these occur as a result of community actions, commercial interests, or government policy. Environmental Influences on Diet and Physical Activity Diez Roux and colleagues in the Atherosclerosis Risk in Communities (ARIC) Study addressed the question of whether neighborhood of residence, characterized by indicators of wealth and income, education, and occupation, differed in incidence of coronary heart disease.61 Residential communities of ARIC
46894_CH16_503_532.qxd
526
2/26/10
1:00 PM
Page 526
CHAPTER 16 SOCIAL AND PHYSICAL ENVIRONMENT
Physical Environment Accessibility of recreational resources
Sport and leisure time physical activity
Transportation, sidewalks, bike lanes Design of public spaces Land use, density, street connectivity, and urban form Aesthetic quality
Walking
Stress and psychosocial factors
Availability and relative cost of “healthy” foods
Diet
Food and tobacco advertising
Smoking
Proximate Biological Factors Blood pressure Body mass index Clinical Diabetes cardiovascular Blood lipids disease Stress response Others
Availability of tobacco Noise
Sleep disturbance and stress Inflammation Endothelial function Heart rate variability Arrhythmia
Air pollution Social Environment Safety and violence Social support and cohesion
Stress and psychosocial factors
Social norms
Behaviors (diet, physical activity, smoking)
Figure 16-12 Schematic Representation of Possible Pathways Linking Residential Environments to Cardiovascular Risk. Source: Reprinted with permission from Journal of Urban Health: Bulletin of the New York Academy of Medicine, Vol 80, AV Diez Roux, p 572, © 2003 The New York Academy of Medicine.
participants in Maryland, Minnesota, Mississippi, and North Carolina were defined as aggregates of census-defined blocks of an average of 1000 persons. During follow-up of more than 13,000 participants for an average of 9.1 years, 615 coronary events occurred. Residence in the most disadvantaged versus the most advantaged neighborhoods was associated with a hazard ratio of 3.1 among Whites and 2.5 among Blacks, after controlling for personal income, education, and occupation. In neither group were these results changed by adjustment for established coronary risk factors. The likelihood that personal factors combine with environmental characteristics in determining behavior led to study of the joint effects of these influences on walking in accordance with recommendations.62 It
was demonstrated in an Australian community, as shown in Table 16-7, that both favorable physical factors and individual factors contributed to meeting recommended levels of walking. The greatest impact was, perhaps not surprisingly, when both sets of factors were optimum. The policy implication may be less obvious, however, that achieving recommended behavior to the greatest degree is unlikely if either set of factors is overlooked in policy development. Addressing environmental influences on both eating and physical activity, French and others addressed issues regarding food supply, eating out, physical activity, and inactivity, as well as effects of advertising, promotion, and pricing on these behaviors.63 Here both social and physical environment are considered as, for practical purposes, they need to be. The range
Table 16-7
Joint Influence of Individual and Physical Environmental Correlates on Walking as Recommended (n 1743)a,b Physical Environmental Factors Individual Factors Low Medium High % OR (95% CI) % OR (95% CI) % OR (95% CI) Low 8.7 1.00 16.0 2.00 (1.09–3.69) 21.7 3.29 (1.77–6.09) Medium 19.7 2.76 (1.54–4.93) 20.4 2.90 (1.61–5.22) 26.0 3.52 (1.95–6.35) High 27.9 4.55 (2.56–8.11) 29.4 4.49 (2.47–8.15) 41.6 7.83 (4.41–13.91) Defined as 30 min sessions of walking 5 days/week. Adjusted for age, sex, children under 18 at home, education, household income, work outside home, SES of area of residence and social environmental factors.
a
b
Source: Reprinted with permission from Journal of Science and Medicine in Sport, Vol 9, B Giles-Corti, p 359, © Sports Medicine Australia. Published by Elsevier Ltd.
46894_CH16_503_532.qxd
2/26/10
1:00 PM
Page 527
CURRENT ISSUES 527
of topics addressed indicates the complexity of determinants of these behaviors, but it also identifies numerous potential points of intervention. The review concluded with an inventory of environmental strategies and policy recommendations, whose focus was obesity prevention. Areas of action under which specific interventions were proposed were: community organization/action, financial and economic incentives, food assistance programs, food packaging and labeling, media and advertising, schools and worksites, and transportation and urban/rural development. Research was called for in each of these areas to determine feasibility, acceptability, and effectiveness of the proposed approaches. A further example of focus on the physical environment is a review by Humpel and others that identified 19 quantitative studies of factors influencing physical activity.64 Factors considered in the review were accessibility of facilities, opportunities for activity, weather, safety, and aesthetic attributes. Several items were identified across the group of studies that were pertinent to these categories of influence. Weather and safety were less strongly associated with physical activity than were accessibility, opportunity, and aesthetics. An accompanying commentary by Sallis noted that a parallel literature in the fields of transportation and urban planning finds walking to be determined by neighborhood design.65 Places where walking to shops is feasible, in contrast to automobile-dependent designs, are accompanied by more frequent walking and biking. Investment in research in this area by the Robert Wood Johnson Foundation is supporting a program called Active Living Policy and Environmental Studies, beginning with research to improve environmental measurement relevant to physical activity.
CURRENT ISSUES A broad range of concepts in cardiovascular epidemiology and prevention has been addressed here, and a correspondingly broad perspective on current issues seems necessary. Some of the larger questions arising in the context of the social and physical environment, social status and social conditions, and social determinants of disease can be subsumed under issues of theory, practice, and further research and noted briefly. Theory The argument is compelling that aspects of social organization and the ways in which individuals are af-
fected by it influence health, both personally and collectively. Specificity of health consequences would not be expected from influences so general as social status and social conditions, which rather would increase the likelihood of suffering what limitations of health are prevalent in a given society at a given time. The particular conditions that afflict a society at a point in history—such as atherosclerotic and hypertensive diseases throughout most of the world’s population today—appear to be determined jointly by the social factors addressed here that amplify the impact of the particular risk factors for a given disease. From this point of view, recognizing the social determinants of disease does not obviate exploiting knowledge of major risk factors to prevent these cardiovascular diseases nor, conversely, does knowledge of the risk factors suffice to achieve sustainable improvement in health if social determinants remain unchanged. What old afflictions remain or new ones arise will replace cardiovascular diseases, equally potentiated by the persisting inequities in society. Practice If the theory simplistically expressed above is sound, its implications for practice in prevention of cardiovascular diseases seem clear: Actions with promise of reducing the force of both social determinants and specific risk factors are needed, in concert. The time scale is an important consideration. Death and disability from cardiovascular diseases will come soon to many who are now at increased risk, and increased risk will come soon to many or most in societies where risk factors are already, or increasingly, epidemic. It seems inescapable that significantly reducing the effects of the social determinants will take long enough for a generation or more of many populations to suffer the now-prevalent diseases. For these reasons, action to reduce or prevent risk should not await progress in ameliorating the social determinants. At the same time, addressing the social determinants should move forward with the greatest feasible support, both to reinforce efforts specific to cardiovascular disease prevention, now, and to minimize the impact of the other chronic diseases and sources of disease and disability, for the future. Research Of the many questions already posed regarding the way forward in improving population health through reducing or eliminating social inequities, those that have priority might be: What actions, based on present knowledge, have the greatest promise of improving cardiovascular health, now and for the future?
46894_CH16_503_532.qxd
528
2/26/10
1:00 PM
Page 528
CHAPTER 16 SOCIAL AND PHYSICAL ENVIRONMENT
How can those actions be initiated and sustained, on a scale sufficient to have measurable impact? What potential impact can those actions be expected to have on health, beyond specific cardiovascular outcomes? How can those actions be judged, through sufficiently rigorous design and evaluation, as to their benefits, adverse effects, and cost-effectiveness? And who is accountable to sponsor, support, and advocate for this research and to ensure that its results are translated into policy development and implementation? REFERENCES 1. World Health Organization. Commission on Social Determinants of Health. WHO/EPI/EQH/ 01/2006. Geneva (Switzerland): World Health Organization; 2006.
10. Sorlie PD, Johnson NJ, Backlund E. Socioeconomic status and cardiovascular disease mortality: National Longitudinal Mortality Study. In: Report of the Conference on Socioeconomic Status and Cardiovascular Health and Disease, November 6–7, 1995. Bethesda, MD: National Heart, Lung and Blood Institute, National Institutes of Health; 1995:23–26. 11. Kennedy BP, Kawachi I, Prothrow-Stith D. Income distribution and mortality: cross sectional ecological study of the Robin Hood index in the United States. Br Med J. 1996; 312:1004–1007.
2. Committee for the Study of the Future of Public Health. The Future of Public Health. Division of Health Care Services, Institute of Medicine. Washington, DC: National Academy Press; 1988.
12. Anand SS, Razak F, Davis AD, et al. Social disadvantage and cardiovascular disease: development of an index and analysis of age, sex and ethnicity effects. Int J Epidemiol. 2006;35:1239–1245.
3. Hinkle LE. Ecological observations of the relation of physical illness, mental illness, and the social environment. Psychosom Med. 1961;23: 289–296.
13. Unnatural causes … is inequality making us sick? Available at http://www.unnatural causes.org. Accessed July 25, 2009.
4. Reeder LG. Occupation and socioeconomic status as variables in heart disease research: a critique. Am J Med Sci. 1959;238:297–307.
14. Murray CJ, Kulkarni SC, Michaud, C, et al. Eight Americas: investigating mortality disparities across races, counties, and race-counties in the United States. PLoS Med. 2006 September;3(9):e260.
5. Hinkle LE, Whitney H, Lehman EW, et al. Occupation, education, and coronary heart disease. Science. 1968;161:238–246. 6. Marmot MG, Kogevinas M, Elston MA. Social/economic status and disease. Ann Rev Public Health. 1987;8:111–135. 7. Marmot MG, Adelstein AM, Robinson N, Rose GA. Changing social-class distribution of heart disease. Br Med J. 1978;2:1109–1112. 8. Kunst AE, Groenhof F, Mackenbach JP, Health EW. Occupational class and cause specific mortality in middle aged men in 11 European countries: comparison of population based studies. EU Working Group on Socioeconomic Inequalities in Health. BMJ. May 30, 1998;316 (7145):1636–1642. 9. Lenfant C. Conference on Socioeconomic Status and Cardiovascular Health and Disease. Circulation. 1996;94: 2041–2044.
15. Kaplan GA, Keil JE. Socioeconomic factors and cardiovascular disease: a review of the literature. Circulation. 1993;88:1973–1998. 16. Link BG, Phelan J. Social conditions as fundamental causes of disease. J Health Soc Behav. 1995;36:80–94. 17. Phelan JC, Link BG, Diez-Roux A, Kawachi I, Levin B. “Fundamental causes” of social inequalities in mortality: a test of the theory. J Health Soc Behav. Sep 2004;45(3):265–285. 18. Marmot M. The Status Syndrome. New York, NY: Henry Holt and Company; 2004. 19. Pollitt RA, Rose KM, Kaufman JS. Evaluating the evidence for models of life course socioeconomic factors and cardiovascular outcomes: a systematic review. BMC Public Health. Jan 20, 2005;5:7.
46894_CH16_503_532.qxd
2/26/10
1:00 PM
Page 529
REFERENCES 529
20. Cassel J, Patrick R, Jenkins D. Epidemiological analysis of the health implications of culture change: a conceptual model. Ann NY Acad Sci. 1960;84:938–949. 21. Cassel J, Tyroler HA. Epidemiological studies of cultural change. Arch Environ Health. 1961;13:31–39. 22. Tyroler HA, Cassel J. Health consequences of cultural change—II. J Chronic Dis. 1964;17: 167–177. 23. Syme SL, Hyman MM, Enterline PE. Cultural mobility and the occurrence of coronary heart disease. J Health Hum Behav. 1965;6:178–189. 24. Scotch NA. Sociocultural factors in the epidemiology of Zulu hypertension. Am J Pub Health. 1963;53:1205–1213. 25. Epstein FH, Eckoff Rd. The epidemiology of high blood pressure––geographic distributions and etiologic factors. In: Stamler S, Stamler R, Pullman TN. The Epidemiology of Hypertension. New York, NY: Grune & Stratton; 1967:155–166. 26. Stamler J. Established major coronary risk factors. In: Marmot M, Elliott P, eds. Coronary Heart Disease Epidemiology: From Aetiology to Public Health. Oxford (England): Oxford University Press; 1992:35–66. 27. Omran AR. Epidemiologic transition in the United States: the health factor in population change. Popul Bull. 1977;32:1–42. 28. Pearson TA, Jamison DT, Trejo-Gutierrez J. Cardiovascular disease. In: Jamison DT, Mosley WH, Measham AR, Bobadilla JL, eds. Disease Control Priorities in Developing Countries. Oxford, England: Oxford University Press; 1993:577–594.
Diseases, Injuries, and Risk Factors in 1990 and Projected to 2020. The Harvard School of Public Health, Boston; 1996. 31. Lopez AD, Mathers CD, Ezzati M, Jamison DT, Murray CJL, eds. Global Burden of Disease and Risk Factors. The International Bank for Reconstruction and Development/ The World Bank, Washington, DC; 2006. 32. Jamison DT, Mosley WH, Measham AR, Bobadilla JL, eds. Disease Control Priorities in Developing Countries. Oxford (England): Oxford University Press; 1993 33. Jamison DT, Breman JG, Measham AR, et al., eds. Disease Control Priorities in Developing Countries. 2nd ed. International Bank for Reconstruction and Development/The World Bank. Washington, DC; 2006. 34. Forsdahl A. Are poor living conditions in childhood and adolescence an important risk factor for arteriosclerotic heart disease? Br J Prev Soc Med. 1977;31:91–95. 35. Barker DJP. The origins of coronary heart disease in early life. In: Henry CJK, Ulijaszek SJ, eds. Long-Term Consequences of Early Environment. Cambridge (England): Cambridge University Press; 1996:155–162. 36. Barker DJP, ed. Fetal and Infant Origins of Adult Disease. London: British Medical Journal; 1992. 37. Barker DJP. Mothers, Babies and Health in Later Life. Edinburgh (UK): Churchill Livingstone; 1998. 38. Barker DJP, Osmond C. 1: Infant mortality, childhood nutrition, and ischaemic heart disease in England and Wales. In: Barker DJP, ed. Fetal and Infant Origins of Adult Disease. London: British Medical Journal; 1992:23–37.
29. Trowell H, Burkitt D. Joint enquiry into Western diseases. In: Trowell HC, Burkitt DP, eds. Western Diseases: Their Emergence and Prevention. Cambridge, MA: Harvard University Press; 1981:427–435.
39. Barker DJP, Winter PD, Osmond C, et al. 13: Weight in infancy from ischaemic heart disease. In: Barker DJP, ed. Fetal and Infant Origins of Adult Disease. London: British Medical Journal; 1992:141–149.
30. Murray CJL, Lopez AD, eds. The Global Burden of Disease: A Comprehensive Assessment of Mortality and Disability from
40. Fall CHD, Osmond C, Barker DJP, et al. Fetal and infant growth and cardiovascular risk factors in women. Br Med J. 1995;310:428–432.
46894_CH16_503_532.qxd
530
2/26/10
1:00 PM
Page 530
CHAPTER 16 SOCIAL AND PHYSICAL ENVIRONMENT
41. Barker DJ, Osmond C, Forsen TJ, Kajantie E, Eriksson JG. Trajectories of growth among children who have coronary events as adults. N Engl J Med. Oct 27, 2005;353(17):1802–1809. 42. Elford J, Shaper AG, Whincup P. Early life experience and cardiovascular disease— ecological studies. J Epidemiol Community Health. 1992;46:1–11. 43. Elford J, Whincup P, Shaper AG. Early life experience and adult cardiovascular disease: longitudinal and case-control studies. Int J Epidemiol. 1991;20:833–844. 44. Paneth N, Susser M. Editorial. Early origin of coronary heart disease (the “Barker hypothesis”). Br Med J. 1995; 310:411–412. 45. Kelleher CC, Lynch JW, Daly L, et al. The “Americanisation” of migrants: evidence for the contribution of ethnicity, social deprivation, lifestyle and life-course processes to the mid-20th century coronary heart disease epidemic in the US. Soc Sci Med. 2006;63: 465–484. 46. Marmot MG, Syme SL. Acculturation and coronary heart disease in Japanese-Americans. J Epidemiol. 1976;104:225–247. 47. Poulter N, Khaw KT, Hopwood BEC, et al. Blood pressure and its correlates in an African tribe in urban and rural environments. J Epidem Commun Health. 1984;38:181–186. 48. Poulter N, Khaw KT, Hopwood BEC, et al. The Kenyan Luo migration study: observations on the initiation of a rise in blood pressure. BMJ. 1990;300:967–972. 49. Poulter N. Blood pressure in urban and rural East Africa: the Kenyan Luo Migrant Study. In: Cruikshank JK, Beevers DG, eds. Ethnic Factors in Health and Disease. London, (UK): Wright, 1989:61–68. 50. Singh GK, Siahpush M. Increasing inequalities in all-cause and cardiovascular mortality among US adults aged 25–64 years by area socioeconomic status, 1969–1998. Int J Epidemiol. Jun 2002;31(3):600–613. 51. Labarthe D, Reed D, Brody J, Stallones R. Health effects of modernization in Palau. Am J Epidemiol. 1973;98:161–174.
52. Gupta R, Gupta VP, Ahluwalia NS. Educational status, coronary heart disease, and coronary risk factor prevalence in a rural population of India. Br Med J. 1994; 309: 1332–1336. 53. Brook RD, Franklin B, Cascio W, et al. Air pollution and cardiovascular disease. A statement for healthcare professionals from the Expert Panel on Population and Prevention Science of the American Heart Association. Circulation. 2004;109:2655–2671. 54. Pope III CA. Air pollution. In: Marmot M, Elliott P, eds. Coronary Heart Disease Epidemiology: From Aetiology to Public Health. 2nd ed. Oxford (England): Oxford University Press; 2005:480–494. 55. Miller KA, Siscovick DS, Sheppard L, et al. Long-term exposure to air pollution and incidence of cardiovascular events in women. New Engl J Med. 2007;356:447–458. 56. Dockery DW, Stone PH. Cardiovascular risks from fine particulate air pollution. N Engl J Med. 2007;356:511–513. 57. Diez Roux AV, Auchincloss AH, Franklin TG, et al. Long-term exposure to ambient particulate matter and prevalence of subclinical atherosclerosis in the Multi-Ethnic Study of Atherosclerosis. Am J Epidemiol. 2008;167: 667–675. 58. Krewski D, Jerrett M, Burnett RT, et al. Extended follow-up and spatial analysis of the American Cancer Society Study linking particulate air pollution and mortality. Health Effects Institute Research Report 140. Boston, MA: Health Effects Institute; 2009. 59. Hillemeier H, Lynch J, Harper S, Casper M. Data Set Directory of Social Determinants of Health at the Local Level. Atlanta: US Department of Health and Human Services, Centers for Disease Control and Prevention; 2004. 60. Dragano N, Bobak M, Wege N, et al. Neighbourhood socioeconomic status and cardiovascular risk factors: a multilevel analysis of none cities in the Czech Republic and Germany. BMC Public Health. 2007;7:255. doi:10.1186/ 1471-2458-7-255.
46894_CH16_503_532.qxd
2/26/10
1:00 PM
Page 531
REFERENCES 531
61. Diez Roux AV, Merkin SS, Arnett D, et al. Neighborhood of residence and incidence of coronary heart disease. N Engl J Med. 2001; 345:99–106.
64. Humpel N, Owen N, Leslie E. Environmental factors associated with adults’ participation in physical activity. A review. Am J Prev Med. 2002;22:188–199.
62. Giles-Corti B. People or places: What should be the target? J Sci Med Sport. 2006;9:357–366.
65. Sallis JF, Kraft K, Linton LS. How the environment shapes physical activity. A transdisciplinary research agenda. Am J Prev Med. 2002; 22:208.
63. French SA, Story M, Jeffery RW. Environmental influences on eating and physical activity. Annu Rev Public Health. 2001;22:309–335.
46894_CH16_503_532.qxd
2/26/10
1:00 PM
Page 532
46894_CH17_533_550.qxd
2/25/10
4:13 PM
Page 533
Causation and Prevention: Theory, Practice, and Further Research
533
P A R T
4
46894_CH17_533_550.qxd
2/25/10
4:13 PM
Page 534
46894_CH17_533_550.qxd
2/25/10
4:13 PM
Page 535
C H A P T E R
17 What Causes Cardiovascular Diseases? however reached—do have important practical consequences for decisions on treatment guidelines or prevention policies, discussed in subsequent chapters.
SUMMARY As a bridge between the evidence discussed in the foregoing chapters and considerations of policy and practice that follow, this chapter addresses concepts of causation and approaches to causal judgment; views on causation of the atherosclerotic and hypertensive diseases; and the fundamental distinction between two complementary perspectives on these topics— one theoretical, the other practical. The natural history of atherosclerosis and hypertension concerns not only their major outcomes—coronary heart disease, stroke, and several related conditions—but their main determinants as well. The latter include both individual characteristics—genes and family history, patterns of behavior, metabolic and physiologic conditions, and psychosocial traits—and population-wide social and environmental conditions. The idea is implicit throughout the preceding chapters that many of the characteristics and conditions discussed here are part of the causation of cardiovascular diseases. But the question “What causes cardiovascular diseases—atherosclerosis, hypertension, and their cardiovascular outcomes?” warrants more direct consideration. It has more than one answer. This is due in part to a difference between theoretical and practical views of the question, one comprehensive and the other selective. Each view is important, and the two are complementary. One view encompasses a complex “n-dimensional” framework for causal understanding, whereas the other limits consideration to those few factors most important for a practical framework for prevention. Differences in concepts of causation may also lead to different answers and to sometimes contentious debate. Such differences notwithstanding, causal judgments—
INTRODUCTION The atherosclerotic and hypertensive diseases— foremost, coronary heart disease, stroke, and related conditions—exist on a global scale and account for an immense burden of morbidity and mortality. The loss of life and well-being, enormous medical expenditures and lost productivity, and disparities within and among nations make prevention of these conditions a paramount public health priority. Knowledge of their determinants includes innumerable connections between characteristics of individuals and their risks and between social and environmental conditions of populations and their rates of disease and risk distributions. Two questions follow: Do we know what causes these diseases? Is this knowledge sufficient for public health action? As background to subsequent chapters that address strategies of prevention and a plan for public health action, this discussion addresses the first of these questions—how evidence about causation is understood and what, as a result, is known. The second question is addressed in the chapters that follow. A Retrospective The British Medical Journal for December 29, 1909, reported: “Professor Osler delivered an address on arterio-sclerosis. Though there were sixty-two theories of its causation, he thought the three main factors were time, tension, and toxins.”1, p 1800 “Time” denoted the
535
46894_CH17_533_550.qxd
536
2/25/10
4:13 PM
Page 536
CHAPTER 17 WHAT CAUSES CARDIOVASCULAR DISEASES?
concept of atheromata as a manifestation of senility, at age 70 years, although heredity might explain appearance of these lesions at much earlier ages, including “fatty degeneration” in the aorta in childhood. “Tension” comprised both “the tension of life,” exemplified by a man working in the New York Stock Exchange, and “muscular tension, due to overexertion.” “Toxins” were both “exogenous” (excess of food, alcohol, tobacco, tea, and coffee) and “endogenous” (“waste products which irritated the endothelium and kept up a high tension”). Osler’s recommendation for prevention was reported as follows: “For those with tendencies towards arteriosclerosis the guiding motto was: ‘Nothing too much’—the life of the tortoise, not that of the hare.” His concluding remarks were paraphrased, enigmatically: “Success was largely a matter of survivorship.” From 62 “theories,” Osler chose three “main factors” in formulating a recommendation for prevention. Nearly a century after his assessment, the number of “theories” or “factors” implicated in the causation of atherosclerotic and hypertensive diseases has multiplied several-fold. Even greater selectivity is necessary today to identify what can be considered as “main factors.” A Current Account In 1981, Hopkins and Williams published results of a survey of the literature by which they found 246 factors associated with coronary heart disease, by the criterion of an odds ratio or risk ratio statistically significant at p 0.05.2 No recent repetition of this exercise has been published, but many more qualifying associations would be expected today. The question “Do we know what causes these diseases?” now calls for explanation of a multifaceted disease process that has been investigated extensively at several levels: Among populations and communities, it exhibits marked variation in morbidity and mortality with sharply divergent trends; within families, it clusters and occurs more often among relatives than expected from the population at large; within individuals, it progresses predictably from early indicators of risk to fully developed disease; it affects specific regions of the circulation, especially the heart, brain, carotid arteries, lower extremities, and aorta, with characteristic derangements and lesions in the arterial wall; structural and functional variation at cell surfaces are implicated; and genes, regulatory proteins, lipid transport molecules, and other molecular phenomena are involved. The range of potential explanatory factors also extends across many levels—from determinants of national and global policies for land use and food production to genetic variation and influences on
gene expression. These factors include intrinsic personal characteristics such as age, sex, race or ethnicity, and heredity; individual patterns of behavior, such as diet, physical activity, and use of tobacco and alcohol; physiologic and metabolic processes, such as regulation of lipid metabolism, blood pressure, blood coagulation, and glucose concentration and insulin activity; psychosocial factors; and extrinsic social and environmental conditions. Some of these factors overlap or interact, and for each factor another underlying set of its own determinants has been established or is proposed. It is reasonable to ask whether, in principle, an intelligible understanding of causation of atherosclerotic and hypertensive diseases should fully accommodate this complexity—all aspects of the disease process, each of the explanatory factors, and their determinants. Would such a comprehensive formulation, if achievable, be informative? A counter-example is found in recent companion essays by Stamler and colleagues on “six established major risk factors.”3,4 Here the emphasis is on “adverse diet, diet-related above-optimal levels of serum total cholesterol (TC) and blood pressure (BP), overweight/ obesity, diabetes mellitus (DM), and cigarette smoking.”3, p 19 Selection of these factors as “established” is justified on the basis of extensive multidisciplinary findings that have “demonstrated their significant role in the etiology of epidemic CHD”; they are “major” on grounds of high prevalence, strong impact on risk, and potential for being prevented or reversed.3, p 20 The choice may be akin to Osler’s, in which the myriad acknowledged factors are reduced to a manageable few. Does this highly selective approach to addressing the question “What causes these diseases?” lack credibility for absence of the wealth of information regarding other factors? If not, this may speak to the utility of a pragmatic view of causation, in the service of preventive policies and practices. A Public Perception Mixed views about causation are not confined to the scientific community. As Figure 17-1 illustrates, public confidence in medical news may suffer when seemingly random items are daily fare.5 Here, whether researchers, the journals, or the media (note the newscaster’s control over the “random” choice) contribute more to this problem is left to the viewer’s prejudices.
CAUSAL JUDGMENT However many factors may be raised for consideration, which if any among them is properly considered a cause? Causation is a fundamental concept in
46894_CH17_533_550.qxd
2/25/10
4:13 PM
Page 537
CAUSAL JUDGMENT 537
Figure 17-1 “Today’s Random Medical News”: A Reflection of Differences in Understanding of Causation. Source: Copyright © Universal Press Syndicate.
Problem: Incomplete and Multifactorial Causation A fundamental problem in studying causation of the atherosclerotic and hypertensive diseases (and other chronic diseases as well) is the fact of incomplete correspondence between exposure and disease, or between cause and effect: Presence of a factor, or “exposure,” does not invariably result in disease, and presence of the disease does not invariably indicate presence of the factor, or “exposure.” A graphic representation of this concept shows the population as a whole in the largest circle (Figure 17-2). Within the circle is a subset of the population who share some particular exposure. Another distinct subset shares a particular disease in common. Some members of the population are in both of these subsets, and therefore these inner circles overlap to some degree. This
generally occurring situation might be termed “incomplete causation,” which is characteristic of the chronic diseases as we know them. The subsets of the population within the circle are the same four categories familiar to many as those in the conventional “2 2” or “fourfold” table— those with both exposure and disease (a), those with neither exposure nor disease (d), and those with either exposure or disease but not the other (b) and
Exposure and Disease Disease 1 2
b
a
c Disease
Exposure
Exposure
science and is central to epidemiology and public health. Acceptance of epidemiologic findings as reflecting causation, when appropriate, is a critical step in translating research into policy and practice. When differences in understanding and interpretation of causation arise, not surprisingly, they can foster controversy. Decisions regarding preventive policies and practices may be impeded as a result.
1
a
b
2
c
d
d (a) In the Population
(b) In the 2 3 2 Table
Figure 17-2 Exposure and Disease (a) In the Population (b) In the 2x2 Table.
46894_CH17_533_550.qxd
538
2/25/10
4:13 PM
Page 538
CHAPTER 17 WHAT CAUSES CARDIOVASCULAR DISEASES?
(c). To measure the quantitative relation among these four groups within a population is the objective of the “analytic” epidemiologic study, whether through a cohort or case-control design. This simplest representation illustrates the starting point of evaluating the relation between exposure and disease that can be followed in brief outline to its logical conclusion. It is not difficult to imagine how the counts of a, b, c, and d would be expected to relate to one another if the overlap between circles of exposure and disease were, on the one hand, mere coincidence or, on the other hand, great enough to suggest to the suspecting mind that there could be a causal connection between them. Applying an appropriate statistical procedure to evaluate this overlap, or possible connection, would be to test for evidence in the data of an “association” between exposure and disease. A rule is needed in advance for deciding the outcome, whether to declare that an association is present or absent. But even a decision in favor of association leads only part of the way toward answering the question of causation—because association is not equivalent to causation. Given association, three questions follow: (1) Does the exposure cause the disease?, (2) How else could the apparent connection be explained?, and (3) How does one decide among the available explanations? The conclusion regarding question (1) depends heavily on whether the alternatives evaluated under question (2) can be eliminated. In what might be called “the differential diagnosis of causation” (by analogy to the clinical concept of formulating, then ruling out, competing diagnostic possibilities) the task is to evaluate each of three broad types of explanation: chance, bias or error, and causation. The tools of statistical evaluation can provide an estimate of the probability that a particular result would be observed if there were truly no connection between exposure and disease. Beyond evaluation of chance, the differential diagnosis must consider “error.” This refers to error in a special sense. It includes flaws in design, conduct, or analysis of a study. Error extends to issues of possibly unknown as well as known characteristics of the groups of individuals whose exposure or disease status is being compared. At issue in part is whether this comparison suffers from any systematic error, or bias, that would be expected to create an artificial picture suggesting association (or concealing one). Because data needed to evaluate potential biases may be unavailable, questions of bias may remain unresolved. Error also concerns understanding of the possible roles of factors other than the exposure of interest in the apparent relation between that exposure
and the disease. Here the concern is whether association between the exposure and disease appears only because both are connected with an intermediate factor, whose presence is the more direct explanation of the apparent relation—such an intermediate factor is termed a confounding variable, or simply confounder. Imagination may offer many possible confounders, some or all of which may not have been addressed in a given study and therefore cannot be taken into account. This leaves the question of error of this kind in some doubt. If the differential diagnosis to this point has rejected chance and diminished error as the basis for association, it remains to address the remaining explanation, causation, more directly. Solution: Development of Causal Thinking In Causal Thinking in the Health Sciences: Concepts and Strategies in Epidemiology, Susser in 1973 traced the background of causal thinking and elaborated conceptual models applicable in epidemiology, including aspects of judgment that finally enter into most, if not all, questions of causation in the health sciences.6 His continuing reflections on this topic were published nearly 20 years later, in 1991, “What Is a Cause and How Do We Know One? A Grammar for Pragmatic Epidemiology.”7 The value of this essay derives especially from its refinement of considerations raised in coming to judgments about causation and its differentiation among types of epidemiologic studies regarding the distinct contribution of each to these judgments. Formal application of Susser’s “grammar” to the causation of the atherosclerotic and hypertensive diseases could be a rewarding, if monumental, task. A second, very accessible, general approach to causal thinking is that of MacMahon and Pugh, whose 1970 text posed the following question:8 What, then, leads one to think of certain relationships as causal? The word cause is an abstract noun and, like beauty, will have different meanings in different contexts. No definition will be equally appropriate in all branches of science. Epidemiology has the practical purpose of discovering relations which offer possibilities of disease prevention and for this purpose a causal association may usefully be defined as an association between categories of events or characteristics in which an alteration in the frequency or quality of one category is followed by a change in the other. . . . [T]he idea of changeability is a basic component of epidemiologic concepts of causation.8, pp 18–19
46894_CH17_533_550.qxd
2/25/10
4:13 PM
Page 539
CAUSAL JUDGMENT 539
The implications of this view of causation are considered in the conclusion of this chapter. At this point, it offers an appropriately broad perspective from which to continue the discussion. In a simple world, cause and effect might seem self-evident and straightforward. The lighted match is quickly and confidently identified as the source of pain in the finger that is exposed. A higher level of complexity, but similarly simple conclusion, is represented by the statement of John Stuart Mill: The cause, then, philosophically speaking, is the sum total of the conditions positive and negative taken together; the whole of the contingencies of every description, which being realized, the consequent invariably follows.9 All questions about this view of causation aside, the implied invariability of causation is intellectually appealing. This concept was retained, for example, in the well-known Henle-Koch postulates, formulated more than a century ago to establish causality in the relation between a parasite and a disease (as quoted by Evans):10 1. The parasite occurs in every case of the disease in question. 2. It occurs in no other disease as a fortuitous and nonpathogenic parasite. 3. After being fully isolated from the body and repeatedly grown in pure culture, it can induce the disease anew.10, p 177 But by the 1950s, this approach was seen to pose difficulties in the area of chronic diseases. The common chronic diseases were thought to be due not to a single, unique exposure, but to a combination of factors. In addition, presence of those factors might not always result in the expected disease. Occurrence of disease without the agent and presence of the agent without the disease clearly defied the basic tenets of Henle and Koch, as well as the principle stated by Mill. Yerushalmy and Palmer, Lilienfeld, and Sartwell each contributed their views to the Journal of Chronic Diseases in 1959 on ways to overcome the limitations of the Henle-Koch postulates.11–13 For example, a “characteristic” or “suspected factor” should simply be found more frequently among persons with than without disease, and persons with a possibly causal factor should develop disease more frequently than those without. These and other “guideposts” and “criteria” addressed specificity in the relation between a particular exposure and disease, the relation of duration and intensity of exposure to incidence of disease, and several other properties of the evidence. There emerged from these first speculations an approach to causal interpretation that has remained influential for more than four decades.
A Conventional Approach Two especially prominent references on the practice of causal analysis are Smoking and Health: Report of the Advisory Committee to the Surgeon General of the Public Health Service, 1964, and a paper by Sir Austin Bradford Hill, The environment and disease: association or causation?, 1965.14,15 The approach to evaluating epidemiologic associations presented in these two sources was closely dependent on the prior ideas of Sartwell and others. Referring to “criteria for judgment” by which the causal significance of an association could be assessed, the 1964 report to the Surgeon General listed five relevant properties of an association: consistency, strength, specificity, temporal relation, and coherence. Soon after, in 1965, Hill elaborated a list of nine properties to be addressed, once an association has met standards of statistical significance, independence from other factors, and absence of explanation by spurious relationships. To the preceding elements Hill added four: biological gradient, plausibility, experiment, and analogy. The nine elements were described as follows: “strength” of the relationship as measured by magnitude of the measure of association (e.g., the relative risk or odds ratio); “consistency” of the association, being found in studies of different populations and under different methods and circumstances of investigation; “specificity,” or the uniqueness or rarity of other associations involving the same factor or disease; “temporality,” or unequivocal evidence concerning presence of the factor prior to onset of the disease; “biological gradient,” or variation in strength occurring with variation in degree or intensity of exposure (dose-response curve); “plausibility,” in relation to current biological knowledge; “coherence,” or conformity with general knowledge about the disease itself; “experiment,” or change in disease outcome following change in exposure; and “analogy,” or prior finding of a similar exposure-disease relationship. Hill added: Here, then, are nine different viewpoints from all of which we should study association before we cry causation. What I do not believe––and this has been suggested––is that we can usefully lay down some hard-and-fast rules of evidence that must be obeyed before we accept cause and effect. None of my nine viewpoints can bring indisputable evidence for or against the causeand-effect hypothesis and none can be required as a sine qua non. What they can do, with greater or less strength, is to help us make up our minds on the fundamental question––is there
46894_CH17_533_550.qxd
540
2/25/10
4:13 PM
Page 540
CHAPTER 17 WHAT CAUSES CARDIOVASCULAR DISEASES?
any other ways of explaining the set of facts before us, is there any other answer equally, or more, likely than cause and effect?15, p 299 With some elaboration on these “considerations” (and not “criteria”) assembled by Hill and now very widely applied in causal analysis, it is clear that this final step in differential diagnosis of association—like the preceding ones—depends on the expertise of the epidemiologist. The science and art of epidemiology are required to identify and evaluate relevant evidence by objective and rigorous methods and, on the basis of that evaluation, to come to a reasoned judgment that can be supported on the strength of the logic as well as the underlying observations themselves. Causal conclusions can be reached with reasonable confidence on the basis of explicit procedures for first taking into account the details of every study that contributes to the relevant body of evidence; second, carrying out the differential diagnosis described above; and third, weighing the results to reach an overall judgment. The outcome of sound causal analysis is a reasoned conclusion that takes into account the strengths and limitations of the available evidence and of the evaluation itself, such that any variation in conclusions from one analysis to another should, in principle, be identifiable and soluble.
CAUSAL CONSTRUCTS Beyond judgment about a causal role for any one factor is development of constructs or models to incorporate multiple factors thought to determine the actual occurrence of disease. Many examples of such constructs, linking one or another of the determinants with cardiovascular disease outcomes, are illustrated in preceding chapters. Other examples have been intended to encompass one or more components of the cardiovascular diseases in a comprehensive way. How to consider multiple causes in an integrated framework has itself been a development in causal thinking. From Single Agent to “n-Dimensional Complex” The idea that a number of preconditions are required for relevant exposures to occur and produce disease led to the metaphor of the “chain of causation.” Links in the chain can be discovered through appropriate investigation and potentially broken through targeted intervention. It is unnecessary to identify every link if one or more critical ones can be identified and their connections interrupted, as in immunization to prevent infection when an otherwise effective exposure occurs. Notably, the discovery of each new link re-
quires further research to connect it—at both ends— with links already known. MacMahon and Pugh found the chain analogy insufficient to account for interactions among many antecedent factors in the causation of disease.8 According to MacMahon (personal communication, 1996), Pugh deserves the credit for proposing the “web of causation” to fill this need, having in MacMahon’s words “invented it out of whole cloth.” With this more accommodating metaphor, they could construct a twodimensional graphic scheme of the causation of icterus (jaundice) as a sign of serum hepatitis resulting from use of contaminated syringes in treatment of persons with syphilis. Their observations about this web are instructive for the present discussion: When it is considered that only a few of the major components are shown, that these are indicated as broad classes of events rather than as the multiple minor events which make up each class, that each component shown is itself the result of a complex genealogy of antecedents, and that the myriad effects of these components other than those contributing to the development of icterus are not shown, then it becomes evident that the chains of causation represent only a fraction of the reality, and the whole genealogy may be thought of more appropriately as a web, which in its complexity and origins lies quite beyond our understanding. Fortunately it is not necessary to understand causal mechanisms in their entirety to effect preventive measures. Knowledge of even one small component may allow significant degrees of prevention.8, pp 23–24 A subsequent contribution was Rothman’s scheme, presented in Figure 17-3, illustrating what he described as “sufficient causal complexes, each having five component causes,” more recently termed
Sufficient Cause I
Sufficient Cause II
D
E A
C B
Sufficient Cause III
G
H A
F B
I
J A
F C
Figure 17-3 Conceptual Scheme for the Causes of a Hypothetical Disease. Source: Reprinted with permission, American Journal of Epidemiology, Vol 104, KJ Rothman, p 589, © The Johns Hopkins University School of Public Health, 1976.
46894_CH17_533_550.qxd
2/25/10
4:13 PM
Page 541
CAUSAL CONSTRUCTS 541
“the causal pie model.”16,17 The underlying concept was that illness is caused by compounds of multiple conditions that, together, are sufficient to produce the illness; that various combinations of component causes might produce the same illness; and that the multiple sufficient causes (here I, II, and III) need not include any component cause in common (although in this illustration cause A, being common to all, may be a “necessary cause”). Rothman noted, as did MacMahon and Pugh, that knowledge of the total makeup of a sufficient cause is unnecessary for prevention, “in that blocking the causal role of but one component of a sufficient cause renders the joint action of the other components insufficient, and prevents the effect.”16, p 588 Stallones expanded the discussion of causal frameworks further.18 He suggested a view of causality extending beyond the two-dimensional web to encompass the possibility of multiple interrelated outcomes, such as coronary heart disease and stroke, rather than only one as in MacMahon’s and Pugh’s example: Whatever constructs we may devise must be understood to represent our biased views of what a representation of reality should be. Because it is not the reality, the value of the model depends upon its utility, and utility depends upon the purpose for which the model is used. Rothman has spoken of a constellation of component causes [citation to Rothman, 1976], but I think his concept is closely akin to the web of causation model, in which the causes converge upon an effect. If the purpose is more global than studying one disease at a time, then the multiple regression/web of causation model is inadequate. An approach that holds promise is to consider the interdependence of a number of diseases, characteristics of individuals, and environmental and social variables as elements in a constellation which is n-dimensional, and within which directed pathways are incidental to the complex as a whole.18, p 76 This approach is attractive in the present context, with many recognized relevant exposures and multiple outcomes. This is so in spite of sharing the inherent incompleteness of the “web.” If the full ramifications of the web were largely unknown, those of the “n-dimensional complex” would be more so. These views of causality indicate progressive concern for comprehensiveness, and increasingly they risk loss of direct practical utility as a result. On the one hand, encyclopedic inclusiveness is of value for intellectual purposes aimed at complete scientific un-
derstanding. On the other, practicality implies identification of “directed pathways,” in Stallones’s terms. Particular pathways or specific factors can be selected for attention on the basis of their relative importance, perhaps in terms of utility for developing preventive strategies, as indicated by MacMahon and Pugh. Scientists working in different regions of the complex may well see causation differently. This likelihood threatens common understanding among scientists, counter to the interest in integrating knowledge across disciplines and areas of investigation. Two illustrations show how far apart two such approaches can be, one beginning from the gene, the other from population differences in rates of disease occurrence. From Molecules to Populations Sing and colleagues addressed the problem of linking DNA variations with interindividual variation in risk of coronary artery disease.19 Three levels of analysis were represented schematically as shown in Figure 7-2. First is the genotype, second is the network of intermediate traits, and third is the probability of developing coronary artery disease. Multiple genes contribute to each of the intermediate traits, which are regulatory phenomena illustrated by hemostasis, lipid and carbohydrate metabolism, and blood pressure regulation. (Each “trait,” of course, is itself a complex regulatory system, the number of whose recognized component processes and determinants expands continually with advancing knowledge.) The resultant of these four traits leads to varying probabilities of expressing coronary artery disease, conditional on environment and age. This figure contrasts with several figures in previous chapters, as those figures are specific to one or another single factor and equal or greater detail was presented for that factor alone. As a grand summary, then, this scheme goes considerably further in linking genes, through intermediate phenotypes, to individual probabilities of disease, though omitting the voluminous detail already established at each level. Even so, this scheme goes only part way toward the full range of considerations important for public health. It omits the level of population differences, which for Rose was the cardinal point of concern.20 In “Sick Individuals and Sick Populations,” he emphasized causes of two kinds whose distinction was critical: causes of incidence and causes of cases. The distinction is analogous to that between the factors required to cause an epidemic of a disease and the specific agent established as causing individual cases of that disease. For example, epidemics of an infectious disease known to occur sporadically, that is, in seemingly isolated, infrequent cases, require additional
46894_CH17_533_550.qxd
542
2/25/10
4:13 PM
Page 542
CHAPTER 17 WHAT CAUSES CARDIOVASCULAR DISEASES?
explanatory factors beyond the infecting organism itself. Thus, causes of incidence operate to produce differences in disease rates between populations or within a population over time, whereas causes of cases account only for differences in risk among individuals within a population at a particular time. From the public health viewpoint, Rose argued, the primary question is why a disease is common in one population and rare in another; to focus only, or primarily, on differences in individual risks may fail to identify the most potent, population-wide influences on the occurrence of disease and the furthestreaching preventive measures. To give primary consideration to utility of particular factors for public health intervention offers a rational approach for their selection, out of the extensive complex of causes, to receive particular attention. This approach reflects a common view that narrowly construed causal pathways are most useful in epidemiology; that is, they have the greatest practical applicability to disease prevention and control. The value of the more comprehensive frameworks must also be fully appreciated, however. From the perspective of interdisciplinary research, the most inclusive array of causal factors may be the most likely to foster innovative collaborations and stimulate studies across levels of understanding. For example, a potential link between the scheme of Sing and others and the population perspective of Rose would be to investigate especially those genetic factors thought to differ in distribution among populations that differ in rates of coronary artery disease. The potential role for genetic epidemiology in this aspect of research on cardiovascular diseases seems clear from this perspective.
SOCIETIES
COMMUNITIES
FAMILIES
Causes versus Mechanisms and Etiology versus Pathogenesis The n-dimensional complex and web of causation are both intended to include all phenomena related to the production of disease, from all levels of observation. The array of relevant levels is represented in Figure 17-4, adapted from Stallones’s original figure to suggest an appropriate connection for contemporary epidemiology, beyond that of traditional epidemiology, across the full spectrum.18 Accordingly, for example, details of the intrinsic and extrinsic pathways of blood coagulation are incorporated in the same framework as differential trends in coronary heart disease mortality between countries during the 20th century. The sharpness of the distinction often made between research on outcomes and research on mechanisms, the latter being construed as exclusively “fundamental,” is greatly diminished in this perspective. All observations, whether in laboratory, clinical, or population settings, gain plausibility from demonstration of linkages with the remainder of the complex. Those most firmly linked in a number of directed pathways would contribute most importantly to understanding of the process as a whole. Etiology refers literally to the study of causation but is often equated with causation itself. Distinction between etiology and pathogenesis is also often blurred, the latter referring strictly to the development or progression of a disease once it has been caused. Whether the relation of a factor associated with occurrence of the atherosclerotic and hypertensive diseases is thought to concern etiology or pathogenesis is rarely made explicit. The multiple outcomes at issue invite inconsistencies. For example, the etiology of coronary heart disease may be considered as operat-
INDIVIDUALS
ORGANS
CELLS
MOLECULAR AND SUBMOLECULAR PARTICLES
Traditional Epidemiology Clinical Research Pathology, Physiology Cell Biology Molecular Biology Contemporary Epidemiology
Figure 17-4 The Biomedical and Community Health Sciences Arrayed on a Scale of Biological Organization. Source: Adapted with permission from the Annual Review of Public Health, Vol 1, © 1980, by Annual Reviews, Inc.
46894_CH17_533_550.qxd
2/25/10
4:13 PM
Page 543
CAUSATION OF ATHEROSCLEROTIC AND HYPERTENSIVE DISEASES 543
ing at the onset of atherosclerosis in childhood or at the moment of precipitation of an acute myocardial infarction as the first clinical manifestation of the disease. What is meant by causation in the present context, then, should be understood as the complex of factors that operate at any point from prior to the first manifestation of atherosclerosis or hypertension to the last, perhaps fatal, step in the course of disease.
CAUSATION OF ATHEROSCLEROTIC AND HYPERTENSIVE DISEASES Risk Factors The earliest known statement on prevention of coronary heart disease and stroke was that of the National Health Education Committee, Inc., published in 1959: A Statement on Arteriosclerosis, Main Cause of “Heart Attacks” and “Strokes.”21 The conditions addressed in that report as contributing or predisposing factors are discussed in several preceding chapters: overweight, elevated blood cholesterol level, elevated blood pressure, excessive cigarette smoking, and heredity. The Statement also noted that moderate physical activity appeared to reduce the hazards of arteriosclerosis. The connection of each factor with the disease was supported by reference to studies published in the 1940s and 1950s. The now commonplace term risk factor was introduced in 1961 in a report by Kannel, Dawber, Kagan, Revotskie, and Stokes on the first 6 years of experience in the Framingham Study.22 (Authority for this claim rests with the late Frederick Epstein, whose diligent and thoroughly reliable scholarship revealed no prior use of the term, according to informed sources [Kannel WB, Higgins I, Higgins M, personal communication, 1996].) Referring to serum cholesterol concentration, blood pressure, and elec-
trocardiographic evidence of left ventricular hypertrophy, Kannel and colleagues wrote: Combinations of the three risk factors under consideration appear to augment further the risk of subsequent development of coronary heart disease. It has been demonstrated . . . that the incidence of coronary heart disease rises progressively as these factors are combined. . . . Whether or not the correction of these abnormalities once they are discovered will favorably alter the risk of development of disease, while reasonable to contemplate and perhaps attempt, remains to be demonstrated. . . . As additional longitudinal observations are made, it is hoped that additional risk factors will be determined. This will allow further identification of susceptible individuals and hopefully suggest methods of control.23, pp 47–48 Emphasis was placed on the evidence in the Framingham Study that these conditions preceded the appearance of coronary heart disease, in contrast to findings of prior clinical studies among existing cases of the disease. However, demonstration of change in outcome with change in exposure (MacMahon and Pugh’s later “changeability”) was clearly not a requirement. Further, no distinction was made between left ventricular hypertrophy, a sign of existing cardiac disease commonly associated with long-standing high blood pressure, and elevated blood pressure itself, notwithstanding the presumed sequential relation between these two factors. The authors’ hope that additional risk factors might be identified was amply fulfilled and documented by Hopkins and Williams, who reported 20 years later on a survey of coronary risk factors.2 The 246 diverse factors reviewed and referenced in detail are categorized in Table 17-1. The criterion for
Table 17-1
Coronary Heart Disease Risk Factors and Suggested Associations Category Constitutional and demographic factors Environmental exposure Habits, lifestyle, and psychosocial factors Physical and biochemical measurements Serum or blood measurements Platelet and coagulation tests Medical conditions Dietary excesses and positive associations Dietary deficiencies and inverse associations or possible protective factors Drug liabilities Total
No. 16 5 54 16 44 16 45 21 23 6 246
% 6.5 2.0 22.0 6.5 17.9 6.5 18.3 8.5 9.3 2.4 100.0
Source: Data from PN Hopkins and RR Williams, A Survey of 246 Suggested Coronary Risk Factors, Atherosclerosis, Vol 40, pp 1–52, © 1981.
46894_CH17_533_550.qxd
544
2/25/10
4:13 PM
Page 544
CHAPTER 17 WHAT CAUSES CARDIOVASCULAR DISEASES?
inclusion of each factor was at least one publication that suggested either its direct or inverse association with occurrence of coronary heart disease. The value of this heroic exercise was lost on two especially vocal critics of the risk-factor concept, McCormick and Skrbanek,23 who took the number of 246 factors as sufficient evidence of the fallacy of this approach to understanding causation, a view shared by some others. The further elaboration of many of these factors and the numerous determinants of these factors in the ensuing 25 years has doubtless multiplied the count of candidate factors, thereby greatly expanding the n-dimensional complex of causation.
taken on several occasions. The linkage of genes, intermediate phenotypes, and coronary artery disease as shown in Figure 7-2 is one example.19 Greater detail at the intermediate level, and multiple possible outcomes anticipating the n-dimensional concept, was presented by Stallones in the 1960s (Figure 17-5).24 Outcomes affecting the heart, brain, and arterial circulation were distinguished, as were subtypes of coronary and cerebral vascular diseases. Contributions of hemostatic function, atherosclerosis, and hypertension were placed closest to these outcomes as “immediate precursors,” and a network of associated factors was arrayed more distant from the outcomes. (This figure is said to have prompted Stallones to claim the “St. Sebastian award” for the number of arrows featured in the scheme [Reed DM, personal communication, 1996].)
Causal Pathways Synthesis of the plethora of suggested or identified risk factors into a coherent picture has been under-
DIET
STRESS
FAT HORMONES
TOTAL CALORIES
SALT
PHYSICAL ACTIVITY
HEREDITY
SMOKING HYPERLIPIDEMIA
? COAGULATION CLOT LYSIS
MYOCARDIAL INFARCTION
ANGINA PECTORIS
CORONARY HEART DISEASE
ATHEROSCLEROSIS
THROMBOSIS
HYPERTENSION
HEMORRHAGE
CEREBRAL VASCULAR DISEASE
HYPERTENSIVE DISEASE
Figure 17-5 Relations Between Major Clinical Components of the Cardiovascular Diseases, Their Immediate Precursors, and Selected Associated Factors of Epidemiologic Concern. Source: Reprinted from National Heart, Lung and Blood Institute, National Institutes of Health, Bethesda, Maryland.
46894_CH17_533_550.qxd
2/25/10
4:13 PM
Page 545
CAUSATION OF ATHEROSCLEROTIC AND HYPERTENSIVE DISEASES 545
Hopkins and Williams, in turn, attempted to organize selected factors among the 246 they reviewed into a causal framework for heart attack (Figure 17-6).2 Their approach was based on a concept of pathogenesis of atherosclerosis progressing from the earliest endothelial abnormalities to cholesterol deposition in the arterial wall, increased likelihood or risk of thrombosis, and precipitation of acute ischemia or arrhythmia. Factors especially linked with these successive phases of atherosclerosis were termed, respectively, initiators, promoters, potentiators, and precipitators. In stating that “the major CHD risk
factor tends to have multiple roles in atherogenesis,”2, p 18 the authors reflected the view that importance of a given factor was dependent on its recurrent appearance in multiple causal pathways. For example, from Figure 20-6, this approach would especially implicate high-fat diet and cigarette smoking. Another approach to synthesis of current understanding into a practical picture of causation was taken by Pearson and colleagues (Figure 17-7).25 Their review was presented in the context of the projected cardiovascular disease experience of developing countries in the decades ahead. Multiple end points were
Initiators Initiate atherogenesis by undermining the integrity of the arterial endothelium
Blood pressure Serum cholesterol (>300 mg/dl) Diabetes Homocysteinemia Cigarettes (carbon monoxide?) High-fat diet (free fatty acids?) Cortisol (Type-A behavior?)
Serum LDL or VLDL Serum HDL High-saturated fat diet Diabetes, insulin resistance Atherogenic lipoproteins Dietary chromium or silicon? Dietary vitamins B6 or C?
Promoters Promote the buildup of cholesterol in the artery wall
Potentiators Increase the likelihood of a microor macrothrombus, thereby potentiating the effects of both initiators and promoters
Precipitators Precipitate a heart attack by inducing acute ischemic events or arrhythmias
Cigarette smoking Oral contraceptives High saturated fat diet Platelet adherence/aggregability Platelet factors III or IV Thromboxane A2, prostacyclin Low fibrinolytic activity
Physical exertion Cold weather All potentiators Cigarettes RBC Agglutination Catecholamines Dietary magnesium?
HEART ATTACK
Note: HDL, high-density lipoprotein; LDL, low-density lipoprotein; RBC, red blood cell; VLDL, very–low-density lipoprotein.
Figure 17-6 What Causes Heart Attacks? A Synthesis by Hopkins and Williams, 1981. Source: Reprinted with permission from Atherosclerosis, Vol 40, PN Hopkins, RR Williams, p 18, © 1981, with kind permission from Elsevier Science Ltd, Bay 15K, Shannon Industrial Estate, Co. Clare, Ireland.
46894_CH17_533_550.qxd
546
2/25/10
4:13 PM
Page 546
CHAPTER 17 WHAT CAUSES CARDIOVASCULAR DISEASES?
End points
Nonmodifiable risk factors Physiological risk factors
Hypertensive heart disease
Age Male sex Family history
Hypertension
Elevated LDL cholesterol Obesity Decreased LDL cholesterol
Hemorrhagic stroke
Coronary heart disease Atherothrombotic stroke
Behavioral risk factors Diabetes
Sedentary lifestyle
Peripheral vascular disease
Diet • Saturated fat • Salt • Cholesterol • Total energy content Heavy alcohol consumption Smoking
Note: LDL, low-density lipoprotein.
Figure 17-7 Relationship Among Risk Factors for Cardiovascular Disease. Source: Reprinted with permission from Disease Control Priorities in Developing Countries, edited by DT Jamison et al., Copyright © 1993, The International Bank for Reconstruction and Development/The World Bank. Used with permission of Oxford University Press.
recognized, as in the scheme shown previously in Figure 17-4, with segregation of primarily hypertensive from primarily atherosclerotic outcomes in the heart and brain and the addition of peripheral vascular disease. These conditions were not differentiated in their relation to the risk factors, however, because the three broad categories—nonmodifiable, behavioral, and physiological risk factors—all bear on the end points jointly in this scheme. Partition of the risk factors in this way was intended to suggest different aspects for policy development. For example, the behavioral risk factors could be addressed through health education and the physiological ones through medical interventions. Interactions among these factors, especially the most important factors, were acknowledged, however, and the multifactorial nature of cardiovascular diseases was seen as providing “a rich and complex range of possibilities in the development of preventive strategies.”25, p 580 “The Present Reality” The concept that many factors relate to the occurrence of atherosclerotic and hypertensive diseases and their several outcomes is supported by decades of laboratory, clinical, and population research. Because
the disease process is protracted and multifaceted, even the discovery of one factor, such as a viral infection that was critical in triggering the onset of atherosclerosis, would be insufficient to explain its varied rates of progression and frequencies of outcomes in different populations and over time. Nor would it seem that programming in fetal or early neonatal life (see Chapter 16) would suffice to dispose of all subsequent influences as being unimportant to progression and outcomes. The number of factors and determinants accepted as causally related to this disease process can be expected to increase rather than to diminish as presently suspected factors are investigated in more intricate detail. Identified candidate genes will probably proliferate in number and may lead to recognition of a great many genetic variants and intermediate environmental factors interacting to affect individuals who have the genotype in question. The concept of the n-dimensional complex seems logically necessary to accommodate current and anticipated knowledge of causal factors. In addition, it can have a salutary effect in bringing into a common framework research findings from multiple levels and scientific disciplines, thereby potentially facilitating communication among
46894_CH17_533_550.qxd
2/25/10
4:13 PM
Page 547
CURRENT ISSUES 547
workers in different specific areas. For advancement of understanding, the richer the multifactorial array, the better. Thus, for the scientific and intellectual purpose of epidemiology, this is an essential view of causation. Still, against the criterion of utility noted by MacMahon and Pugh, Stallones, and others, a causal scheme that accommodates all valid observations fails as a direct consequence of its comprehensiveness. To formulate practical implications from such extensive knowledge of causation requires a process of judicious selection. Assembly of the selected elements that lead to well-founded and coherent prevention policies requires emphasis on factors considered most fundamental to the disease process, amenable to intervention, beneficial to public health, and free of undue risks. In the view of Rose, the interventions best justified on these grounds are those tending to restore “biological normality.”20 From the discussions of preceding chapters, two cardinal examples would be reestablishment of patterns of diet and physical activity that are associated with low rates of atherosclerotic and hypertensive diseases and that are also characteristic of the human species throughout most of our existence. Elimination of toxic exposures, principally tobacco smoke, is a third such example. The ramifications of preventive strategies based on these three areas of intervention alone would clearly affect many of the intermediate factors and mechanisms recognized as being associated with disease progression and outcomes. Additional selected factors are reasonably included in intervention strategies for specific populations or groups or in circumstances that offer special promise of benefit from other approaches. A recent approach to the task of simplifying representation of the causes of atherosclerotic and hypertensive diseases in developing a comprehensive strategy for prevention is shown in Figure 17-8. The figure is the first of several panels in a larger framework taken from A Public Health Action Plan to Prevent Heart Disease and Stroke.26 It is intended as a summary of the determinants of onset and progression of these diseases, identifying six points at which different types of intervention approaches can
be applied. The point of departure is the mix of unfavorable social and environmental conditions that underlie adverse behavioral patterns on the population level. The adverse behavioral patterns, especially regarding dietary imbalance and physical inactivity, lead to development of the major risk factors—mainly adverse blood lipid profile, high blood pressure, obesity, diabetes, and smoking. In the absence of effective control of risk factors once they are present, a first coronary or cerebrovascular event may occur, resulting in sudden death, or in survival, often with disability and with high risk of recurrence. Survivors are likely to die of eventual cardiovascular complications or decompensation. The utility of this representation of the causal pathway will be considered further in Chapter 22.
CONCLUSION In what sense are the “main factors” or “established major risk factors” distinct from the rest of the myriad factors now identified? Evidence for their causal role and their actual or potential changeability is strong; practical means for intervention to modify them have been devised and extensively investigated; and they offer the potential for major impact on the occurrence of atherosclerotic and hypertensive diseases in populations. An understanding of causation that includes a selective emphasis on preventive utility and therefore on the established major risk factors is no less essential to fulfilling the purposes of epidemiology than is the comprehensive causal complex from which it derives. Epidemiology and public health have the luxury as well as the necessity of entertaining both views of causation in pursuit of the complementary purposes of advancing knowledge and serving the health of the public.
CURRENT ISSUES Do we know what causes these diseases? The theoretical view remains open to new knowledge and anticipates that “emerging” factors will continually arise
The Present Reality Unfavorable Social and Environmental Conditions
Figure 17-8 The Present Reality
Adverse Behavioral Patterns
Major Risk Factors
First Event/ Sudden Death
Disability/ Risk of Recurrence
Fatal CVD Complications/Decompensation
46894_CH17_533_550.qxd
548
2/25/10
4:13 PM
Page 548
CHAPTER 17 WHAT CAUSES CARDIOVASCULAR DISEASES?
from further research and take their place in the causal complex. High expectations surround the field of genomics, but anticipation extends to study of inflammation, infection, and other areas as well. The practical view selects as a “directed pathway” a set of “established major risk factors” to be targeted through intervention approaches—across a spectrum from social and environmental determinants and population-wide behavioral patterns to risk factors at the individual level. While the two views can be seen as complementary, some tension exists between them: From the theoretical view, there is a pressing need for further research; from the practical view, needed action lags far behind already-established knowledge. What we know and need to know to take effective public health action, and priorities in public health research, are topics of the following chapters. REFERENCES 1. Arterio-sclerosis. Editorial. Br Med J. December 29, 1909:1800. 2. Hopkins PN, Williams RR. A survey of 246 suggested coronary risk factors. Atherosclerosis. 1981;40:1–52. 3. Stamler J. Established major coronary risk factors: historical review. In: Marmot M, Elliott P, eds. Coronary Heart Disease Epidemiology: From Aetiology to Public Health. 2nd ed. Oxford: Oxford University Press; 2005:18–31. 4. Stamler J, Neaton JD, Garside DB, Daviglus ML. Current status: six established major risk factors—and low risk. In: Marmot M, Elliott P, eds. Coronary Heart Disease Epidemiology from Aetiology to Public Health. 2nd ed. Oxford: Oxford University Press; 2005:32–70. 5. Borgman J. Today’s random medical news from the New England Journal of PanicInducing Gobbledygook (cartoon). Cincinnati Enquirer. 1997. 6. Susser M. Causal Thinking in the Health Sciences: Concepts and Strategies in Epidemiology. New York: Oxford University Press; 1973. 7. Susser M. What is a cause and how do we know one? A grammar for pragmatic epidemiology. Am J Epidemiol. 1991;133:635–648.
8. MacMahon B, Pugh TF. Epidemiology: Principles and Methods. Boston, MA: Little, Brown & Co; 1970. 9. Mill JS. A System of Logic. London: Parker, Son and Bowin, 1856. 10. Evans AS. Causation and disease: the HenleKoch postulates revisited. Yale J Biology Med. 1976;49:175–195. 11. Yerushalmy J, Palmer CE. On the methodology of investigations of etiologic factors in chronic diseases. J Chron Dis. 1959;10:27–40. 12. Lilienfeld A. On the methodology of investigations of etiologic factors in chronic disease— some comments. J Chron Dis. 1959;10:41–46 13. Sartwell PE. On the methodology of investigations of etiologic factors in chronic diseases. Further comments. J Chron Dis. 1959;11: 61–63. 14. Advisory Committee to the Surgeon General of the Public Health Service. Smoking and Health: Report of the Advisory Committee to the Surgeon General. Atlanta, GA: Public Health Service, US Department of Health, Education and Welfare; 1964. 15. Hill AB. The environment and disease: association or causation? Proc Soc Med. 1965;58: 295–300. 16. Rothman KJ. Reviews and commentary. Causes. Am J Epid. 1976;104:587–592. 17. Rothman KJ, Greenland S. Causation and causal inference in epidemiology. Am J Pub Health. 2005;95(suppl 1): S144–S150. 18. Stallones RA. To advance epidemiology. Ann Rev Public Health. 1980;1:69–82. 19. Sing CF, Haviland MB, Templeton AR, et al. Biological complexity and strategies for finding DNA variations responsible for inter-individual variation in risk of a common chronic disease, coronary artery disease. Ann Med. 1992;24: 539–547.
46894_CH17_533_550.qxd
2/25/10
4:13 PM
Page 549
REFERENCES 549
20. Rose G. Sick individuals and sick populations. Int J Epidemiol. 1985;14:32–38. 21. White PD, Wright IS, Sprague HB, et al. A Statement on Arteriosclerosis: Main Cause of “Heart Attacks” and “Strokes.” New York: National Health Education Committee, Inc.; 1959. 22. Kannel WB, Dawber TR, Kagan A, et al. Factors of risk in the development of coronary heart disease—six-year follow-up experience: the Framingham Study. Ann Intern Med. 1961; 55:33–50. 23. McCormick J, Skrbanek P. Coronary heart disease is not preventable by population interventions. Lancet. 1988;ii:839–841.
24. Stallones RA. Prospective epidemiologic studies of cerebrovascular disease. Public Health Monogr. 1966;76:51–55. 25. Pearson TA, Jamison DT, Trejo-Gutierrez J. Cardiovascular disease. In: Jamison DT, Mosely WH, Measham AR, Bobadilla JL, eds. Disease Control Priorities in Developing Countries. Oxford (England): Oxford University Press; 1993:577–594. 26. US Department of Health and Human Services. A Public Health Action Plan to Prevent Heart Disease and Stroke. Atlanta, GA: US Department of Health and Human Services, Centers for Disease Control and Prevention; 2003.
46894_CH17_533_550.qxd
2/25/10
4:13 PM
Page 550
46894_CH18_551_566.qxd
2/25/10
4:14 PM
Page 551
C H A P T E R
18 Strategies of Prevention Discussion of the population-wide and high-risk strategies of CVD prevention anticipates subsequent chapters in which global risk assessment and risk scores, development of recommendations, guidelines and policies, the rationale for a public health strategy for prevention, and plans for taking action are addressed.
SUMMARY Prevention is the pathway from the present reality of progressive loss of cardiovascular health to a future in which cardiovascular health (and health in general) is maintained throughout life. Part IV continues with the concepts and language of prevention. From Rose came clear articulation of the duality of prevention strategies—the “population-wide” and the “high-risk” strategies—and their underlying epidemiologic rationale. From Strasser came the contrasting idea of “primordial prevention”—that is, intervention to protect whole societies from epidemics of the major risk factors themselves. To explore the broad range of opportunities for prevention of cardiovascular diseases, strategies, stages, approaches, and settings are considered. The impacts of different strategies on risk distributions are graphically illustrated. Stages of prevention are outlined, as well as the distinction among four goals of prevention expressed in Healthy People 2010—prevention of risk factors, detection and treatment of risk factors, early identification and treatment of heart attacks and strokes, and prevention of recurrent cardiovascular events. Multiple approaches to prevention are identified—policy and environmental change; population-wide behavior change; risk-factor detection and treatment; emergency care and acute case management; rehabilitation and long-term care; and end-of-life care. Such approaches are considered as elements of a framework for planning and implementing public health strategies for heart disease and stroke prevention. Although these concepts of prevention evolved mainly in industrial countries where epidemic CVD first appeared, they are found applicable in developing countries as well.
INTRODUCTION The landmark 1988 report on the state of official public health agencies in the United States, The Future of Public Health, declared the mission of public health to be “fulfilling society’s interest in assuring conditions in which people can be healthy. Its aim is to generate organized community effort to address the public interest in health by applying scientific and technical knowledge to prevent disease and promote health.”1, p 7 Three core functions were identified for public health agencies at all levels of government: assessment, policy development, and assurance. In fulfilling these functions, public health agencies evaluate the health of the community, develop responsive public health policies through science-based decision making, and assure that services fundamental to the well-being of the population are available. Although the preceding chapters focused on causes of societal health problems, knowledge of means of prevention is also essential and is the subject of this and following chapters. To make the transition from causation to prevention, it is useful to expand the picture of atherosclerotic and hypertensive diseases presented in Figure 17-8 to include with “the present reality” an alternative view—“a vision of the future.” For each
551
46894_CH18_551_566.qxd
552
2/25/10
4:14 PM
Page 552
CHAPTER 18 STRATEGIES OF PREVENTION
element of the present reality, a desired alternative can be displayed: Social and environmental conditions may become more favorable to health; population-wide behavioral patterns may promote health; the distribution of risk in the population may be improved and prevalence of increased risk reduced; incidence of acute coronary or cerebrovascular events may be reduced, with decreased case-fatality and lesser occurrence of sudden death; functional capacity may be preserved or more rapidly restored; recurrence of acute events may become less frequent; and good quality of life may be maintained very nearly until death (Figure 18-1). Central to this scheme is the concept of prevention, representing types of actions for moving from the present reality to this vision of this future. What is prevention? In common usage, according to Webster’s, it is an “act of preventing or hindering; obstruction of action, access, or approach; thwarting; . . . ”2, p 1960 To the extent that these implicitly negative meanings reflect the usual understanding of “prevention,” it should not be surprising that the terms “disease prevention” or “preventive medicine” lack popular appeal and engagement. Here and in subsequent chapters, “prevention” is intended in the positive sense of transforming states of health from conditions of the present reality to those envisioned in a healthier future, as illustrated in Figure 18-1. This sense is conveyed by the language of prevention. It is further reflected in strategies of prevention and the intervention approaches through which such strategies are implemented. Adding these intervention approaches and US national goals for heart
Social and Environmental Conditions Favorable to Health
disease and stroke prevention to the skeleton Figure 18-1 will complete a graphic framework that represents a comprehensive public health strategy. In subsequent chapters the idea, and the ideal, of prevention is examined from several perspectives—those of evidence and decision making for prevention; prevailing preventive recommendations, guidelines and policies; the case to be made for prevention; and actions being urged in order to meet the global challenge of epidemic cardiovascular diseases.
CONCEPTS AND LANGUAGE OF PREVENTION The late Geoffrey Rose’s culminating scientific contribution, The Strategy of Preventive Medicine (1992), begins by recognizing the limited focus of medical practice, which is “largely concerned with responding to the needs of sick individuals.”3, p vii Even when it includes risk identification and disease prevention, medical practice addresses only “a vulnerable minority of individuals.” This approach is inadequate to address the “essential determinants of the health of society. . . its mass characteristics: the deviant minority can only be understood when seen in its societal context, and effective prevention requires changes which involve the population as a whole.” Thus, “The radical strategy is to identify and if possible to remedy the underlying causes of our major health problems.” The thinking of Rose and his contemporaries and subsequent contributions have generated several concepts of prevention and a language of prevention that apply not only to cardiovascular diseases but also to
A Vision of the Future Behavioral Patterns Promote Health
Low Population Risk
Few Events/ Only Rare Deaths
Full Functional Capacity/ Low Risk of Recurrence
Good Quality of Life Until Death
PREVENTION
The Present Reality Unfavorable Social and Environmental Conditions
Adverse Behavioral Patterns
Major Risk Factors
First Event/ Sudden Death
Disability/ Risk of Recurrence
Figure 18-1 PREVENTION—Link Between “The Present Reality” and “A Vision of the Future.”
Fatal CVD Complications/Decompensation
46894_CH18_551_566.qxd
2/25/10
4:14 PM
Page 553
CONCEPTS AND LANGUAGE OF PREVENTION 553
other chronic or noncommunicable diseases as well. The resulting terms are used in varying and sometimes conflicting ways, and there is no special authority to standardize them. It is useful for present purposes to adopt more or less consistent usage, even though cited sources may often differ. A catalogue of terms can be organized under somewhat arbitrary headings of “Strategy,” “Stage,” “Approach,” and “Setting,” as in Table 18-1. Strategy: The distinction between the “population-wide” and “high-risk” strategies was articulated most clearly and effectively by Rose, in his 1981 report, “Strategies of Prevention: Lessons from Cardiovascular Disease.”4 The high-risk strategy refers to the identification and treatment of those at the extreme of risk, the “vulnerable minority”; the populationwide strategy (initially called the “mass strategy” and often simply the “population strategy”) is the “radical strategy” intended to address the “mass characteristics,” the “underlying causes of our major health problems,” noted previously. The population-wide approach targets the whole population because, for example, most coronary heart disease events or strokes in a population with high incidence of atherosclerosis and hypertensive diseases arise from levels of risk factors that are not extreme, and such cases are therefore not preventable by the high-risk approach. (See the discussion that follows in relation to Figure 18-5 and Figure 18-6a-e.) Potential risks and benefits of intervention through each of these strategies—which are seen as complementary, not competing, alternatives—were discussed in Rose’s initial report. This concept of dual approaches to prevention was central to the 1982 Report of a WHO
Table 18-1
Concepts and Language of Prevention
Strategy • Population-wide/high-risk • Health promotion/disease prevention Stage • Primordial/primary/secondary/tertiary Approach • Policy and environmental change/behavior change/risk factor detection and control/emergency care and acute case management/rehabilitation and long-term case management/end-of-life care • Lifestyle change/pharmacotherapy • Single-/multi-factor intervention Setting • Community/worksite/school/healthcare facility/ religious organization
Expert Committee, Prevention of Coronary Heart Disease, and has since been widely adopted.5 Rose subsequently elaborated on this concept in his commentary “Sick Individuals and Sick Populations:”6, p 32 Aetiology confronts two distinct issues: the determinants of individual cases, and the determinants of incidence rate. If exposure to a necessary agent is homogeneous within a population, then case/control and cohort methods will fail to detect it: they will only identify markers of susceptibility. The corresponding strategies in control are the ‘high-risk’ approach, which seeks to protect susceptible individuals, and the population approach, which seeks to control the causes of incidence. The two approaches are not usually in competition, but the prior concern should always be to discover and control the causes of incidence. The duality of “health promotion” and “disease prevention” has a parallel history, beginning with publication in 1979 of Healthy People: The Surgeon General’s Report on Health Promotion and Disease Prevention.7 However, the historical roots of the term “health promotion” are much deeper and are traceable at least to an 1850 report of Lemuel Shattuck on “promotion of public and personal health.” The term is also institutionalized in the name of the International Union for Health Promotion and Education (IUHPE), established some 50 years ago [www.iuhpe.org]. “Health promotion” applies to the broadest societal determinants of health, and “disease prevention” applies to the individual or population-wide manifestations of risk or disease. Health promotion encompasses social values, principally equity in health between and within countries, and a global scale of engagement in advocacy, knowledge development, improvement of health policy and practice, and strengthening capacity for health promotion and health education [www.iuhpe.org]. “Health promotion” has an explicitly positive sense, which further distinguishes this concept from that of disease prevention. However, the two are often juxtaposed in the mission of such organizations as the American College of Preventive Medicine (ACPM), which “seeks to improve population health status through evidence-based disease prevention and health promotion research, policies, practices, and programs” [www.acpm.org]. Stage: Understanding the progression of atherosclerotic and hypertensive diseases from underlying social and environmental conditions, through
46894_CH18_551_566.qxd
554
2/25/10
4:14 PM
Page 554
CHAPTER 18 STRATEGIES OF PREVENTION
population-wide behavioral patterns, to individual risks and events suggests several stages as opportunities for prevention (Figure 21-1). Terms used to distinguish among stages of intervention include “primordial,” “primary,” “secondary,” and (less often) “tertiary” or (still less often) “quaternary” prevention. Mensah and others have described these terms as encountered in several standard references.8 Usages were generally consistent, but difficulties arise when the target of prevention is an intermediate factor in the progression of disease. Thus, although “primary” and “secondary” prevention are widely encountered terms, their use in different contexts conveys different meanings and introduces confusion. When, for example, the focus is on acute coronary or cerebrovascular events, primary prevention means averting these events; in this case, secondary prevention means preventing recurrences among those who have survived the first acute event. Rehabilitation may be included in secondary prevention, although it is often denoted separately. By contrast, when high blood pressure is the focus, primary prevention means preventing this risk factor itself; secondary prevention is rarely used in this context. Whether primary prevention refers to preventing risks or to preventing events must be made explicit to avoid this confusion. “Primordial prevention” is perhaps the least understood and most widely misused term, according to its original definition. The term was introduced by Strasser, of the Cardiovascular Diseases Unit of the World Health Organization, in 1978. 9 Speculating on the future of cardiovascular disease prevention, he noted first the familiar roles of tertiary prevention, by which he referred to rehabilitation after myocardial infarction; secondary prevention, for example, treatment to avert recurrent coronary events; and primary prevention, to prevent the occurrence of clinical events such as myocardial infarction or stroke. He then presented an innovative approach:9, p 228 From the viewpoint of world health for tomorrow, however, one has to go one step further. While the epidemic of risk factors has pervaded the consumer societies, it still has not reached the majority of the developing world. Real grassroot prevention should start by preserving entire risk-factor-free societies from the penetration of risk factor epidemics. Here lies the possibility of averting one of tomorrow’s world health problems. For expressing this important concept, I wish to propose the term of protoprophylaxis or primordial prevention.
“Primordial prevention” is less widely recognized than the high-risk and population strategies, and “proto-prophylaxis” seems not to have caught on at all. This may result from its seeming (but mistaken) irrelevance in Western industrialized societies, which by 1978 were already far advanced in their riskfactor epidemics. At this time developing countries were still regarded as inexperienced with respect to cardiovascular disease. Lack of awareness that cardiovascular diseases were already becoming established there resulted in complacency rather than rapid application of Strasser’s idea, when it could have had greatest impact. Instead, the term primordial prevention is often used to denote intervention early in life to prevent risk-factor development, especially in children. Primordial prevention in this usage conjures a sense of personal developmental stages and the “life course” concept of health promotion and disease prevention. This is the basis for policy that “seeks to promote the well-being of the young, both because of its intrinsic value and because doing so will improve the health of the population at all ages.”10, p 155 “Prevention of risk factors in the first place” was the language of choice to represent this concept in the 1994 report on priorities for cardiovascular disease prevention from the US National Heart, Lung and Blood Institute.11 Terminology aside, the idea of intervention early in life to prevent adult cardiovascular diseases has been strongly advocated, if not as widely practiced.12 Primordial prevention as conceived by Strasser is commonly cited as a strategy of prevention in developing countries. But from the point of view of protecting successive new generations from their own risk-factor epidemics, primordial prevention and the life course idea would converge logically in any population. Primordial prevention therefore has great potential importance everywhere today. Approach is used here to denote one or another specific type of intervention for promoting health or preventing disease. Additional methods could be considered; those mentioned here serve for illustration. “Policy and environmental change” is an approach, or group of approaches, by which public health agencies acting with other sectors of society can intervene to improve the health of communities. Often this is achieved by changing “systems” through which society functions. A 2001 report of the Association of State and Territorial Directors of Health Promotion and Public Health Education (ASTDHPPHE) subtitled “New Directions for Public Health” defines “policies” as “laws, regulations, and rules (both formal and informal)” and “environmental interven-
46894_CH18_551_566.qxd
2/25/10
4:14 PM
Page 555
CONCEPTS AND LANGUAGE OF PREVENTION 555
tions” as “changes to the economic, social, or physical environments.”13, p iii This is a broad approach: “It is becoming increasingly clear that public health practitioners must address these policies, these environments, and the support and obstacles they provide relative to healthy behaviors as the fundamental means of intervention.”13 p ii The “fundamental” nature of this approach reflects its role at the earliest stage of disease development, or most far-reaching potential for societal impact; it is the furthest “upstream” approach in contrast to the “downstream” approaches to individuals. Figure 18-2, from a 2006 review by Brownson and others, illustrates components of environmental and policy change relevant to chronic disease prevention.14 Behavior change refers to population-wide change in attitudes, behaviors, or habits that are deleterious for health or put individuals at risk. This approach relies on health education via communications through the media or other means of public infor-
mation, sometimes targeted to one or more particular groups within a population. “Risk factor detection and control” is an approach to identify members of a population likely to benefit from individual-level intervention and link them with sources of risk-factor management, to reduce already existing risk. Important aspects are: targeting of screening to those most likely to have undetected risk; effective intervention to reduce risk; and long-term follow-up to sustain risk reductions and avert acute coronary or cerebrovascular events. This approach is a major part of “primary prevention” of these events. Emergency care and acute case management address the urgent situation of recognizing and responding to an impending or unfolding clinical event. These involve public awareness of signs and symptoms of possible heart attack or stroke; knowledge of the value of timely communication for emergency transport, as in the United States, through the 9-1-1
Change normative attitudes toward tobacco use, physical activity, and healthy eating
Prevent tobacco use and exposure
Physical Environment/Access • Clean indoor-air policies • Presence of and access to walking/bicycling trails • Access to healthy foods
Increase tobacco use cessation
Economic Environment • Cost of tobacco • Parking cash-outs • Affordable fruits and vegetables
Increase physical activity
Reduce chronic disease morbidity & mortality Increase quality of life
Increase healthy eating
Communication Environment • Make smoking unattractive • Raise awareness of benefits of stair use • Make healthy foods attractive
Improve energy balance
Assemble the necessary resources (funding for programs and policies, a well-trained workforce)
Figure 18-2 Conceptual Framework for Understanding the Prevention of Chronic Diseases Through Environmental and Policy Approaches. Source: Reprinted with permission from Annual Review of Public Health, Vol 27 © 2006 by Annual Reviews. www.annualreviews.org.
46894_CH18_551_566.qxd
556
2/25/10
4:14 PM
Page 556
CHAPTER 18 STRATEGIES OF PREVENTION
telephone system; emergency dispatch of qualified medical service technicians; and availability and access to qualified hospital emergency departments for rapid diagnostic evaluation and treatment. This approach thus denotes a group of interventions that falls between primary and secondary prevention, as it does not address prevention of first events, nor does it directly work to reduce risk of recurrences. “Rehabilitation and long-term case management” concern efforts both to restore functional status in the aftermath of an acute event and to reduce risk of subsequent events among survivors of a first attack. The latter function is clearly secondary prevention. Secondary prevention may also follow detection of inapparent or subclinical CVD, absent the occurrence of an acute event, as through a screening or diagnostic examination, that also warrants long-term case management. End-of-life care is the furthest “downstream” approach, aimed at preventing avoidable adversity due to progression and complications of advanced vascular disease. Because the burdens of disability, dependency, and cost, as well as personal, family, and social issues, intensify as the end of life nears, this often-overlooked phase of cardiovascular diseases warrants emphasis and advance planning. To this effect, the Canadian Cardiovascular Health Strategy and Action Plan illustrates attention to this phase of prevention as “end-oflife planning and care” [www.chhs-scsc.ca]. Lifestyle change and pharmacotherapy are two additional dimensions of prevention at the individual level. Each can have a significant impact on risk in both primary and secondary prevention. “Lifestyle change” typically refers to interventions to improve individual patterns of behavior, especially regarding diet, physical activity, and tobacco use or exposure. It is sometimes intended to include other specific behaviors or habits, such as coping with stress, economic hardship, racism, or other adverse conditions. Each of these types of intervention confronts significant obstacles. For lifestyle change, these include access to needed services, efficacy of available interventions, ability of individuals to make desired choices, and costs. For pharmacotherapy, or more simply use of drugs or medications, issues arise regarding efficacy and safety, access and affordability, long-term adherence and overall benefits versus risks, and others. Lifestyle change is usually considered the first step in intervention to reduce risk, with deferral of drugs until effects of lifestyle change can be evaluated. Even if drugs are added, interventions to promote healthy behavior are intended to continue. “Single-factor” and “multi-factor” intervention distinguish between approaches in which one or multiple factors are considered as the target. The earliest
studies of prevention typically focused on one or another single target, such as blood pressure or cholesterol. Current guidelines and policies more often advocate multifactor, “integrated,” or “global,” assessment of risk and use of multicomponent interventions. Setting: The question where prevention takes place has led to recognition of several types of settings, each with potential opportunities and constraints. “Community” ordinarily refers to a geographic area, perhaps most often to a locality such as a neighborhood, town, or city, including all of its population, organizations, institutions, and the like. In this sense, interventions designed for a community might be more comprehensive than those specifically targeting worksites, schools, healthcare settings, or religious organizations, each of which would have features unique to the environment and population to be addressed. However, in a larger sense “community” may mean the population or citizenry of any place, whether at a local, state, or national level, or a body of persons related though common interests, institutional membership, and so on. Intervention embracing all of these components would be expected to achieve synergy and increased impact over intervention in any one setting alone. What level of social organization is most promising as a focus of prevention? This question is commonly considered in the context of “social determinants of health” or “determinants of population health.” This concept is represented graphically in The Future of the Public’s Health in the 21st Century, a 2002 report from the Institute of Medicine (Figure 18-3).15 This version of a “socioecological model” of population health is adapted from earlier work by Dahlgren and Whitehead and by Worthman.16,17 If the scope of determinants is truly as wide as indicated, the most promising interventions would involve the outer ring— the “broad social, economic, cultural, health, and environmental conditions and policies at the global, national, state, and local levels.”15, p 52 Consideration of “settings” for prevention, then, may entail much more than attention to a particular establishment or institution, or even a whole community, in order to include its larger context. A simplified view of the community can still serve to illustrate relations among components of a broad cardiovascular health strategy. These relations are depicted in something of a “Rubic’s cube” for prevention, presented by Pearson and others in a Scientific Statement of the American Heart Association, in 2003. The intent of the report was to illustrate approaches, beyond the more customary statements regarding secondary and primary prevention, to address a strategy with “the potential not only to prevent the first heart attack or stroke in the person at average risk (a
46894_CH18_551_566.qxd
2/25/10
4:14 PM
Page 557
STRATEGIES OF PREVENTION 557
Bro a
So cia
In
Innate Individual traits: age, sex, race and biological factors --The biology of disease
licies at the b nd po gl s a
ks or tw
r
Over the life span
n itio nd co
d
,
l ta
so c
a
l ia
ral, health, a nd e , cultu mi c nvi ron m k i n r g o w c o d en ndit n a ion ing v s i L nd commun ily a ity ne am f l, b l e a hav u d i io div
o on ec
oba l, na
Living and working conditions may include: • Psychosocial factors • Employment status and occupational factors • Socioeconomic status (income, education, occupation) • The natural and built environmentsc • Public health services • Health care services
l, na t io
sta
an te,
d local levels
Notes: Adapted from Dahlgren and Whitehead, 1991. The dotted lines between levels of the model denote interaction effects between and among the various levels of health determinants (Worthman, 1999). a Social conditions include, but are not limited to: economic inequality, urbanization, mobility, cultural values, attitudes and policies related to discrimination and intolerance on the basis of race, gender, and other differences. b Other conditions at the national level might include major sociopolitical shifts, such as recession, war, and governmental collapse. c The built environment includes transportation, water and sanitation, housing, and other dimensions of urban planning.
Figure 18-3 Social-Ecological Framework: Levels of Influence on Behavior. Source: Reprinted with permission from The Future of the Public’s Health in the 21st Century, Institute of Medicine, National Academy of Sciences, Washington, DC, National Academies Press, © 2003.
population in which large numbers of CVD deaths still occur) but also to avoid the need for intensive and expensive pharmacotherapies to control risk factors such as hypertension, hyperlipidemia, and diabetes, once they become established.”18, p 645 Figure 18-4 presents a proposed conceptual framework for public health practice in CVD prevention, linking components of the community setting with selected essential public health services to improve behaviors, manage specific risk factors when present, and recognize early manifestations of disease.
STRATEGIES OF PREVENTION “Strategies of prevention,” referring to the complementary population-wide and high-risk strategies as described previously, are cited often. A closer examination of their respective outcomes is therefore worthwhile. For this purpose, Rose’s graphic repre-
sentation serves well to illustrate the relation between levels of individual risk across the distribution of a single risk factor and the attributable proportion of events occurring at each level of that factor (Figure 18-5). Although an earlier form of this figure appeared in the 1982 report from the WHO, based on data from the Framingham Heart Study, the present version of this figure used data from more than 360,000 men screened for the Multiple Risk Factor Intervention Trial.5,3, p 23;19 The figure includes three elements: a histogram of the distribution of serum cholesterol concentration, showing the frequencies of nine strata; the continuously increasing age-adjusted rate of coronary deaths (per 1000 men in 6 years) across the cholesterol distribution; and the percentage of deaths attributable to cholesterol concentrations greater than the lowest level (less than 4 mmol/L) at each successively higher level of cholesterol. The highest level of cholesterol (more than 7.5 mmol/L) contributed only 8% of deaths,
46894_CH18_551_566.qxd
558
2/25/10
4:14 PM
Page 558
CHAPTER 18 STRATEGIES OF PREVENTION
Essential Public Health Services Policy/Legislation Assuring Personal Health Services Organizational Partnerships Education/Media Surveillance
Community Setting
Whole Communities Schools Religious Organizations Healthcare Facilities
Early Recognition of Symptomatic Disease
Hyperlipidemia Hypertension
Tobacco
Sedentary Lifestyle
Diet
Worksites
Risk Factor/Risk Behavior
Figure 18-4 A Conceptual Framework for Public Health Practice in CVD Prevention. Source: Reprinted with permission from Circulation, Vol 107, TA Pearson, TL Bazzare, SR Daniels, et al., p 647, © 2003 American Heart Association, Inc.
20
20
8% 22% 4% 19%
10
10 13% 9%
8%
0
0 4
5 6 7 Serum Cholesterol (mmol/l)
8
Figure 18-5 Prevalence Distribution (Bars) of Serum Cholesterol Concentration Related to Age-Adjusted Mortality from Coronary Heart Disease (CHD) (Broken Curve) in Men Aged 40–59 Years. Source: Reprinted with permission from G Rose, The Strategy of Preventive Medicine, p 23, © G Rose, 1992. Oxford University Press, Oxford, England. Also reprinted from The Lancet, Vol 328, MJ Martin, WS Browner, SB Hulley, LH Kuller, and D Wentworth, Serum Cholesterol, Blood Pressure, and Mortality: Implications from a Cohort of 361,662 Men, p 4, © 1986, with permission from Elsevier.
Prevalence %
CHD Deaths/1000/6 Year
17%
and the commonest level (from 5 to 5.5 mmol/L) contributed 17%, or more than twice as many, of deaths. Every stratum except the lowest two contributed as many deaths or more than did the highest-risk stratum. The next set of figures, based on Rose’s figure, illustrates the changes in risk distribution expected from each of three strategies for reducing CHD risk by lowering cholesterol. Figure 18-6a shows a smoothed distribution curve for cholesterol. Figure 18-6b shows the distribution if the high-risk strategy were fully successful in converting those at highest risk (greater than 7 mmol/L) to the levels of lowest risk (3.5–4.5 mmol/L): The highest strata would disappear and the lowest would increase in prevalence. Figure 18-6c illustrates the population strategy, in which each stratum is converted to the next lower cholesterol level: The single highest stratum would disappear and the proportion of the population at each level of increased risk would move to the next lower level of risk. (A teaching exercise developed by Rose demonstrates that, under these assumptions, the population strategy brings about a greater reduction in events because of high blood pressure than does the high-risk strategy.) It is important to recognize that both the population and high-risk strategies are intended to reduce risk in a population in which an adverse distribution of risk is already present. Strasser’s concept of pri-
46894_CH18_551_566.qxd
2/25/10
4:14 PM
Page 559
STRATEGIES OF PREVENTION 559
mordial prevention differs fundamentally, as seen in Figure 18-6d. Rather than reducing risk by shifting the distribution downward, to the left, primordial prevention would keep the distribution far to the left from the beginning. Higher levels associated with increased risk would not occur in the first place. One further variation is the strategy of “maintaining low risk” as articulated by Stamler and oth-
ers since the late 1990s.20 Primordial prevention, in preventing epidemic development of the risk factors, might allow for an upper tail of the distribution where increased risk occurs but is rare. Maintaining low risk, however, might preserve low risk in all individuals, so that none developed the major risk factors. This scenario is represented by Figure 18-6e, in which the upper tail of the distribution is truncated.
20
20
8% 22%
19% 10
10
9%
CHD Deaths/1000/6 Year
13% 8% 0
0 4
5 6 7 Serum Cholesterol (mmol/L)
A
17%
8
20
20
8% 22% 4% 19%
10
10 13% 9%
19% 10
10
D
8
8% 17%
0 5 6 7 Serum Cholesterol (mmol/L)
B
8
17% 20
20
8%
4% 19%
10
10 13%
Prevalence %
22%
CHD Deaths/1000/6 Year
4
20
22% 4% 19% 10
10 ? 13% 9%
0 4
5 6 7 Serum Cholesterol (mmol/L)
8%
0
0 4
5 6 7 Serum Cholesterol (mmol/L)
8%
0
E 9%
20
8%
8
Figure 18-6 Effects of Alternative Strategies on the Distribution of Risk in the Population.
8
Prevalence %
0
CHD Deaths/1000/6 Year
5 6 7 Serum Cholesterol (mmol/L)
13% 9%
C
0 4
22% 4%
8%
0
20
8%
Prevalence %
CHD Deaths/1000/6 Year
17% 20
Prevalence %
4%
Prevalence %
CHD Deaths/1000/6 Year
17%
46894_CH18_551_566.qxd
560
2/25/10
4:14 PM
Page 560
CHAPTER 18 STRATEGIES OF PREVENTION
In the WHO report of 1982, from the Expert Committee chaired by Rose, the further observation was made that:5, p 12 Prediction can be considerably strengthened by considering simultaneously the results for several risk factors. Thus in one typical study [citing the Pooling Project Research Group Report], it was found that 43% of CHD cases occurred among the 20% of the study population with the highest estimated risk. . . . Unfortunately it is also true that in the study cited more than half of the CHD cases occurred in those who were not recognizably at special risk, and this is one limitation of the high-risk strategy. In this connection, publications by Mendis and WHO point to the combined effect of the high-risk and population strategies when a composite or “global” risk score is used to define the high-risk stratum.21,22 These observations stem from the fact that cardiovascular risk is distributed widely from the lowest risk level of each factor, a principle that is fundamental to understanding strategies of prevention.
INTERVENTION APPROACHES The generic term PREVENTION at the center of Figure 18-1 can now be replaced by a more detailed
panel. The series of intervention approaches, ranging from policy and environmental change to end-oflife care is most suitable for this purpose (Figure 18-7). This form of the figure makes explicit the link of each approach to the corresponding stages of the present reality, on the one hand, and the vision of the future, on the other. Policy and environmental change has been described previously as a type of upstream intervention for improving social and environmental conditions, and changing systems, at the level of society as a whole. But changes in policy and environment can take place anywhere along the progression to overt heart disease and stroke and further downstream as well, such as increased community availability of emergency transportation, extended coverage for rehabilitation services, or improved access to support services for end-of-life care. Behavior change at the population level, as through widespread information and education, can bring about changes in culture as evidenced in public attitudes regarding smoking. It can also reinforce interventions further downstream, by strengthening community supports for relevant system changes. Risk-factor detection and control becomes applicable at the individual level, once risk factors have developed. This remains a necessary aspect of prevention throughout the life span of those affected and is part of both acute and long-term case management. Need for emergency care and acute case management arises
A Vision of the Future Social and Environmental Conditions Favorable to Health
Policy and Environmental Change change
Behavioral Patterns Promote Health
Behavior Behavior Change change
Few Events/ Only Rare Deaths
Low Population Risk
Risk RiskFactor Factor Detection Detection and andControl Control
Emergency Emergency Care/Acute Care/Acute Case Case Management Management
Intervention Approaches
Full Functional Capacity/ Low Risk of Recurrence
Rehabilitation/ Long-term Case Management
Good Quality of Life Until Death
End-ofLife Care
The Present Reality Unfavorable Social and Environmental Conditions
Adverse Behavioral Patterns
Major Risk Factors
First Event/ Sudden Death
Disability/ Risk of Recurrence
Fatal CVD Complications/Decompensation
Figure 18-7 Intervention Approaches to Move from “The Present Reality” to “A Vision of the Future.” Source: Adapted from A Public Health Action Plan to Prevent Heart Disease and Stroke, p 6, US Department of Health and Human Services, Centers for Disease Control and Prevention, Atlanta, GA, 2003.
46894_CH18_551_566.qxd
2/25/10
4:14 PM
Page 561
A DEVELOPING COUNTRY PERSPECTIVE 561
for victims of heart attacks, strokes, and other cardiovascular emergencies—not only in the first event but also potentially, for those who survive, on multiple occasions. Limited to survivors of one or more previous events, rehabilitation and long-term case management may serve to improve functional capacity of the individual and reduce the risk of a recurrence. End-of-life planning and care may be effective in improved quality of life until, ultimately, death occurs. These multiple pathways, with their vertical links and horizontal reach, constitute the beginnings of a framework for action—the foundation of a comprehensive public health strategy. A further step is to incorporate the larger goals for heart disease and stroke prevention that direct efforts to bring about the changes, from upstream to downstream, expected from the intervention approaches. For the United States, goals for improving the nation’s health are updated every 10 years. Accordingly, Healthy People 2010 was released in January 2000 and presents the goals for the current decade.23 First, two overarching goals for the nation’s health are recognized—to increase quality and years of healthy life and to eliminate disparities: “The first goal of Healthy People 2010 is to help individuals of all ages increase life expectancy and improve their quality of life.”23, p 8 ”The second goal . . . is to eliminate health disparities among segments of the population, including differences that occur by gender, race or ethnicity, education or income, disability, geographic location, or sexual orientation.”23, p 11 Effective action through any of the six types of intervention approaches identified here would be expected to result in some progress toward each of these goals. For the goals specific to heart disease and stroke prevention, however, the picture is quite different. On the basis of the goal statement for Focus Area 12, Heart Disease and Stroke, the following can be distinguished (Figure 18-8):23, p 12-3 prevention of risk factors; identification and treatment risk factors; early identification and treatment of heart attacks and strokes; and prevention of recurrent events. These and the overarching goals can be added to complete the framework as in Figure 18-9. Goal 1, prevention of risk factors, requires intervention before risk factors develop—through policy and environmental change and population-wide behavior change. Goal 2, detection and treatment of risk factors, requires intervention once risk factors are present— through risk-factor detection and control. (Detection and management of subclinical cardiovascular disease, important for recognizing disease progression and risk in advance of an acute event, is most appropriately included here but was not identified sepa-
The national goal for heart disease and stroke prevention: - prevention of risk factors - detection and treatment of risk factors - early identification and treatment of heart attacks and strokes - prevention of recurrent cardiovascular events
Figure 18-8 The National Goal for Heart Disease and Stroke Prevention. Source: Adapted from Healthy People 2010. 2nd ed. With Understanding and Improving Health and Objectives for Improving Health. 2 vols. Washington, DC: US Government Printing Office, November 2000.
rately from “risk factors” in Healthy People 2010.) Goal 3, early identification and treatment of heart attacks and strokes, can have effect only when warning signs and symptoms become apparent, through emergency care and acute case management. Goal 4, prevention of recurrent cardiovascular events, applies only to survivors of previous events, through rehabilitation and (for those with recognized subclinical disease) long-term case management, and ultimately through end-of-life care. The resulting framework is being employed currently in the United States to guide public health strategies for heart disease and stroke prevention at local, state, and national levels, as will be discussed in Chapter 22, “Taking Action.”
A DEVELOPING COUNTRY PERSPECTIVE In two successive editions of Disease Control Priorities in Developing Countries, the World Bank identifies “those interventions that, on the basis of their cost-effectiveness, governmental policy should seek to encourage or discourage.”24, p 4; 25 Both editions address the broad range of health conditions of major public health significance in the developing world. In the second edition, interventions are categorized as “population-based” (previously “public health”) or “personal” (previously “clinical”) and are distinguished from “policy instruments” that “encourage, discourage, or undertake interventions.” These are described as follows:26, p 5 • Population-based primary prevention is directed toward entire populations or population subgroups. These interventions fall into
46894_CH18_551_566.qxd
562
2/25/10
4:14 PM
Page 562
CHAPTER 18 STRATEGIES OF PREVENTION
Action Framework for a Comprehensive Public Health Strategy To Prevent Heart Disease and Stroke Social and Environmental Conditions Favorable to Health
Behavioral Patterns Promote Health
Policy and Environmental Change
A Vision of the Future Few Events/ Only Rare Deaths
Low Population Risk
Full Functional Capacity/ Low Risk of Recurrence
Good Quality of Life Until Death
Behavior Change Risk Factor Detection and Control
Emergency Care/Acute Case Management
Intervention Approaches
Rehabilitation/ Long-term Case Management
End-ofLife Care
The Present Reality Unfavorable Social and Environmental Conditions
Adverse Behavioral Patterns
Major Risk Factors
First Event/ Sudden Death
Fatal CVD Complications/Decompensation
Disability/ Risk of Recurrence
The Healthy People 2010 Goals Increase Quality and Years of Healthy Life Eliminate Disparities Goal 1
Goal 2
Goal 3
Goal 4
Prevention of risk factors
Detection and treatment of risk factors
Early indentification and treatment of heart attacks and strokes
Prevention of recurrent cardiovascular events
Figure 18-9 Strategic Framework for a Comprehensive Public Health Strategy to Prevent Heart Disease and Stroke. Source: Adapted from A Public Health Action Plan to Prevent Heart Disease and Stroke, p 6, US Department of Health and Human Services, Centers for Disease Control and Prevention, Atlanta, GA, 2003.
three broad categories: personal behavior change, control of environmental hazards, and population-oriented medical interventions (for example, immunization, mass chemoprophylaxis, and screening and referral). • Personal interventions are directed toward individuals and can be provided at home; at clinics (community, private, work-based, or school-based); at district hospitals; or at referral hospitals. Several types of interventions are distinguished according to their aims: primary prevention, cure, acute management, secondary prevention (or chronic care), rehabilitation, and palliation. Policy instruments were described as “activities that governments or other entities that wish to encourage or discourage interventions or to expand the potential interventions could undertake. The following are major instruments of
policy: information, education, and communication . . . taxes and subsidies on commodities, services, and pollutants . . . regulation and legislation . . . direct expenditures . . . research and development.”26, p 59 Some elements of this perspective (immunization, mass chemoprophylaxis, cure) have at most limited relevance to cardiovascular diseases or other chronic noncommunicable conditions. In all other respects, the concepts of intervention and policy instruments addressed here are largely coherent with the discussion of strategies, stages, approaches, and settings, described previously, although they are combined in differing ways. The role of government in intervention through application of policy instruments is fundamental to the World Bank perspective. A similar view is reflected widely in other sources, an important example being the report cited earlier, A Race Against Time––The Challenge of Cardiovascular Disease in
46894_CH18_551_566.qxd
2/25/10
4:14 PM
Page 563
CURRENT ISSUES 563
Developing Economies.27 Attention is called to macroeconomic implications of cardiovascular diseases in developing countries, where three levels of prevention are recognized: “macroeconomic and whole-of-government interventions that affect everyone”; “population-based interventions”; and “providerbased intervention.” Macroeconomic interventions are characterized as deriving from government ministries other than health and as implicitly political in nature:27, pp 59–65 • tobacco production and consumption— including subsidies, taxes, advertising and control strategies, and incentives to grow crops other than tobacco; • nutrition—including food production, processing and marketing subsidies, such as those in relation to animal or vegetable fats, and the salt content of foods; • education—including decisions about curricula in schools (e.g., physical or drug education, nutrition, and cooking) and assistance in managing stress;
• urban planning—including recreational spaces, transport systems, and city design that encourages physical interaction with the environment. Population-based interventions are considered the province of other agencies as well as government agencies and have explicit health goals. Policies and programs related to tobacco, nutrition, and specific cardiovascular risk factors are cited as examples that use media projects, advocacy, and social programs for their implementation. Provider-based prevention seeks to modify risk factors, as they become more prevalent over time, through more individualized and medical interventions.
CURRENT ISSUES Strategies for primary prevention of cardiovascular diseases continue to be discussed and debated, in terms of high-risk versus population-wide intervention, single versus multiple risk-factor targets, and single drug versus multidrug management. Table 18-2 is based
Table 18-2
A Comparison of Approaches to the Primary Prevention of Major CVD Prevention Approach RRR Predicted Reduction in Major CVD (%) “High-Risk” Approach Management Group Identified for Treatment Top 10% Top 20% Top 30% Treatment decision based on total Statin 30% 6% 9% 12% cholesterol Treatment decision on blood -blocker/diuretic 22% 6% 8% 10% pressure Treatment decision based on total 13% 21% 28% cholesterol Treatment decision based on blood Aspirin, statin, ACE inhibitor 68% 18% 25% 31% pressure and -blocker/diuretic Treatment decision based on overall 17% 28% 37% absolute risk Framingham 10-year CHD event risk ⱖ 30% ⱖ 20% ⱖ 15% Treatment decision based on overall absolute risk
Statin -blocker/diuretic Aspirin, statin, ACE inhibitor and -blocker/diuretic
30% 22% 68%
Population “shifting mean” approach Reduce mean total cholesterol in the population Reduce systolic blood pressure Reduce both mean total cholesterol and mean systolic blood pressure
5% 15% 21% 4% 11% 16% 11% 34% 49% “Shift” the risk factor distribution bya
—
5% 12%
10% 22%
15% 32%
— —
16% 26%
29% 45%
40% 59%
RRR: relative risk reduction (high-risk approach). a Everyone reduces their total cholesterol ⫹ blood pressure level by the same absolute amount. Source: Reprinted with permission from European Heart Journal, Vol 25, J Emberson, P Whincup, R Morris, et al., © The European Society of Cardiology, 2004, p 487.
46894_CH18_551_566.qxd
564
2/25/10
4:14 PM
Page 564
CHAPTER 18 STRATEGIES OF PREVENTION
on the follow-up experience of a cohort of nearly 6000 40- to 59-year-old men enrolled in the British Regional Heart Study between 1978 and 1980.28 Data from clinical trials were used to estimate the expected impact of candidate interventions in reducing the observed event rate for fatal or nonfatal coronary events and strokes among the 450 cases actually observed. The high-risk approach, based on use of statins to reduce total cholesterol concentrations limited to the top 10% of the distribution at baseline, would be expected to result in only a 6% reduction in events; the proportion would only reach 12% by including three times as many men in treatment (the top 30% of the distribution). Treatment based on the overall risk estimate and using a multidrug combination would reduce the events by 11% for those with initial risk of 30% or greater, by 34% for those with initial risk of 20% or greater, and by 49% in those with 15% or greater, respectively, in 10 years. By contrast, shifting the distributions of both total cholesterol and systolic blood pressure downward by 5%, 10%, and 15% of the mean value at baseline through population-wide measures would reduce events by 26%, 45%, and 59%, respectively. It was concluded that the limited impact of the high-risk approach relative to population-wide measures, under the assumptions of the analysis, means a lower threshold for treatment would be necessary to match the potential benefit of population-wide measures by high-risk treatment. This raises the question of the distinction between high-risk individuals within a population and being an “average” member of a population whose overall risk is high. A contrary view is taken by Manuel and others, who compared the projected effect on coronary death rates of three intervention strategies, on the basis of risk-factor distributions in the Canadian population.29 The “population health strategy” would lower cholesterol uniformly in the whole population. The “single raised risk factor strategy” would treat persons with baseline values greater than 6.2 mmol/L (11.1% of the population) with statins to lower cholesterol. With the “high baseline risk strategy,” the Framingham risk score would identify those with risk of coronary heart disease death in 10 years as predicted from multiple risk factors, and all persons with baseline risk of 15% or greater (12.9% of the population) would be treated. Under assumptions of only a 2% decrease in cholesterol concentration in the whole population, and 100% adherence to the treatment regimen in the two high-risk strategies, far greater numbers of death would be prevented by the baseline risk approach than either the single risk factor or population-wide approach, with the latter hav-
ing the least effect. On this basis, the authors concluded that “population health strategies that target the majority of the population (people at low coronary heart disease risk) have little effect on population health outcomes because the population risk is low in this group.”29, p 661 Jackson and others, in a companion article, posed the question, “Does Rose’s population prevention axiom still apply in the 21st century?”30, p 617 The results of the analysis by Manuel and colleagues were discussed in the context of North American experience (in Canada and the United States) with previous downward shifts in distributions of cholesterol. It was concluded that there was little likelihood of more than a further 2% reduction in population-wide levels in the future. (There is room to doubt this key assumption, given examples of populations with more favorable such distributions and the evidence that favorable change can occur, reviewed in Chapters 11 and 12; it seems far from certain that the limit of substantial further improvement has been reached.) In large part under this assumption, however, the value of the population strategy was considered diminished relative to that of the high baseline risk strategy. In addition, the latter strategy redefines high risk by including those with existing cardiovascular disease as well as multiple risk factors. The effect is to reduce the diffusion of risk across the population while concentrating it among those to be treated. This strategy depends on individual clinical assessment of the whole population, which has very limited practical applicability in most circumstances. Additional considerations in assessing potential impacts of alternative strategies of prevention include feasibility, cost, and the full range of potential health benefits, in the context of a given population at a given time. Jackson and others answer their opening question, as follows: “Rose’s population and absolute risk axioms are both relevant to cardiovascular disease prevention in the 21st century, although their applicability is likely to vary by country and by time.”30, p 618 They suggest further that population strategies are most relevant in this century to low- and middle-income countries, where the population strategy has yet to be implemented widely. A further qualification is needed: Given the high and, in the case of obesity, diabetes, and high blood pressure, increasing prevalence of risk factors in the United States, as well as marked disparities among population subgroups in cardiovascular disease burden, there remains an essential role for the population strategy in North America as well. Consistent with this view is the assessment by Pearson on the continuing roles of combined strate-
46894_CH18_551_566.qxd
2/25/10
4:14 PM
Page 565
REFERENCES 565
gies, much as Rose proposed more than two decades ago:31, p 49 “Primary prevention requires assessment of risk in asymptomatic people, to yield cost-effective benefits. CVD prevention at the societal level should target deleterious behavior in community settings, using effective public health interventions. Policy options that involve multiple preventive approaches offer the best opportunity to minimize the economic and social burdens of CVD.” REFERENCES 1. Committee for the Study of the Future of Public Health. The Future of Public Health. Washington DC: Division of Health Care Services, Institute of Medicine. National Academy Press; 1988. 2. Nielson WA, Knott TA, Carhart PW, eds. Webster’s New International Dictionary of the English Language. 2nd ed. Unabridged. Springfield, MA: G&C Merriam Company, Publishers; 1961. 3. Rose G. The Strategy of Preventive Medicine. Oxford (England): Oxford University Press; 1992. 4. Rose G. Strategy of prevention: lessons from cardiovascular disease. BMJ. 1981;282: 1847–1851. 5. World Health Organization Expert Committee. Prevention of Coronary Heart Disease. WHO Technical Report Series 679. Geneva: World Health Organization; 1982. 6. Rose G. Sick individuals and sick populations. Int J Epidemiol. 1985;14:32–38. 7. US Department of Health, Education, and Welfare. Healthy People. The Surgeon General’s Report on Health Promotion and Disease Prevention, Public Health Service, Office of the Assistant Secretary for Health and Surgeon General. DHEW (PHS) Publication No. 7955071. Washington, DC: US Government Printing Office; 1979. 8. Mensah GA, Dietz WH, Harris VB, et al. Prevention and control of coronary heart disease and stroke—nomenclature for prevention approaches in public health. A statement for public health practice from the Centers for Disease Control and Prevention. Am J Prev Med. 2005;29(5S1):152–157.
9. Strasser T. Reflections on cardiovascular diseases. Interdisc Sci Rev. 1978;3:225–230. 10. Forrest CB, Riley AW. Childhood origins of health: a basis for life-course health policy. Health Aff. 2004;23:155–164. 11. National Heart, Lung and Blood Institute. Report of the Task Force on Research in Epidemiology and Prevention of Cardiovascular Diseases. Washington, DC: National Institutes of Health, Public Health Service, US Department of Health and Human Services; 1994. 12. WHO Expert Committee. Prevention in Childhood and Youth of Adult Cardiovascular Diseases: Time for Action. Technical Report Series 792. Geneva: World Health Organization; 1990. 13. Association of State and Territorial Directors of Health Promotion and Public Health Education. Policy and Environmental Change. New Directions for Public Health. Final Report. US Centers for Disease Control and Prevention; 2001. 14. Brownson RC, Haire-Joshu D, Luke DA. Shaping the context of health: a review of environmental and policy approaches in the prevention of chronic diseases. Annu Rev Public Health. 2006;27:341–370. 15. Committee on Assuring the Health of the Public in the 21st Century. The Future of the Public’s Health in the 21st Century. Board on Health Promotion and Disease Prevention, Institute of Medicine. Washington: The National Academies Press; 2003. 16. Dahlgren G, Whitehead M. Policies and Strategies to Promote Social Equity in Health. Stockholm: Institute for the Futures Studies; 1991. 17. Worthman CM. Epidemiology of human development. In: Panter-Brick C, Worthman CM, eds. Hormones, Health, and Behavior: A Socio-Ecological and Lifespan Perspective. Cambridge: Cambridge University Press; 1999:47–104. 18. Pearson TA, Bazzarre TL, Daniels SR, et al. American Heart Association Guide for
46894_CH18_551_566.qxd
566
2/25/10
4:14 PM
Page 566
CHAPTER 18 STRATEGIES OF PREVENTION
Improving Cardiovascular Health and the Community Level. A Statement for Public Health Practitioners, Healthcare Providers, and Health Policy Makers from the American Heart Association Expert Panel on Population and Prevention Science. Circulation. 2003; 107:645–651. 19. Martin MJ, Hulley SB, Browner WS, Kuller LH, Wentworth D. Serum cholesterol, blood pressure, and mortality: implications from a cohort of 361,662 men. Lancet 1986;2: 933–936. 20. Stamler J, Stamler S, Neaton JD, et al. Low risk-factor profile and long-term cardiovascular and noncardiovascular mortality and life expectancy. Findings for 5 large cohorts of young adult and middle-aged men and women. JAMA. 1999;282:2012–2018. 21. Mendis S. Cardiovascular risk assessment and management in developing countries. Vasc Health Risk Manag. 2005;1:15–18. 22. World Health Organization. Prevention of Cardiovascular Disease. Guidelines for Assessment and Management of Cardiovascular Risk. Geneva: World Health Organization; 2007. 23. US Department of Health and Human Services. Healthy People 2010. 2nd ed. With Understanding and Improving Health and Objectives for Improving Health. 2 vols. Washington DC: US Government Printing Office; 2000. 24. Jamison DT, Mosely WH, Measham AR, Bobadilla JL, eds. Disease Control Priorities in Developing Countries. Oxford (England): Oxford University Press; 1993.
25. Jamison DT, Breman JG, Measham AR, et al., eds. Disease Control Priorities in Developing Countries. 2nd ed. Washington, DC: International Bank for Reconstruction and Development/The World Bank; 2006. 26. Gaziano T, Jamison DT, Shahid-Salles S. Intervention categories and pertinent policy instruments. Appendix 2.A In: Jamison DT, Breman JG, Measham AR, et al., eds. Disease Control Priorities in Developing Countries. 2nd ed. Washington, DC: International Bank for Reconstruction and Development/The World Bank; 2006:59. 27. Leeder S, Raymond S, Greenberg H. A Race Against Time: The Challenge of Cardiovascular Disease in Developing Countries. New York: The Trustees of Columbia University in the City of New York; 2004. 28. Emberson J, Whincup P, Morris R, Walker M, Ebrahim S. Evaluating the impact of population and high-risk strategies for the primary prevention of cardiovascular disease. European Heart Journal. 2004;25:484–491. 29. Manuel DG, Lim J, Tanesuputro P, Anderson M, Alter DA, Laupacis A, Mustard CA. Revisiting Rose: strategies for reducing coronary heart disease. BMJ. 2006;332:659–662. 30. Jackson R, Lynch J, Harper S. Preventing coronary heart disease. BMJ. 2006;332:617–618. 31. Pearson TA. The prevention of cardiovascular disease: Have we really made progress? Health Aff. 2007;26:49–60.
46894_CH19_567_590.qxd
2/25/10
4:14 PM
Page 567
C H A P T E R
19 Evidence and Decision Making ing the Cochrane Collaboration, each of these organizations presents recommendations for policy or practice. Many current clinical interventions have become established and others rejected on the basis of systematic reviews of evidence. Fewer public health or population-wide interventions have been evaluated in this way, although the Guide to Community Preventive Services provides some important examples. Current issues in the area of evidence-based guidelines and decision making for CVD prevention include continuing controversy over the nature of evidence needed to justify action; delay in translating existing evidence into recommendations and policies; and a typical lag of several years before evidencebased guidelines and decisions are widely implemented in practice.
SUMMARY Preventing cardiovascular diseases requires that effective policies and practices are formulated and implemented by those able to take meaningful action—from informed individuals to healthcare providers and policymakers. Recommended preventive policies and practices are expected to be “evidence-based,” meaning that they are founded on good science, appropriately evaluated. This is the underpinning of “evidence-based public health” as well as “evidence-based medicine.” The evidence on which decisions about policies and practices are made is potentially broad in scope, and procedures for its systematic review or evaluation are increasingly codified. What constitutes evidence is itself sometimes controversial. The boundary may go far beyond a narrow view of the research continuum to include highly subjective information. Whether there is a “gold standard” for quality or admissibility of evidence—the randomized control trial is the usual candidate—is debated especially in the arena of public health decision making. Standardized procedures for reviewing and evaluating evidence for health-related policies and practices have been developed by numerous organizations and agencies, with some focusing specifically on prevention or treatment of cardiovascular conditions. Several of their approaches to evaluating evidence are examined here—those of the Cochrane Collaboration, American College of Cardiology/American Heart Association and European Society of Cardiology, United States Preventive Services Task Force, World Health Organization, and United States Task Force on Community Preventive Services. Differences in approach parallel differences in questions addressed and intended outcomes of evaluation. Except-
INTRODUCTION Questions of causation were discussed earlier (Chapter 17) along with the kinds of evidence that are relevant and considerations in reaching causal conclusions. Concepts of prevention, addressed in Chapter 18, suggest different questions: Did an intervention produce a different health outcome than would have been expected in the absence of the intervention? Were there beneficial effects that outweigh any harmful effects? Is use of the intervention under other circumstances likely to be equally beneficial and no more harmful? These and further questions about intervention require evidence of different kinds. Evaluating this evidence may also differ from the procedures described for causal interpretation of associations. This chapter concerns the nature of
567
46894_CH19_567_590.qxd
568
2/25/10
4:14 PM
Page 568
CHAPTER 19 EVIDENCE AND DECISION MAKING
evidence regarding interventions and its evaluation in decision making for prevention, especially in public health. Where does evidence for intervention come from? One common view is that it is the product of a research continuum—a process that begins in the laboratory and culminates in clinical application, a progression “from bench to bedside.” In the cardiovascular arena, this view was captured in a graphic representation known as “Levy’s arrow,” after Robert I. Levy, who as Director of the National Heart, Lung and Blood Institute (NHLBI) published the figure in 1982 (Figures 19-1A,B).1, p 218 Levy’s arrow was complicated somewhat by inclusion of five stages of research, from “basic research” to “demonstration programs,” and addition of “control programs,” “professional and public education programs,” and interactions between the Institute and other federal agencies, nonfederal health organizations, and the medical profession. This “continuum of research” was seen as providing the basis for health practices in prevention and treatment at local, state, national, and international levels. The Levy report actually presented two arrows. The second was greatly simplified and was titled “Biomedical research continuum” (Figure 19-1B). This latter figure has been recognized more widely, if erroneously, as “Levy’s arrow.”1, p 218
This is an adequate depiction of the relation between laboratory and clinical research but is less satisfactory in representing the evidence base for decision making in public health. Green created a rival figure by reducing Levy’s arrow to its core elements and superimposing what he calls the “CDC wedge” (Green, Lawrence W., personal communication, February 29, 2008). The overlay introduces to the research spectrum (which widens from left to right, opposite Levy’s arrow) surveillance, community and state-level effectiveness trials, and program evaluation, all of which yield population-level––in contrast to molecularlevel––evidence. Relevant organizations and agencies are also included with the CDC wedge. The overall effect is no less complex than the true Levy arrow but better represents the scope of research that underlies discovery and evaluation of public health interventions. (Regrettably, this figure is not reproducible without color graphics and a multistage buildup to the complete picture.) The thrust of the argument is that the biomedical model gives insufficient weight to the application end of the sequence, where research is needed to evaluate the true public health impact of interventions, programs, or policies when they are actually implemented at the community or population level. This emphasis was reflected in the mid-1990s in the report on the first decade of work by CDC’s Prevention Research
Other Federal Agencies
Control Programs
NHLBI Focus
Basic Research
Applied Research and Development
Clinical Investigation
Clinical Trials
Demonstration Program
Professional & Public Education Programs
Health Practices (Prevention & Treatment) • Local • State • National • International
Non-Federal Health Organizations • Public
• Private
• Voluntary
Medical Profession
Figure 19-1A Biomedical Research Continuum, Including “Levy’s Arrow.” NHLBI = National Heart, Lung and Blood Institute. Source: Reprinted with permission from RI Levy, Circulation, Vol 65, p 218, © 1982 American Heart Association.
46894_CH19_567_590.qxd
2/25/10
4:14 PM
Page 569
NATURE OF EVIDENCE 569
Idea Generation
NHLBI Focus
Basic and Clinical Research
Knowledge Acquisition
Idea Communication
Idea Utilization and Development
Applied Research and Development
Knowledge Validation
Idea Diffusion Into Practice Demonstration and Education Programs
Health Practice
Knowledge Transfer
Figure 19-1B Biomedical Research Continuum, Including “Levy’s Arrow.” NHLBI = National Heart, Lung and Blood Institute. Source: Reprinted with permission from RI Levy, Circulation, Vol 65, p 218, © 1982 American Heart Association.
Centers, where preference was expressed for the term “biomedical and community health research” over the narrower, less salutary “biomedical research.”2 The scientific basis for health practice goes beyond evidence itself, however broad the range of disciplines and settings involved, to include evaluating the evidence for decision making. It has come to be expected that any decision or recommendation regarding a public health or medical practice will be “evidence-based.” This implies not only that a relevant body of evidence should be identified but that the process for its evaluation should be objective and explicit. This chapter addresses both the nature of evidence and processes for its evaluation, as conducted by each of several authoritative bodies in the cardiovascular or broader public health areas. Brownson has suggested a picture of this entire process as a “research-policy interface” with progression from hypothesis to scientific evidence, synthesis, decision making, and policy (Brownson, Ross C., personal communication, August 27, 2007). In Brownson’s picture, the process extends well beyond the research continuum of the Levy arrow or even the effectiveness trials and program evaluation of the CDC arrow. Research synthesis is a necessary step as evidence is brought to bear on decisions about policy and practice. For this reason, it is useful to understand not only the nature of evidence that is relevant to practice but also approaches to evaluating it. Two additional considerations would expand this picture further: First, not all knowledge entering into decisions about policy and practice comes from research. Inputs beyond research evidence need to be recognized and taken into account. Second, the relation between research and practice need not be uni-
directional as the arrows suggest. Green makes the case for recognizing “practice-based evidence” as well as “evidence-based practice” by which health practice can be a source, and not only a product, of evidence for policy decisions.3 This calls attention to evaluation of practices, programs, and policies, including the principles, methods, and contributions of such research to public health decision making. It also leads to discussion of the nature of evidence relevant to health-related policy and practice and approaches to evaluating this evidence, the main topics of this chapter.
NATURE OF EVIDENCE What is evidence for prevention? First, the full scope of prevention represents a broad range of interventions and outcomes, as suggested in a recent review by Pearson:4, p 49 Evidence-based guidelines have been developed for secondary, primary, and community-based prevention. To improve compliance with secondary prevention guidelines, programs must better organize and monitor care. Primary prevention requires assessment of risk in asymptomatic people, to yield cost-effective benefits. CVD [cardiovascular disease] prevention at the societal level should target deleterious behavior in community settings, using effective public health interventions. Policy options that involve multiple preventive approaches offer the best opportunity to minimize the economic and social burdens of CVD.
46894_CH19_567_590.qxd
570
2/25/10
4:14 PM
Page 570
CHAPTER 19 EVIDENCE AND DECISION MAKING
This brief statement points to several aspects of evidence for prevention. Reference to “evidence-based guidelines” implies derivation of practice recommendations through a process of identifying, assessing, and translating relevant information, or “evidence,” into proposed lines of action. The distinction among “secondary, primary, and communitybased prevention” (discussed in Chapter 18) indicates multiple levels of action, each based on a distinct body of evidence. For example, the evidence base for guidelines on improving organization and monitoring of patient care would be distinct from that for public health interventions addressing health-related behaviors in communities. Policy options involving multiple preventive approaches “to minimize economic and societal burdens of CVD” would be based on a correspondingly wide array of evidence. Another way to understand the scope of prevention relates to Figure 18-8, in the preceding chapter. In representing the determinants of onset and progression of the atherosclerotic and hypertensive diseases, the figure also demonstrates multiple opportunities, in principle, for intervention—targeting social and environmental conditions, population-wide behavior patterns, established risk factors, acute events, recurrent events, or end-of-life care. The rationale for intervention at each of these points, in accordance with current concepts of prevention and as components of a comprehensive public health strategy, was addressed in Chapter 18, “Strategies of Prevention.”
EVIDENCE-BASED DECISION MAKING “Evidence for prevention” relates both to clinical and public health practice and to decision making for health policy. In contrast to judging evidence on causation, based largely on the relevant science, Teutsch observes that systematic reviews and syntheses of evidence for decision making in health policy must address a broader set of issues.5 This process, when conducted in a pluralistic society, must be more deliberative and publicly transparent and include other information that may or may not be considered from a narrow view of “evidence” as limited to science. Beyond studies of interventions and their outcomes, this would include information on the distribution of risk, extent of disease burden and disparities, historical trends in risk factors and disease, economics of prevention, models and forecasts of change in the course of disease, and societal values bearing on health promotion and disease prevention. To bridge research and public health practice, Evidence-Based Public Health (2003) calls for wider
adoption of evidence-based strategies in public health. It presents methods and tools for using relevant literature and other information sources to document a public health problem and to implement and evaluate actions taken in response. Fielding, in a “Foreword,” identifies multiple user groups for such evidence: public health practitioners; policy-makers, from local to international levels; stakeholders, including the general public and special interest groups; and population health researchers who evaluate policies and programs.7 Evidence-based public health (or “EBPH”) is variously defined but consistently features science as its foundation and decision making for improving health of populations as its purpose.8 Conceptual origins of EBPH are traced to evidence-based medicine (or “EBM”), proposed in the early to mid-1990s to encourage reliance on science, rather than subjective judgment or consensus, as the basis for clinical practice. Treatment decisions for individual patients, though distinct from guidelines applied to defined groups of patients, are likewise proposed to be evidence based. Eddy argues for a unified approach to both of these levels of clinical decision making.9 In differentiating between EBPH and EBM, Brownson and others indicate that EBPH is more typically based on observational and quasiexperimental studies than randomized experiments. It is usually limited by a smaller body of evidence, especially in circumstances where a long interval from intervention to outcome poses difficulties in design, conduct, and interpretation of relevant studies. Within EBPH, they recognize two types of evidence: one includes analytic data on risk factor-disease relationships, burden of disease, and preventability; the other concerns relative effectiveness of specific interventions.6 EBPH, in their view, incorporates evidence regarding causation and overall assessment of a public health problem as well as intervention effectiveness. Consistent with this view, an extensive array of evidence is needed for public health decision making, according to the Community Preventive Services Task Force and its Guide to Community Preventive Services: Systematic Reviews and Evidence-Based Recommendations.10 Table 19-1 illustrates the quantitative and qualitative factors and underlying questions considered important for public health decision making—that is, for EBPH.11 The range of evidence, or information, required includes several necessarily subjective, judgmental, aspects. Quantitative factors include the magnitude or public health importance of a particular health problem (“size of the problem”); evidence that intervention can work, at least under controlled conditions
46894_CH19_567_590.qxd
2/25/10
4:14 PM
Page 571
EVIDENCE-BASED DECISION MAKING 571
Table 19-1
Quantitative and Qualitative Factors in Decision Making Factor Specific Questions Size of the problem It is important? What is the public health burden? Problem preventability
What is the efficacy? Can it work at least in ideal circumstances? What do we know about the biological plausibility? Is it logical (theory-based)?
Intervention effectiveness
What is the effectiveness? Does it work in real-world settings? Would it work in the settings being considered (is it generalizable)? How much less effective would it be compared with ideal settings? Is there better evidence for alternative interventions?
Benefits and harms
What are all the consequences of the intervention? What are the trade-offs?
Intervention cost
Is it affordable?
Comparison of benefits and costs
What is the value? How does it compare to other alternatives?
Incremental gain
What are the additional cost and benefits (value) compared to what is already being done (if anything)?
Feasibility
Are adequate time and money available?
Acceptability
Is it consistent with community priorities, culture, values, the political situation?
Appropriateness
Is it likely to work in this specific setting? Are there ways to better understand the context for intervention in various populations?
Equitability
Does it distribute resources fairly?
Sustainability
Are resources and incentives likely to support conditions to maintain the intervention?
Source: Reprinted from American Journal of Preventive Medicine, Vol 28(5S), LM Anderson, RC Brownson, MT Fullilove, et al., p 229, © 2005, with permission from Elsevier.
(“problem preventability,” “efficacy”); evidence that it does work in varied settings (“intervention effectiveness”); and adverse as well as positive effects of intervention (“benefits and harms,” “consequences,” or “trade-offs”). Economic considerations are of mixed character as they are partly quantitative but also involve judgment: cost and affordability, costs and benefits relative to alternative decisions, and added costs and benefits above and beyond those accruing from current policies and practices (“incremental gain”). More fully qualitative factors are considerations of feasibility, acceptability, appropriateness, equitability, and sustainability. The breadth of evidence or information relevant to public health decision making is described elsewhere in other terms, for example, in the context of obesity prevention (Table 19-2).12 A proposed framework for translating evidence into action in obesity prevention notes contributions of several types of ev-
idence, first within familiar categories of observational epidemiology, monitoring and surveillance, experimental studies, and program or policy evaluation. Less closely akin to epidemiology is “extrapolated evidence,” based on modeling to estimate program effectiveness, with or without an economic component, or on “indirect (or assumed) evidence” consisting of inferences from other areas of knowledge, such as business practices. In a category termed “experience,” three additional types of evidence are cited—“parallel evidence,” “theory and programme logic,” and “informed opinion”— the latter including both expert and lay opinion. These last sources could be quantitative but would typically represent subjective judgment. From this point of view, information of many kinds may contribute to decision making in the arena of health policy, or specifically in cardiovascular disease prevention. However, even within a narrower
46894_CH19_567_590.qxd
572
2/25/10
4:14 PM
Page 572
CHAPTER 19 EVIDENCE AND DECISION MAKING
Table 19-2
Description of Types of Evidence and Information Relevant to Obesity Prevention Type of Evidence or Information Description Observational Observational epidemiology Epidemiological studies that do not involve interventions but may involve comparisons of exposed and non-exposed individuals, e.g. cross-sectional, case-control, or cohort studies Monitoring and surveillance Population-level data that are collected on a regular basis to provide time series information, e.g. mortality and morbidity rates, food supply data, car and TV ownership, birth weights, and infant anthropometry Experimental Experimental studies
Programme/policy evaluation
Extrapolated Effectiveness analyses
Economic analyses Indirect (or assumed) evidence
Experience Parallel evidence
Theory and programme logic
Informed opinion
Intervention studies where the investigator has control over the allocations and/or timings of interventions, e.g. randomized controlled trials, or non-randomized trials in individuals, settings, or whole communities Assessment of whether a programme or policy meets both its overall aims (outcome) and specific objectives (impacts) and how the inputs and implementation experiences resulted in those changes (process) Modelled estimates of the likely effectiveness of an intervention that incorporate data or estimates of the programme efficacy, programme uptake, and (for population effectiveness) population reach Modelled estimates that incorporate costs (and benefits), e.g. intervention costs, costeffectiveness, or cost-utility Information that strongly suggests that the evidence exists, e.g. a high and continued investment in food advertising is indirect evidence that there is positive (but proprietary) evidence that the food advertising increases the sales of those products and/or product categories Evidence of intervention effectiveness for another public health issue using similar strategies, e.g. the role of social marketing or policies or curriculum programmes or financial factors on changing health-related behaviours such as smoking, speeding, sun exposure, or dietary intake. It also includes evidence about the effectiveness of multiple strategies to influence behaviours in a sustainable way, e.g. health-promoting schools approach comprehensive tobacco control programmes, or co-ordinated road toll reduction campaigns. The rationale and described pathways of effect based on theory and experience, e.g. linking changes in policy to changes in behaviours and energy balance, or ascribing higher levels of certainly of effect with policy strategies like regulation and pricing compared with other strategies such as education The considered opinion of experts in particular field, e.g. scientists able to peer review and interpret the scientific literature, or practitioners, stakeholders, and policy-makers able to inform judgments on implementation issues and modeling assumptions (incorporates “expert” and “lay knowledge”)
Source: Adapted with permission from B Swinburn, T Gill, S Kumanyika, Obesity Reviews, Vol 6, p 27, © 2005 The International Association for the Study of Obesity.
perspective limited to quantitative, research-based evidence, views diverge sharply regarding admissibility of evidence based on underlying research methods. These views are fundamental to the perceived scientific rigor of evidence-based decisions, whether for clinical or public health practice and policy. The central and often polarizing issue concerning evidence for prevention is the place of randomized
controlled trials (RCTs)—experimental studies in which participants are assigned at random (that is, independently and with known prior probability) to receive one or another of the interventions being investigated.13–16 The most critical contribution of random assignment is that it provides theoretical justification for statistical analysis and interpretation of the results. Random assignment is expected to bal-
46894_CH19_567_590.qxd
2/25/10
4:14 PM
Page 573
APPROACHES TO EVALUATION OF EVIDENCE 573
ance the distribution of known and unknown characteristics of participants among treatment groups. Accordingly, it can be judged with high confidence in a properly designed and conducted RCT, with appropriate statistical analysis, that the treatment rather than extraneous factors was the cause of observed differences in outcome between groups of participants. Under some circumstances (e.g., testing active drug versus a well-matched placebo, but not in many behavioral interventions) random assignment can also serve to mask the treatment allocation so neither participants nor evaluators can be influenced by having this knowledge. Both theoretical validity and balance in characteristics among treatment groups depend on a large number of participants being assigned because in a small study, chance variation may still result in troubling differences in makeup of the groups. “Large” may mean 50 assignees or more, not ordinarily a daunting number of individuals. But when communities are the assignment units, rather than individuals, 50 is likely to seem a very large number; often as few as two communities may constitute the study population. In this case, the critical functions of random assignment cannot be fulfilled, even though drawing straws would avoid investigator bias in making the intervention assignment. For those who hold strictly to the view that only the RCT is sufficiently rigorous, community-level trials as usually conducted do not provide admissible evidence for decision making. But for those who conduct intervention research at the community (or other group-assignment) level, the RCT is generally considered either impracticable because in part of limited feasibility of studying a sufficiently large number of communities or inappropriate because of the nature of the intervention or other considerations. Although those demanding evidence from RCTs point to the limitations of nonrandomized studies, others faced with practicalities of decision making for public health policy and practice cite limitations of RCTs as well. Here the cardinal issue is the contrast between RCTs and community-based intervention studies in terms of internal and external validity. “Internal validity” concerns the reliability of attributing the effect of an intervention to the intervention itself and not to extraneous factors. As discussed previously, random assignment is a fundamental requirement in making this attribution with highest reliability. Other elements of study design, conduct, analysis, and interpretation are also necessary, beyond random assignment, for the RCT to achieve the intended rigor. For example, selecting participants to minimize loss over the course of the study
usually entails exclusion of people who might be thought unreliable. Medication may be provided at no cost; systematic follow-up may reinforce adherence to treatment; and payment may be made as an incentive to full participation. Such factors reinforce the internal validity of the RCT. However, like random assignment, these are unlikely features of a communitybased intervention study. “External validity,” by contrast, denotes the degree to which results can be generalized to other individuals or communities not represented in a study. The exclusions and other elements of the ideal RCT seriously limit evidence on the potential effects of an intervention when it is applied to the general population; external validity is typically sacrificed in the interest of internal validity. The question remains, then, whether the results of the RCT are truly applicable in the “real world” of practice. This problem is often discussed in terms of “efficacy” (effect of intervention under the special conditions of a clinical trial) and “effectiveness” (effect of intervention in the general population of interest, beyond the constraints of the clinical trial). More balanced attention to external validity has been urged in applying research to practice.17 Community studies undertaken to evaluate the impact of population-wide policies and programs have no direct correspondence at the individual level. Here, the model of the RCT faces practical obstacles that may be difficult or impossible to overcome. This and the foregoing considerations continue to fuel arguments regarding the role of the RCT in public health decision making. It remains the “gold standard” from one point of view and continues to be vigorously advocated for health promotion research and evaluation of community interventions.15 From the perspective of many working in community health, however, the RCT is considered as largely inapplicable. The nature of evidence underlying decisions about clinical and public health policy and practice can be seen as multifaceted, varying in sources, methods, and interpretation. Differences in judgment as to what constitutes admissible evidence would not be unexpected. There are also differences in the questions to be decided, adding further to variation in approaches to evaluating evidence. These approaches are illustrated next, by examples from several widely recognized authorities in clinical and public health decision making.
APPROACHES TO EVALUATION OF EVIDENCE Efforts to systematize evaluation of evidence regarding interventions for prevention and treatment of disease
46894_CH19_567_590.qxd
574
2/25/10
4:14 PM
Page 574
CHAPTER 19 EVIDENCE AND DECISION MAKING
have become highly developed—more so than for causation, discussed in the preceding chapter. Many organizations and agencies have become sponsors of evidence reviews, each with its own methods and procedures. Their particular interests determine the topical focus and approach. The common aim is to make scientifically sound recommendations that will have a favorable impact on policy or practice. The usual intent is to make fullest use of the most rigorous available information and arrive at decisions that are “evidence-based.” This characterization is now expected of any judgment claimed to be supported by sound science. However, organizations differ in their specific purposes and the reviews they conduct reflect these differences—beginning from the questions being addressed: What is the soundest scientific conclusion about known effects of a clinical, or individual-level, intervention? What is the best-supported recommendation to healthcare providers regarding treatment of individual patients or categorical groups within the population? What health services and other interventions are most appropriate at the community or population level? Differences follow from these questions, partly in a predictable way, regarding the relevant or admissible evidence; methods and procedures for identifying, reviewing, and interpreting evidence; the nature of judgments reached; the audience intended to use the results; and means to promote awareness or application of the findings. Several examples will be presented below, including by way of illustration the types of conclusions or recommendations presented by each in the area of smoking prevention or cessation. Clinical Intervention Clinical decision making is relevant to public health, as reducing burden and eliminating current disparities in health depend in part on health services at the individual level.4 The approaches of the Cochrane Collaboration, American Heart Association/American College of Cardiology (AHA/ACC), US Task Force on Clinical Preventive Services (USPSTF), and World Health Organization (WHO) guidelines processes are appropriate illustrations, as all bear on individuallevel recommendations for CVD prevention. The Cochrane Collaboration Perhaps the most highly systematic approach to evidence synthesis and evaluation in the context of healthrelated interventions is the Cochrane Collaboration, “The reliable source of evidence in health care.”18 The Web site offers a very extensive set of documents about
the Cochrane Collaboration—its history, organization, methods, resources, and products. These include the Cochrane Library, established in 1996, which compiles all previously conducted reviews and is republished electronically every 3 months to include revisions based on comments, corrections, and updates as well as newly completed reviews. The focus of the Cochrane Collaboration is described as follows:19, p 1 Evidence-based health care is the conscientious use of current best evidence in making decisions about the care of individual patients in the delivery of health services. Current best evidence is up-to-date information from relevant, valid research about the effects of different forms of health care, the potential for harm from exposure to particular agents, the accuracy of diagnostic tests, and the predictive power of prognostic factors. Evidence-based clinical practice is an approach to decisionmaking in which the clinician uses the best evidence available, in consultation with the patient, to decide upon the option which suits that patient best. Evidence-based medicine is the conscientious, explicit, and judicious use of current best evidence in making decisions about the care of individual patients. The practice of evidence-based medicine means integrating individual clinical expertise with the best available external clinical evidence from systematic research. A Cochrane review follows a prescribed format with a summary, structured abstract, background, objectives, selection criteria, search strategy, methods, description of studies, quality of studies, results, discussion, and authors’ conclusions. Significantly, the conclusions address “Implications for practice” and, separately, “Implications for research” but do not include recommendations:20, p 12.7.2 Drawing conclusions about the practical usefulness of an intervention entails making tradeoffs, either implicitly or explicitly, between the estimated benefits, harms and the estimated costs. Making such trade-offs, and thus making specific recommendations for an action, goes beyond a systematic review and requires additional information and informed judgments that are typically the domain of clinical practice guideline developers. Authors of Cochrane reviews should not make recommendations. Notably, according to the Cochrane Handbook for Systematic Reviews of Interventions, considerations of “the quality of the evidence, the balance
46894_CH19_567_590.qxd
2/25/10
4:14 PM
Page 575
APPROACHES TO EVALUATION OF EVIDENCE 575
of evidence and harms, values and preferences and resource utilization . . . involves judgments and effort that goes beyond the work of most review authors.”20, p 12.7.1 Evidence is graded by quality in one of four categories: high (“randomized trials; or double-upgraded observational studies”); moderate (“downgraded randomized trials; or upgraded observational studies”); low (“double-downgraded randomized trials; or observational studies”); and very low (“triple-downgraded randomized trials; or down-graded observational studies; or case series/case reports”). A highly structured approach to assessment of the body of evidence is elaborated in the Handbook and is to be documented in a tabular “summary of findings.” The first issue of the Cochrane Database of Systematic Reviews for 2008 was described as containing more than 3000 reviews and 1700 protocols for reviews in progress.21 These included 122 reviews in the “Heart Group” and 150 in the “Stroke Group.” The Heart Group included 20 topics under “Prevention,” of which 5 were related to drugs, 10 to lifestyle, 2 to complementary treatments, and 3 to other interventions. One example serves to illustrate the approach and outcome of such a review—under the heading of “Physician advice for smoking cessation.”22 The stated aims of the review were to assess effectiveness of physician advice to promote smoking cessation by their patients, with various levels of intensity of advice and support from aids to advice; outcomes were smoking cessation and both disease-specific and allcause mortality. Search methods were described that yielded 39 randomized trials conducted between 1972 and 2003. Minimal impact on smoking cessation was
Table 19-3
observed and only one study addressed mortality, with no effect of advice. The authors’ conclusions were expressed briefly in the abstract of the review as follows: “Simple advice has a small effect on cessation rates. Additional manoeuvres appear to have only a small effect, though more intensive interventions are marginally more effective than minimal interventions.”22, p 2 In keeping with standard procedures, no recommendation was given. However, conclusions were elaborated further as “Implications for practice” and “Implications for research.” Potential benefit of “brief simple advice” was recognized as depending on systematic identification of patients who smoke and routine offering of advice to quit, whereas “the marginal benefits of more intensive interventions . . . is small, and cannot be justified as a routine intervention in unselected smokers.” Regarding research, it was concluded that further studies of such interventions would be unlikely to yield new information. (It is perhaps a fine point whether these statements constitute recommendations for practice and research or merely opinions regarding “Implications.”) American College of Cardiology/American Heart Association/(ACC/AHA) The next approach is published in the American College of Cardiology/American Heart Association (ACC)/(AHA) Methodology Manual for ACC/AHA Guideline Writing Committees.23 Table 19-3 presents the rating levels of evidence and recommendations adopted for guidelines developed by these organizations. Regarding levels of evidence, the RCTs occupy a prominent place: The highest, “A,” rating applies only to topics in which multiple RCTs are
Level of Evidence and Classification of Recommendations
Level of evidence A: Data derived from multiple randomized clinical trials or meta-analyses B: Data derived from a single randomized trial, or non-randomized studies C: Consensus opinion of experts, case studies, or standard of care Classification of recommendations I: Conditions for which there is evidence and/or general agreement that a given procedure or treatment is useful and effective II: Conditions for which there is conflicting evidence and/or a divergence of opinion about the usefulness/efficacy of a procedure or treatment IIa: Weight of evidence/opinion is in favor of usefulness/efficacy IIb: Usefulness/efficacy is less well established by evidence/opinion III: Conditions for which there is evidence and/or general agreement that the procedure/treatment is not useful/effective, and in some cases may be harmful Source: Adapted from Methodology Manual for ACC/AHA Guideline Writing Committees, Methodologies and Policies from the ACC/AHA Task Force on Practice Guidelines. © 2006 American College of Cardiology Foundation and American Heart Association, Inc., p 35.
46894_CH19_567_590.qxd
576
2/25/10
4:14 PM
Page 576
CHAPTER 19 EVIDENCE AND DECISION MAKING
available; a “B” rating requires either a single RCT or multiple nonrandomized studies; consensus opinion of experts, in the absence of requirements for an A or B rating, qualifies as level “C.” Recommendations, on the other hand, may receive the highest, “Class I,” rating on the basis of “evidence for and/or general agreement that the procedure or treatment is useful and effective.” Conflicting evidence or divergence of opinion limits the recommendation to “Class II,” which may be further designated “IIa” if evidence is favorable or “IIb” if, somewhat enigmatically, “usefulness/efficacy is less well established by evidence or opinion.” “Class III” identifies conditions with “evidence and/or general agreement that the procedure or treatment is not useful/effective and in some cases may be harmful.” Translation of evidence into the language of recommendations is elaborated in some detail in this approach, as illustrated in Figure 19-2.23 Interpretation is presented for each of 12 types of recommendation, from Class I, Level A, to Class III, Level C, regarding the nature of both the recommendation and the underlying evidence. A dimension of levels of
evidence introduced in the table is the number of risk strata evaluated, in clinical trials or registries, from 3–5 for Level A and 1–2 for Level C. Suggested wording expresses the intent of each class designation, irrespective of the level of evidence on which a recommendation is based. Notably, the Manual indicates that “a recommendation can be a Class I, even if it is based entirely on expert opinion and no research studies have ever been conducted on the recommendation (Level C). Similarly, a Class IIa or IIb can be assigned a Level A if there are multiple randomized controlled trials coming to divergent conclusions.” Further, “Assigning a Level of Evidence B or C should not be construed as implying that the recommendation is weak. Many important clinical questions addressed in the guidelines either do not lend themselves to experimentation or have not yet been addressed by high quality investigations. Even though randomized controlled trials may not be available, the clinical question may be so relevant that it would be delinquent to not include it in the guideline.”23, p 36 Cost-effectiveness is explicitly excluded from consideration as beyond the scope of scientific evidence.
“Estimate of Certainty (Precision) of Treatment Effect”
“SIZE of TREATMENT EFFECT” Class I
Class IIa
Class IIb
Class III
Benefit >>> Risk
Benefit >> Risk Additional studies with focused objectives needed
Benefit ⱖ Risk
Risk ⱖ Benefit No additional studies needed
Procedure/Treatment SHOULD be performed/administered
IT IS REASONABLE to perform procedure/administer treatment
Procedure/Treatment MAY BE CONSIDERED
Procedure/Treatment should NOT be performed/administered SINCE IT IS NOT HELPFUL AND MAY BE HARMFUL
• Recommendation that procedure or treatment is useful/effective • Sufficient evidence from multiple randomized trials or meta-analyses
• Recommendation in favor of treatment or procedure being useful/effective • Some conflicting evidence from multiple randomized trials or meta-analyses
• Recommendation’s usefulness/efficacy less well established • Greater conflicting evidence from multiple randomized trials or meta-analyses
• Recommendation that procedure or treatment not useful/effective and may be harmful • Sufficient evidence from multiple randomized trials or meta-analyses
Level B Limited (2–3) population risk strata evaluated*
• Recommendation that procedure or treatment is useful/effective • Limited evidence from single randomized trial or non-randomized studies
• Recommendation in favor of treatment or procedure being useful/effective • Some conflicting evidence from single randomized trial or non-randomized studies
• Recommendation’s usefulness/efficacy less well established • Greater conflicting evidence from single randomized trial or non-randomized studies
• Recommendation that procedure or treatment not useful/ effective and may be harmful • Limited exidence from single randomized trial or non-randomized studies
Level C Very limited (1–2) population risk strata evaluated*
• Recommendation that procedure or treatment is useful/effective • Only expert opinion, case studies, or standard-ofcare
• Recommendation in favor of treatment or procedure being useful/effective • Only diverging expert opinion, case studies, or standard-of-care
• Recommendation’s usefulness/efficacy less well established • Only diverging expert opinion, case studies, or standard-of-care
• Recommendation that procedure or treatment not useful/effective and may be harmful • Only expert opinion, case studies, or standard-of-care
Suggested phrases for writing recommendations
should is recommended is indicated is useful/effective/beneficial
is reasonable can be useful/effective/ beneficial is probably recommended or indicated
may/might be considered may/might be reasonable usefulness/effectiveness is unknown/unclear/uncertain or not well established
is not recommended is not indicated should not is not useful/effective/beneficial may be harmful
Level A Multiple (3–5) population risk strata evaluated* General Consistency of direction and magnitude of effect
Additional studies with broad objectives needed; Additional registry data would be helpful
Figure 19-2 Applying Classification of Recommendations and Level of Evidence. Source: Reprinted with permission from Methodology Manual for ACC/AHA Guideline Writing Committees, Methodologies and Policies from the ACC/AHA Task Force on Practice Guidelines. © 2006 American College of Cardiology and American Heart Association, Inc., p 37.
46894_CH19_567_590.qxd
2/25/10
4:14 PM
Page 577
APPROACHES TO EVALUATION OF EVIDENCE 577
Again taking recommendations regarding tobacco use for illustration, the American Heart Association/American Stroke Association (AHA/ ASA) recommendations in the context of primary prevention of stroke, following the ACC/AHA process, are as follows:24, pp e884–e885 Abstention from cigarette smoking and (for current smokers) smoking cessation are recommended . . . (Class I, Level of Evidence B). Avoidance of environmental tobacco smoke for stroke prevention should also be considered (Class IIa, Level of Evidence C). The use of counseling, nicotine replacement, and oral smoking-cessation medications has been found to be effective for smokers and should be considered (Class IIa, Level of Evidence B). Another approach sometimes used by the AHA and ACC to synthesize expert opinion is the Delphi method, useful in development of clinical performance measures. Here the task is to prioritize clinical practices for monitoring or performance evaluation as a tool to reinforce guideline implementation. The European Society of Cardiology has adopted guideline development practices closely similar to those of the ACC/AHA, in which classes of recommendations and levels of evidence are virtually identical. This close resemblance reflects wide consensus on approaches to clinical guidelines specifically in the cardiovascular arena.25
United States Preventive Services Task Force (USPSTF) The mission of the USPSTF is “1) to evaluate the benefits of primary and secondary preventive services in apparently healthy persons based on age, sex, and risk factors for disease; and 2) to make recommendations about which preventive services should be incorporated into primary care practice.”26, p v The reach of the Task Force recommendations may extend well beyond the individual practitioner, however, to include “professional societies, coverage policies of many health plans and insurers, health care quality measures, and national health objectives.”26, p v In formulating recommendations, the USPSTF assesses the strength of evidence for a given action and estimates both benefits and harms that may result. Recommendations are then graded on the basis of the strength of evidence and magnitude of net benefit. Tables 19-4 and 19-5 outline the assessment and the form of resulting recommendations from the USPSTF.26 Without specifying the study designs underlying acceptable evidence, the outline in Table 19-4 points to quality of studies, representativeness of populations, consistency of findings, and assessment of health outcomes as considerations in grading evidence of effectiveness as good, fair, or poor. On the basis of an estimate of benefits and harms, and importantly the net difference that ranges from “substantial” to “zero/negative,” the evidence is coded from “A” (good, substantial) to “D” (good or fair,
Table 19-4
Strength, Grade, and Quality of Overall Evidence Appendix A The USPSTF grades the quality of the overall evidence for a service on a 3-point scale (good, fair, poor). Good: Evidence includes consistent results from well-designed, well-conducted studies in representative populations that directly assess effects on health outcomes. Fair: Evidence is sufficient to determine effects on health outcomes, but the strength of the evidence is limited by the number, quality, or consistency of the individual studies, generalizability to routine practice, or indirect nature of the evidence on health outcomes. Poor: Evidence is insufficient to assess the effects on health outcomes because of limited number or power of studies, important flaws in their design or conduct, gaps in the chain of evidence, or lack of information on important health outcomes. Strength of Overall Evidence and Estimate of Net Benefit Determine the Grade. Strength of Overall Evidence of Effectiveness Good Fair Poor
Estimate of Net Benefit (Benefit Minus Harms) Substantial Moderate Small Zero/Negative A B C D B B C D I—Insufficient Evidence
Source: Reprinted with permission from the US Preventive Services Task Force, The Guide to Clinical Preventive Services 2006, US Department of Health and Human Services, Agency for Healthcare Research and Quality, p 187.
46894_CH19_567_590.qxd
578
2/25/10
4:14 PM
Page 578
CHAPTER 19 EVIDENCE AND DECISION MAKING
Table 19-5
How the US Preventive Services Task Force Grades Its Recommendations Appendix A
The US Preventive Service Task Force (USPSTF) grades its recommendations based on the strength of evidence and magnitude of net benefit (benefits minus harms). A. The USPSTF strongly recommends that clinicians provide [the service] to eligible patients. The USPSTF found good evidence that [the service] improves important health outcomes and concludes that benefits substantially outweigh harms. B. The USPSTF recommends that clinicians provide [the service] to eligible patients. The USPSTF found at least fair evidence that [the service] improves important health outcomes and concludes that benefits outweigh harms. C. The USPSTF makes no recommendation for or against routine provision of [the service]. The USPSTF found at least fair evidence that [the service] can improve health outcomes but concludes that the balance of benefits and harms is too close to justify a general recommendation. D. The USPSTF recommends against routinely providing [the service] to asymptomatic patients. The USPSTF found at least fair evidence that [the service] is ineffective or that harms outweigh benefits. I. The USPSTF concludes that the evidence is insufficient to recommend for or against routinely providing [the service]. Evidence that [the service] is effective is lacking, of poor quality, or conflicting, and the balance of benefits and harms cannot be determined. Source: Reprinted with permission from the US Preventive Services Task Force, The Guide to Clinical Preventive Services 2006, US Department of Health and Human Services, Agency for Healthcare Research and Quality, p 186.
zero/negative), or “I” (poor, no assessment of net benefit). These categories then translate into the language of recommendations, shown in Table 19-5. In an important caveat, the USPSTF emphasizes individuality of patients and of clinical decisions, which should not be based on evidence alone but tailored to the patient and the situation. This emphasis changes the sense of the recommendations from the language of Table 19-5, points A and B (“strongly recommends” or “recommends” “that clinicians provide [the service] to eligible patients [emphasis added]”). Instead, the clinician is advised to “discuss services with ‘A’ and ‘B’ recommendations with eligible patients and offer them as a priority [emphasis added]” (Table 19-6).26, pp vi–vii One effect of this change is to cloud the interpretation of the recommendations, which are based on the effect of the action, not the effect of clinician advice to consider the action. Left out of the assessment is the role of patient understanding or the physician-patient interaction in determining whether the action will be taken or the intended effect of the action will be sustained to confer the expected benefit. Recommendations of USPSTF regarding counseling on tobacco use (released November 2003) illustrate some outcomes of this process:26, p 120 The USPSTF strongly recommends that clinicians screen all adults for tobacco use and provide tobacco cessation interventions for those who use tobacco products. Rating: A Recommendation. The USPSTF strongly recommends that clinicians screen all pregnant women for tobacco
use and provide augmented pregnancy-tailored counseling to those who smoke. Rating: A Recommendation. The USPSTF concludes that the evidence is insufficient to recommend for or against routine screening for tobacco use or interventions to prevent and treat tobacco use and dependence among children or adolescents. Rating: I Recommendation.
Table 19-6
Actions to Be Taken According to the US Preventive Services Task Force Recommendations Preface
• Discuss services with “A” and “B” recommendations with eligible patients and offer them as a priority. • Discourage the use of services with “D” recommendations unless there are unusual additional considerations. • Give lower priority to services with “C” recommendations; they need not be provided unless there are individual considerations in favor of providing the service. • For service with “I” recommendations, carefully read the Clinical Considerations section for guidance, and help patients understand the uncertainty surrounding these services. Source: Reprinted with permission from the US Preventive Services Task Force, The Guide to Clinical Preventive Services 2006, US Department of Health and Human Services, Agency for Healthcare Research and Quality, p vii.
46894_CH19_567_590.qxd
2/25/10
4:14 PM
Page 579
APPROACHES TO EVALUATION OF EVIDENCE 579
World Health Organization (WHO) The WHO approach to assessing levels of evidence and grading recommendations differs in several ways from the others presented here.27 As shown in Table 19-7a, evidence appears to be categorized by level separately according on the one hand to features of the underlying study designs (column 1, “Clinical Trial Data”) or on the other to subject matter (column 2, “Behavioral Risk Factor Data”). On examination, however, “clinical trial data” range from level “1⫹⫹” based on “high-quality meta-analyses, systematic reviews of randomized controlled trials (RCTs), or RCTs with a very low risk of bias” to level 4 (“expert opinion”), with observational studies alone mentioned in levels 2 and 3 (“non-analytic studies”). Thus evidence other than clinical trials would be considered or expert opinion with no reference to study data. Behavioral risk-factor data could be categorized only in level 1 (1⫹⫹, 1⫹, 1⫺) or level 2⫹⫹, the latter defined as “Case-control or cohort studies with a high risk of confounding, bias or chance and a significant risk that the relationship is not causal.”27, p 20 Level 2⫹⫹ is far weaker for this body of evidence than
Table 19-7a
Levels of Evidence, WHO Clinical Trial Data
when it applies to case-control or cohort study findings in the first column. Recommendations for “a pattern of care” (Table 19-7b) are graded as A–D with successively lower levels of supporting evidence, including a designation (√) for “Recommended best practice based on the clinical experience of the guidelines development group.”27, p 21 Within this scheme, a pattern of care might be recommended by consensus on the basis of case report or case series observations alone, or expert opinion in the absence of data qualifying for higherlevel classification. One example of WHO recommendations based on this scheme addresses smoking cessation, which is strongly encouraged for all smokers, with health professional support to do so (grade 1⫹⫹, A). Community Intervention The Guide to Community Preventive Services (GCPS) At the community level, where public health decision making is primarily focused, the methods and procedures of the Task Force on Community Preventive
Behavioral Risk Factor Data
1⫹⫹
High-quality meta-analyses, systematic reviews of randomized controlled trials (RCTs), or RCTs with a very low risk of bias
Systematic reviews of high-quality case-control or cohort studies with a very low risk of confounding, bias, or chance, and a high probability that the relationship is causal
1⫹
Well-conducted meta-analyses, systematic reviews of RCTs, or RCTs with a low risk of bias
Well-conducted case-control and cohort studies with a very low risk of confounding, bias, or chance, and a high probability that the relationship is causal
1⫺
Meta-analyses, systematic reviews of RCTs, or RCTs with a high risk of bias
Case-control and cohort studies with a low risk of confounding, bias, or chance, and a moderate probability that the relationship is causal
2⫹⫹
High-quality systematic reviews of case-control or cohort studies. High-quality case control or cohort studies with a very low risk of confounding or bias and a high probability that the relationship is causal
Case-control or cohort studies with a high risk of confounding, bias, or chance, and a significant risk that the relationship is not causal
2⫹
Well-conducted case control or cohort studies with a low risk of confounding or bias and a moderate probability that the relationship is causal
2⫺
Case control or cohort studies with a high risk of confounding or bias and a significant risk that the relationship is not causal
3 4
Non-analytical studies, e.g., case reports, case series Expert opinion
Source: Adapted with permission from World Health Organization, Prevention of Cardiovascular Disease: Guidelines for Assessment and Management of Total Cardiovascular Risk, © World Health Organization 2007, pp 20–21.
46894_CH19_567_590.qxd
580
2/25/10
4:14 PM
Page 580
CHAPTER 19 EVIDENCE AND DECISION MAKING
Table 19-7b
Grades of Recommendation, WHO Note: The grade of recommendation relates to the strength of the evidence on which the recommendation is based. It does not reflect the clinical importance of the recommendation. A There is robust evidence to recommend a pattern of care. At least one meta-analysis, systematic review of RCTs or, RCT rated as 1⫹⫹ and directly applicable to the target population; or a body of evidence, consisting principally of studies rated as 1⫹, that is directly applicable to the target population, and demonstrating overall consistency of results. B There is evidence to recommend a pattern of care. A body of evidence, including studies rated as 2⫹⫹, is directly applicable to the target population and demonstrating overall consistency of results; or extrapolated evidence from studies rated as 1⫹⫹ or 1⫹. C On balance of evidence, a pattern of care is recommended with caution. A body of evidence, including studies rated as 2⫹, directly applicable to the target population and demonstrating overall consistency or results; or extrapolated evidence from studies rated as 2⫹⫹. D Evidence is inadequate, and a pattern of care is recommended by consensus. Evidence is of level 3 or 4; or extrapolated evidence from studies rated as 2⫹. √ Recommended best practice based on the clinical experience of the guideline development group. Source: Adapted with permission from World Health Organization, Prevention of Cardiovascular Disease: Guidelines for Assessment and Management of Total Cardiovascular Risk, © World Health Organization 2007, pp 20–21.
Services are of particular interest.28,29 At the time of this writing, Community Guide publications in areas most directly related to CVD prevention were those for nutrition, physical activity, tobacco, diabetes, and obesity.30 The work of the Task Force is to evaluate evidence and make recommendations for population-based and public health interventions. Effectiveness is judged on the basis of comparative studies with either concurrent or before/after evaluation of response to intervention versus no (or alternative) intervention. (There is no reference to RCTs.) Each study is reviewed to determine the quality of execution in accordance with several elements of its design and conduct. After assessing this evidence, the next step is translation into recommendations. These processes are outlined in Figure 19-329 and Tables 19-8 and 19-9.28 Figure 19-3 shows the algorithm for identifying types of studies contributing to
the review of a particular intervention. A wide diversity of study designs is identified and suggests that assessment of evidence would be inclusive of relevant data from many sources. Table 19-8 describes assessment of the strength of evidence regarding effectiveness of population-based interventions for Community Guide development. Strength of evidence is rated as “strong,” “sufficient,” “expert opinion,” or “insufficient” as summary conclusions from the indicators shown across the table—design, execution, number of studies, consistency in direction and size of effect, effect size, and contribution of expert opinion, if any. Reference to consistency and effect size is reminiscent of considerations in causal assessment, discussed in Chapter 17, “What Causes Cardiovascular Diseases?” A greater number of studies of lesser quality may be considered equivalent to fewer studies of high quality. Sufficiency and necessary effect size are not defined a priori but are to be judged in each case by Task Force opinion. On the basis of the identified and accepted evidence and its assessment, one of five categories of recommendation applies, as shown in Table 19-9. Strength of evidence translates directly to strength of recommendation, from “strongly recommended” to “discouraged.” Notably, the process does not end here with an unqualified, universal recommendation. Further discussion of GCPS procedures goes beyond this step to consider applicability, or generalizability, of a recommendation to a particular population; evidence of side effects or unintended consequences of intervention; data from economic evaluations, when available; and barriers to implementation. Also, as a by-product of this process, evidence gaps and research needs are identified in order that recommendations may be improved in the future. Examples of this approach are the Task Force recommendations regarding tobacco, summarized and compared with others after presentation of details of the Task Force’s own review.30,31 One of several types of intervention addressed was increasing unit price for tobacco products, for the purpose of reducing tobacco-use initiation by children and adolescents. The conclusion of the GCPS was: “Strongly Recommended . . . Increasing the price for tobacco products reduces the number of adolescents and young adults who use tobacco products and the quantity consumed.”30, p 81 The companion report describes details of the review process used by the Task Force on Community Preventive Services to reach its conclusions on each of several tobacco-related interventions.31 Regarding increasing the unit price for tobacco products, the interventions are first defined, being legislative actions
46894_CH19_567_590.qxd
2/25/10
4:14 PM
Page 581
APPROACHES TO EVALUATION OF EVIDENCE 581
Comparison between exposed and unexposed?
No
Non-Comparative Study e.g., Case series Focus gr oup Case study Descriptive epi. study
Yes
Cross Sectional
Yes
Exposure and outcome determined in the same population at the same time? No
Randomized trial
No Multiple measurements made before, during or after an intervention?
No
Exposure assigned at group level? (e.g., community, county, clinic)
More than one group studied?
No Yes BeforeAfter
Yes
Yes
Group randomized trial
Yes
Time Series Investigators assign exposure?
Yes
No
No
Case Control
Outcome
Exposure assigned randomly?
Non-randomized “trial” (Group or Individual)
Groups defined by?
Exposure Other designs with concurrent comparison groups (e.g., time series study with comparison group)
No
Cohort Design?
Yes
Prospective?
Yes
Prospective Cohort Study
No Retrospective Cohort Study
Figure 19-3 Study Design Algorithm. Source: Reprinted from American Journal of Preventive Medicine, Vol 18 (1S), S Zaza, LK Wright-De Agüero, PA Briss, et al., p 74, © 2000, with permission from Elsevier.
to increase the excise tax on tobacco products. The rationale and prior experience are cited. The review of evidence is summarized with respect to effectiveness, applicability, other positive or negative effects, and economic aspects. Barriers to implementation are discussed. The conclusion is then expressed as in the summary cited previously.
Because experience teaches that recommendations and guidelines for clinical or public health practice tend at best to be adopted only very gradually, it is also of interest that the Community Guide addresses dissemination of evidence-based interventions and not only development of the recommendations themselves (Figure 19-4).27 The dissemination process
Source: Reprinted from American Journal of Preventive Medicine, Vol 18(1S), PA Briss, S Zaza, M Papaioanou, et al., p 40, © 2000, with permission from Elsevier.
b
The categories are not mutually exclusive; a body of evidence meeting criteria for more than one of these should be categorized in the highest possible category. Studies with limited execution are not used to assess effectiveness. c Generally consistent in direction and size. d Sufficient and large effect sizes are defined on a case-by-case basis and are based on Task Force Opinion. e Expert opinion will not be routinely used in the Guide but can affect the classification of a body of evidence as shown. f Reasons for determination that evidence is insufficient will be described as follows: A. Insufficient designs or executions, B. Too few studies, C. Inconsistent, D. Effect size too small, E. Expert opinion not used. These categories are not mutually exclusive and one or more of these will occur when a body of evidence fails to meet the criteria for strong or sufficient evidence.
a
Insufficientf
Expert Opinion
Sufficient
4:14 PM
Evidence of Effectivenessa Strong
Assessing the Strength of a Body of Evidence on Effectiveness of Population-Based Interventions in the Guide to Community Preventive Services Design Suitability— Execution— Greatest, Moderate, Good or Fairb or Least Number of Studies Consistentc Effect Sized Expert Opinione Good Greatest At Least 2 Yes Sufficient Not Used Good Greatest or Moderate At Least 5 Yes Sufficient Not Used Good or Fair Greatest At Least 5 Yes Sufficient Not Used Meet Design, Execution, Number, and Consistency Criteria for Sufficient Large Not Used But Not Strong Evidence Good Greatest 1 Not Applicable Sufficient Not Used Good Greatest or Moderate At Least 3 Yes Sufficient Not Used Good or Fair Greatest, Moderate, At Least 5 Yes Sufficient Not Used or Least Varies Varies Varies Varies Sufficient Supports a Recommendation A. Insufficient Designs or Execution B. Too Few Studies C. Inconsistent D. Small E. Not Used
582
2/25/10
Table 19-8
46894_CH19_567_590.qxd Page 582
CHAPTER 19 EVIDENCE AND DECISION MAKING
46894_CH19_567_590.qxd
2/25/10
4:14 PM
Page 583
APPROACHES TO EVALUATION OF EVIDENCE 583
Table 19-9
Relationship of Strength of Evidence of Effectiveness and Strength of Recommendations Strength of Evidence of Effectiveness Recommendation Strong Strongly recommended Sufficient Recommended Insufficient empirical information supplemented Recommended based on expert opinion by expert opinion Insufficient Available studies do not provide sufficient evidence to assess Sufficient or strong evidence of ineffectiveness or harm Discouraged Source: Reprinted from American Journal of Preventive Medicine, Vol 18(1S), PA Briss, S Zaza, M Papaioanou, et al., p 41, © 2000, with permission from Elsevier.
includes five stages (innovation development, awareness, adoption, implementation, and maintenance), each with multiple activities illustrated in the figure. The focus of dissemination discussed here is on the Community Guide as a whole rather than specific interventions, which might well be targeted in a similar manner. Beginning from the principle that evidencebased interventions are desired for sound policies and practice, effective communication to relevant audiences, assurance of actual adoption and implementation, and attention to sustained action in accordance with intended interventions are all required if proposed interventions are to have substantial public health impact. A continuing issue regarding the evidence base for intervention is the question of applicability of evidence from one population or setting to others. This recalls the tension between internal and external validity in design and interpretation of clinical trials discussed previously. Green and Glasgow have used an approach to program evaluation called RE-AIM (for reach, effectiveness, adoption, implementation, and maintenance) as a frame of reference for examining study results regarding this issue (Table 19-10).17 Each of these terms is defined, and questions are proposed by which to assess adequacy of any one study to predict suitability and outcomes of an intervention in another population than where it has been evaluated. The report anticipates limitations in applicability and offers guidance for integrating available evidence regarding an intervention with additional information about the target population and setting with both theory and local experience to adapt an intervention appropriately. Taking this issue a step further, the same authors propose a set of criteria by which to rate intervention studies as to their external validity. Again, several questions are posed that, if addressed in design, conduct, analysis, and reporting of studies, would potentially strengthen their external validity (Table 19-11). Strengthening applicability or generalizability of studies in these ways would enhance their con-
tribution to decision making for public health policy and practice. Economic Evaluations Economic considerations are often part of decision making for both clinical and community health policy and practice. The scope of these considerations may be as narrow as the unit cost of a daily dose of medication or as broad as the macroeconomic consequences of failure to act now to prevent the massive burden of cardiovascular diseases in low- and middleincome countries over the next two to four decades.33 The procedures for developing evidence-based recommendations have been seen to differ among various organizations regarding whether evidence on economic aspects is incorporated. Several standard methods for economic evaluation of health-related interventions are summarized in Table 19-12.34 A different question is addressed through each of these methods, and each results in answers that contribute in different ways to decision making. Basic principles and details of each method are presented in Prevention Effectiveness: A Guide to Decision Analysis and Economic Evaluation; a method used extensively in current health policy evaluations for developing countries is the focus of Cost-Effectiveness in Health and Medicine.35,36 Application of these methods as described for systematic reviews for the Guide to Community Preventive Services (illustrated for interventions to improve vaccination coverage) encounters the difficulties of finding and selecting relevant studies, accounting for methodological differences, abstracting and adjusting economic results from multiple studies, assessing consistency of adjustment methods across studies, and summarizing the results.37 Criticism of such reviews points to the limited availability and quality of relevant studies, including incomplete specification of pertinent costs and inconsistent methods for evaluation.38 For example, interest in economic
46894_CH19_567_590.qxd
584
2/25/10
4:14 PM
Page 584
CHAPTER 19 EVIDENCE AND DECISION MAKING
Stage 1. Innovation development
Activity • •
2. Awareness
• • •
3. Adoption
• • • •
4. Implementation
• • • • • •
5. Maintenance
•
• •
Examples
Building the rationale for evidence-based interventions Development of Community Guide reviews and findings
•
Identifying target audiences and their needs Identifying communication channels Supporting knowledge transfer to promote understanding of the resource and the findings
•
Adoption of the Community Guide by the target audience(s) Targeting adoption to attitudes and values of intended audience Assessing degree to which intervention is adopted as planned Identification and address of barriers to and facilitators of adoption
Strategies to promote adoption have included the following: • Sharing findings at meetings • Having findings incorporated into CDC program or research guidance • Disseminating findings to public and private sector partners through established relationships
Assessing initial use of the Community Guide in practice Improving the skills of adopters Providing training and technical assistance Integrating with established materials and curricula Applying research on how to increase implementation Promoting additional research to fill gaps in research
•
Promoting ongoing implementation and continued use of the Community Guide recommendations Ensuring adequate financial and technical resources Continuing evaluations of audience needs and perceptions, the process of developing and disseminating the findings, and use of the resource and the findings
•
•
Development, communication testing, and improvement of the methods and process 130 reviews and findings (as of July 2003) across diverse public health areas ranging from tobacco use prevention, to physical activity promotion, to promotion of healthy social environments (by, for example, intervening on education or the environment) Audience analyses
Knowledge transfer has included combinations of the following: • Involving partners in setting priorities and developing findings • Publishing findings in various media and formats [ journal publications, Web site, other documents (brochures), press releases]
•
•
Specific dissemination and implementation activities such as workshops Additional educational efforts needed (such as incorporating Community Guide materials into training programs and academic curricula) Some efforts underway to work with funders and researchers to encourage additional research that may close gaps Efforts to institutionalize the Community Guide have included development or dissemination of Community Guide reviews by CDC staff not primarily assigned to the Community Guide, incorporation of Community Guide recommendations into Federal research and program guidance, and development of research networks to help fill Community Guide-identified research gaps
Figure 19-4 Stages and Activities in the Dissemination of Evidence-Based Interventions in the Community Guide. Source: Reprinted with permission from PA Briss, RC Brownson, JE Fielding, S Zaza, Annual Review of Public Health, Vol 25, p 295, 2004, license and copyright retained by the US Government.
46894_CH19_567_590.qxd
2/25/10
4:14 PM
Page 585
APPROACHES TO EVALUATION OF EVIDENCE 585
Table 19-10
RE-AIM Definitions and Questions to Ask to Assess Applicability (www.re-aim.org) RE-AIM Dimension Definition Questions to Ask What percentage of the target population Participation rate among intended Reach (individual level) came into contact with or began program? audience and representativeness Did program reach those most in need? of these participants Were participants representative of your practice setting? Effectiveness (individual level)
Impact on key outcomes and quality of life Consistency of effects across subgroups
Did program achieve key targeted outcomes? Did it produce unintended adverse consequences? How did it affect quality of life? What did program cost as implemented and what would it cost in your setting?
Adoption (setting and/or organizational level)
Participation rate and representativeness of settings in the evaluation
Did low-resource organizations serving highrisk populations use it? Did program help the organization address its primary mission? Is program consistent with your values and priorities?
Implementation (setting and/or organizational level)
Level and consistency of delivery across program components and different staff members
How many staff members delivered the program? Did different levels of staff implement the program successfully? Were different program components delivered as intended?
Maintenance (individual and setting levels)
At individual level: Long-term effectiveness At setting level: Sustainability and adaptation of program
Did program produce lasting effects at individual level? Did organizations sustain the program over time? How did the program evolve? Did those persons and settings that showed maintenance include those most in need?
Source: Reprinted with permission from LW Green, RE Glasgow, Evaluation of the Health Professions, Vol 29, p 133, © 2006 Sage Publications.
evaluation of the strong recommendation to increase the unit price for tobacco products was met with a dearth of studies meeting criteria for inclusion in a Community Guide review. Only one such study was found, leaving basic economic questions unanswered: What are the costs of interventions to bring about price increases, and what are the outcomes of economic evaluations when cost-effectiveness analysis includes costs of illnesses averted as a result of intervention?32 Such evaluations require cautious interpretation, especially by noneconomists who may be unfamiliar with assumptions and methods of analysis used. Despite such limitations and concerns, costeffectiveness is pivotal in policy development. This is especially true in the arena of priority-setting in the face of competing opportunities for disease pre-
vention and health promotion in low- and middleincome countries.39 Returning to the case of tobacco policy, for example, the estimated cost-effectiveness of increasing tobacco taxes by 33 percent lies anywhere from US$13 to US$195 per disability-adjusted life year (DALY) saved, versus US$55 to US$751 per DALY saved by nicotine replacement therapy. Smoking prevention appears substantially more costeffective than cessation, at least by these interventions and under the assumptions of this analysis.40 Estimating economic costs of disease is also challenging, but figures regarding state or national expenditures for treatment and national, regional, or global costs of disease in terms of lost economic productivity of victims are increasingly common. Further discussion of this dimension of the case for prevention follows in Chapter 24.
46894_CH19_567_590.qxd
586
2/25/10
4:14 PM
Page 586
CHAPTER 19 EVIDENCE AND DECISION MAKING
Table 19-11
Proposed Quality Rating Criteria for External Validity I. Reach and representativeness A. Participation: Are there analyses of the participation rate among potential (a) settings, (b) delivery staff, and (c) patients (consumers)? B. Target audience: Is the intended target audience stated for adoption (at the intended settings such as worksites, medical offices, etc.) and application (at the individual level)? C. Representativeness—Settings: Are comparisons made of the similarity of settings in study to the intended target audience of program settings—or to those settings that decline to participate? D. Representativeness—Individuals: Are analyses conducted of the similarity and differences between patients, consumers, or other subjects who participate versus either those who decline, or the intended target audience? II. Program or policy implementation and adaptation A. Consistent implementation: Are data presented on level and quality of implementation of different program components? B. Staff expertise: Are data presented on the level of training or experience required to deliver the program or quality of implementation by different types of staff? C. Program adaptation: Is information reported on the extent to which different settings modified or adapted the program to fit their setting? D. Mechanisms: Are data reported on the process(es) or mediating variables through which the program or policy achieved its effects? III. Outcomes for decision making A. Significance: Are outcomes reported in a way that can be compared to either clinical guidelines or public health goals? B. Adverse consequences: Do the outcomes reported include quality of life or potential negative outcomes? C. Moderators: Are there any analyses of moderator effects—including of different subgroups of participants and types of intervention staff—to assess robustness versus specificity of effects? D. Sensitivity: Are there any sensitivity analyses to assess dose-response effects, threshold level, or point of diminishing returns on the resources expended? E. Costs: Are data on the costs presented? If so, are standard economic or accounting methods used to fully account for costs? IV. Maintenance and institutionalization A. Long-term effects: Are data reported on longer term effects, at least 12 months following treatment? B. Institutionalization: Are data reported on the sustainability (or reinvention or evolution) of program implementation at least 12 months after the formal evaluation? C. Attrition: Are data on attrition by condition reported, and are analyses conducted of the representativeness of those who drop out? Source: Reprinted with permission from LW Green, RE Glasgow, Evaluation of the Health Professions, Vol 29, p 133, © 2006 Sage Publications.
Table 19-12 Overview of Economic Evaluation Methods Economic Evaluation Method Comparison
Measurement of Health Effects
Economic Summary Measure
Cost analysis
Used to compare net costs of different programs for planning and assessment
Dollars
Net cost Cost of illness
Cost-effectiveness analysis
Used to compare interventions that produce a common health effect
Health effects, measured in natural units
Cost-effectiveness ratio Cost per case averted Cost per life-year saved
Cost-utility analysis
Used to compare interventions that have morbidity and mortality outcomes
Health effects, measured as years of life, adjusted for quality of life
Cost per quality-adjusted life year (QALY)
Cost-benefit analysis
Used to compare different programs with different units of outcomes (health and nonhealth)
Dollars
Net benefit or cost Benefit-to-cost ratio
Source: Reprinted with permission from GA Stone, AB Hutchinson, PS Corso, et al., p 451, Understanding and using the economic evidence. In S Zaza, PA Briss, KW Harris, eds. The Guide to Community Preventive Services: What Works to Promote Health? New York: Oxford University Press; 2005:449–463.
46894_CH19_567_590.qxd
2/25/10
4:14 PM
Page 587
REFERENCES 587
The next chapter presents examples of practice recommendations and public health policies for CVD prevention that are currently advanced by national and international authorities. They are largely based on application of the concepts and approaches described here. They also reflect evolving understanding of CVD risks and their distribution within populations, in specific groups, and across countries or regions.
CURRENT ISSUES In the area of decision making for clinical and public health policy and practice, there remain three broad areas of concern. There is continuing controversy over the nature of evidence needed to justify action. Is useful action being impeded by exaggerated or misplaced concern, especially the demand for RCT-type evidence when it may be inappropriate or infeasible? Existing evidence, whatever its strengths or limitations, is not being translated into recommendations and policies in a timely way. Is progress in translation delayed by lack or insufficiency of evidence, demands of current procedures for systematic review, influences beyond the science in decision making, or other factors, and can any or all of these be remedied? When recommendations are made, there is typically a lag of several years before they are widely implemented in practice. Do the decisions reached on the basis of current evidence and approaches lack credibility or persuasiveness, or are there other explanations for inaction despite development of “evidence-based” recommendations? REFERENCES 1. Levy RI. The National Heart, Lung, and Blood Institute. Overview 1980. The Director’s Report to the NHLBI Advisory Council. Circulation. 1982;65:217–225. 2. Stoto MA, Green LW, Bailey LA, eds. Linking Research and Public Health Practice. A Review of CDC’s Program of Centers for Research and Demonstration of Health Promotion and Disease Prevention. Washington, DC: National Academy Press; 1997. 3. Green LW. Public health asks of systems science: To advance our evidence-based practice,
can you help us get more practice-based evidence? Am J Prev Med. 2006;96;406–409. 4. Pearson TA. The prevention of cardiovascular disease: Have we really made progress? Health Aff. 2007;26:49–60. 5. Teutsch SM, Berger ML. Evidence synthesis and evidence-based decision making: related but distinct processes. Med Decision Making. 2005;25:487–489. 6. Brownson RC, Baker EH, Leet TL, Gillespie KN. Evidence-Based Public Health. Oxford: Oxford University Press; 2003. 7. Fielding JE. Foreword. In: Brownson RC, Baker EH, Leet TL, Gillespie KN. EvidenceBased Public Health. Oxford: Oxford University Press; 2003:v–vii. 8. Kohatsu ND, Robinson JG, Torner JC. Evidence-based public health. An evolving concept. Am J Prev Med. 2004;27:417–421. 9. Eddy DM. Evidence-based medicine: a unified approach. Health Aff. 2005;24:9–17. 10. McGinnis JM. With both eyes open. The Guide to Community Preventive Services. Am J Prev Med. 2005;28:223–225. 11. Anderson LM, Brownson RC, Fullilove MT, Teutsch SM, Novick LF, Fielding J, Land GH. Evidence-based public health policy and practice: promises and limits. Am J Prev Med. 2005;28(5S):226–230. 12. Swinburn B, Gill T, Kumanyika S. Obesity prevention: a proposed framework for translating evidence into action. Obes Rev. 2005;6:23–33. 13. Victora CG, Habicht J-P, Bryce J. Evidence-based public health: moving beyond randomized trials. Am J Public Health. 2004;94:400–405. 14. Ioannidis JPA. Contradicted and initially stronger effects in highly cited clinical research. J Am Med Assoc. 2005;294:218–228. 15. Rosen L, Manor O, Engelhard D, Zucker D. In defense of the randomized controlled trial for health promotion research. Am J Pub Health. 2006;96:1181–1186.
46894_CH19_567_590.qxd
588
2/25/10
4:14 PM
Page 588
CHAPTER 19 EVIDENCE AND DECISION MAKING
16. Sanson-Fisher RW, Bonevski B, Green LW, D’Este C. Limitations of the randomized controlled trial in evaluating population-based health interventions. Am J Prev Med. 2007;33:155–161. 17. Green LW, Glasgow RE. Evaluating the relevance, generalization, and applicability of research. Eval Health Prof. 2006;29:126–153. 18. The Cochrane Collaboration. Evidence-based medicine and healthcare. Available at: http://www.cochrane.org/reviews/docs/ ebm.htm. Accessed March 8, 2008. 19. Starr M, Chalmers I. The evolution of The Cochrane Library, 1988–2003. 2003. Update Software: Oxford. Available at: http://www .cochrane.org/reviews/docs/ebm.htm. Accessed March 8, 2008. 20. Higgins JPT, Green S, eds. Cochrane Handbook for Systematic Reviews of Interventions Version 5.0.0 [updated February 2008]. The Cochrane Collaboration, 2008. Available from: http://www.cochrane-handbook.org. Excerpt quoted available from: http://www .mrc-bsu.cam.ac.uk/Cochrane/handbook/ chapter12. Accessed March 8, 2008. 21. The Cochrane Collaboration. Cochrane reviews and the Cochrane Library; 2008. Available from: http://www.cochrane.org/ reviews. Accessed March 8, 2008. 22. Lancaster T, Stead LF. Physician advice for smoking cessation. Cochrane Database Syst Rev. 2004, Issue 4. Art. No.:CD00165. doi: 10.1002/14651858.CD00165.pub2. Available at: http://www.thecochranelibrary.com. 23. ACC/AHA Task Force on Practice Guidelines. Methodology Manual for ACC/AHA Guideline Writing Committees. Methodologies and Policies from the ACC/AHA Task Force on Practice Guidelines. American College of Cardiology Foundation and American Heart Association, Inc. April 2006. Available at: http://www.americanheart.org/print_presenter .jhtml?identifier=3039683. Accessed January 12, 2008. 24. Goldstein LB, Adams R, Alberts MJ, et al. Primary prevention of ischemic stroke. A guideline from the American Heart
Association/American Stroke Association Stroke Council; Cosponsored by the Atherosclerotic Peripheral Vascular Disease Interdisciplinary Working Group; Cardiovascular Nursing Council; Clinical Cardiology Council; Nutrition Physical Activity, and Metabolism Council; and the Quality of Care and Outcomes Research Interdisciplinary Working Group. Circulation. 2006;113: e873–e923. 25. Committee for Practice Guidelines (CPG) of the European Society of Cardiology (ESC). Recommendations for Guidelines Production. A document for Task Force Members Responsible for the Production and Updating of ESC Guidelines. 2006. Available at: http://www .escardio.org/escardio/Templates. Accessed January 13, 2008. 26. US Preventive Services Task Force. The Guide to Clinical Preventive Services 2006. Recommendations of the US Preventive Services Task Force. Washington, DC: Agency for Healthcare Research and Quality; 2006. Available at: http://www.ahrq.gov/clinic/uspstf/ uspstbac.htm. Accessed October 14, 2007. 27. World Health Organization. Prevention of Cardiovascular Disease. Guidelines for assessment and management of cardiovascular risk. Geneva: World Health Organization; 2007. 28. Briss PA, Zaza S, Pappaioanou M, et al. Developing an evidence-based Guide to Community Preventive Services––methods. Am J Prev Med. 2000;18(1S):35–43. 29. Zaza S, Wright-De Agüero LK, et al. Data collection instrument and procedure for systematic reviews in the Guide to Community Preventive Services. Am J Prev Med. 2000; 18(1S):44–74. 30. Guide to Community Preventive Services. The Community Guide. Available at: http://www .thecommunityguide.org/index.html. Accessed September 25, 2007. 31. Hopkins DP, Husten CG, Fielding JE, Rosenquist JN, Westphal LL. Evidence reviews and recommendations on interventions to reduce tobacco use and exposure to environmental tobacco smoke. A summary of selected guidelines. Am J Prev Med. 2001;20(2S):67–87.
46894_CH19_567_590.qxd
2/25/10
4:14 PM
Page 589
REFERENCES 589
32. Hopkins DP, Briss PA, Ricard CJ, et al. Reviews of evidence regarding interventions to reduce tobacco use and exposure to environmental tobacco smoke. Am J Prev Med. 2001; 20(2S):16–66.
37. Carande-Kulis VG, Maciosek MV, Briss PA, et al. Methods for systematic reviews of economic evaluations for the Guide to Community Preventive Services. Am J Prev Med. 2000; 18(1S):75–91.
33. Leeder S, Raymond S, Greenberg H, et al. A Race Against Time: The Challenge of Cardiovascular Disease in Developing Countries. The Trustees of Columbia University in the City of New York, New York; 2004.
38. Ramsey SD. Methods for reviewing economic evaluations of community preventive services: A cart without a horse? Am J Prev Med. 2000; 18(1S):15–17.
34. Stone GA, Hutchinson AB, Corso PS, Teutsch SM, Fielding JE, Carande-Kulis VG. Understanding and using the economic evidence. In: Zaza S, Briss PA, Harris KW. The Guide to Community Preventive Services: What Works to Promote Health? New York: Oxford University Press; 2005:449–463. 35. Haddix AC, Teutsch SM, Shaffer PA, Duñet DO, eds. Prevention Effectiveness. A Guide to Decision Analysis and Economic Evaluation. New York: Oxford University Press; 1996. 36. Gold MR, Siegel JA, Russell LB, Weinstein MC, eds. Cost-Effectiveness in Health and Medicine. New York: Oxford University Press; 1996.
39. Musgrove P, Fox-Rushby J. Cost-effectiveness analysis for priority setting. In: Jamison DT, Breman JG, Measham AR, et al., eds. Disease Control Priorities in Developing Countries. 2nd ed. Washington, DC: International Bank for Reconstruction and Development/The World Bank; 2006:271–285. 40. Laxminarayan R, Chow J, Shahid-Salles SA. Intervention cost-effectiveness: overview of main messages. In: Jamison DT, Breman JG, Measham AR, et al., eds. Disease Control Priorities in Developing Countries. 2nd ed. Washington, DC: International Bank for Reconstruction and Development/The World Bank; 2006:35–86.
46894_CH19_567_590.qxd
2/25/10
4:14 PM
Page 590
46894_CH20_591_618.qxd
2/25/10
4:15 PM
Page 591
C H A P T E R
20 Recommendations, Guidelines, and Policies rated management algorithms. Examples developed in the United States, in Europe, and by the World Health Organization illustrate such clinical guidelines. Limitations of evidence may leave uncertainty about applying these guidelines to groups of special concern—particular racial/ethnic groups, women, the elderly, and children and adolescents, populations different from those represented in any particular prediction model. Regarding models, opinion may vary about risk factors to be included, outcomes of importance for risk prediction, and use of risk estimation for young adults, whose 10-year absolute risks are generally low. Community guidelines and public policies address the population-wide aspect of CVD prevention at multiple levels, from detailed recommendations for action by local communities and their leaders to broadly stated policies for implementation regionally or world-wide. In between are state- and nationallevel policies—in the United States, examples are state laws on tobacco taxation and the Healthy People goals and objectives for heart disease and stroke prevention. In substance these recommendations retain essentially the same targets of intervention as a half century ago, although they are generally more formal and much more detailed than those proposed in 1959. Current issues in this area concern, first, the extent to which existing recommendations, guidelines, and policies are actually implemented and evaluated and, second, the need for further research to advance policy development and implementation. Effective action in both respects is required to assure that the intended benefits of CVD prevention are realized for the whole population, as well as individuals at high risk.
SUMMARY Recommendations, guidelines, and policies comprise a variety of authoritative statements intended to influence individual behavior, professional practice, or public action. Such statements in the area of prevention of atherosclerotic and hypertensive diseases, from their beginning in 1959 to the present, have continually reflected contemporary knowledge and concepts of prevention. Recommendations for prevention and control of single risk factors, discussed throughout Part III, predominated early in this history and remain important. More comprehensive approaches to prevention that address multiple risk factors are emphasized currently and are illustrated by selected examples here. In some instances, chiefly the guidelines for clinical preventive practice, these statements result from formal synthesis and evaluation of evidence as described in Chapter 22. In others, expert opinion is the basis rather than a structured review of evidence. The concept that CVD risk depends on multiple risk factors, and not necessarily on extreme levels of any one factor, was discussed in Chapter 18. That concept translates into a composite definition of “high risk” based on various combinations of risk factors and outcomes. Prediction models derived from prospective epidemiologic studies are used for generating estimates of individual “absolute” or “total” risk of CVD events. Risks are then categorized, for example, in specified strata—such as from 0 to 10%, 10 to 20%, or greater than 20% risk in 10 years or some other interval from the point of assessment. On the basis of estimated risk, specific interventions would follow in accordance with sometimes highly elabo-
591
46894_CH20_591_618.qxd
592
2/25/10
4:15 PM
Page 592
CHAPTER 20 RECOMMENDATIONS, GUIDELINES, AND POLICIES
INTRODUCTION The terms “recommendations,” “guidelines,” and “policies” are often used interchangeably, although there may be a sense of narrower scope for “recommendations” and broader reach for “policies.” The intent here is to be inclusive of the wide range of authoritative statements whose purpose is to influence personal behavior, professional practice—in both clinical and public health settings—and public action for CVD prevention. Typically, one or another of the major, established risk factors has been the single focus of attention in clinical guidelines. Such statements have been noted in Part III under prevention and control of the respective factors. A contrasting multifactor approach began earlier in development of community interventions and is now widely adopted in clinical preventive approaches to risk prediction and management as well. Guidelines at the individual level are especially subject to change in response to new evidence from clinical trials. As a consequence, today’s recommendations may be revised importantly tomorrow, and any review may soon be outdated. It is nevertheless useful to illustrate the general character of current approaches by selected examples. It can then be seen how the processes of evidence synthesis and evaluation discussed in the preceding chapter culminate in specific guidelines, recommendations, and policies at this time. Links to major sources discussed here should provide access to new or updated guidelines in coming years. Although recommendations for community intervention can also change as new evidence dictates, change is less frequent owing to the relative paucity of research and program evaluation that can advance knowledge and practice at the community level. Here, too, examples that address multiple factors are discussed as follows. Beginnings A Statement on Arteriosclerosis. Main Cause of “Heart Attacks” and “Strokes,” a 21-page booklet published in 1959 by the National Health Education Committee, Inc., was the first known report to address prevention of atherosclerosis and related conditions.1 Chaired by Mrs. Albert D. (Mary) Lasker, the Committee requested a report from an eight-author panel that included five past presidents of the American Heart Association and Paul Dudley White, then President of the International Society of Cardiology, at the head of the list (see Reference). The report was endorsed by 106 members of the American Society for the Study of Arteriosclerosis,
the organization that was to become the Council on Arteriosclerosis of the American Heart Association. The document was intended to provide “a simple guide which would give the average man and woman something he or she could do in cooperation with the physician to minimize the hazards of arteriosclerosis, main cause of ‘heart attacks’ and ‘strokes.’”1, p 1 The factors identified as predisposing an individual to these events were overweight, elevated blood cholesterol level, elevated blood pressure, excessive cigarette smoking, and heredity. A family history of cardiovascular disease (CVD) was acknowledged as unchangeable but was an indication for special concern in the presence of any of the other factors. A further observation was that “regular, moderate, physical activity appears to lessen the hazards of arteriosclerosis.”1, p 1 Scientific support was cited for inclusion of each “contributing factor” from publications appearing in the 1940s and 1950s. It is instructive to recognize that this report antedated nearly 50 years of research since the 1960s. What has this further experience added, and in what respects have current recommendations, guidelines, and policy advanced from this “simple guide”? Interim Developments The intervening decades have provided an immense body of knowledge about atherosclerosis and hypertension, their causes, and their consequences. Results of population studies and clinical and laboratory research, together with years of public health practice, point to multiple strategies and methods of intervention. Part III focused on the epidemiology, including the distributions and determinants of these diseases, with brief attention to means of prevention and control of each factor. Chapter 18 traced development of concepts of prevention. Chapter 19 outlined current approaches to assembling and evaluating evidence regarding interventions in decision making for clinical and public health practice and policy. Here the focus is on the products of these concepts and procedures—major recommendations, guidelines, and policies for prevention of atherosclerotic and hypertensive diseases. An Abundance of Reports and Recommendations In addition to advances in knowledge and concepts of CVD prevention, the third critical ingredient has been the widening interest in prevention by official and voluntary health agencies, health professionals, and the public. As a result, countless formal recommendations on various aspects of cardiovascular disease prevention have been published.
46894_CH20_591_618.qxd
2/25/10
4:15 PM
Page 593
INTRODUCTION 593
One difficulty in attempting a comprehensive collection of these reports over past decades is the breadth of the subject area. The scope of recommendations now includes not only the risk factors themselves and their relation to primary prevention, as in the 1959 report, but also the underlying behaviors, later stages of prevention, population-wide and highrisk strategies, the circumstances of special population groups, information for health professionals and the general public, and materials for use in specific settings—schools, worksites, healthcare facilities, faith organizations, and communities. The wide range of health-related behaviors at issue also broadens the reach of many recommendations. Recommendations on diet, for example, have implications not only for CVD, diabetes, cancer, and other chronic conditions but also for health and well-being more generally. Overview of Current Recommendations, Guidelines, and Policies In 2004, the World Heart and Stroke Forum of the World Heart Federation presented principles for national and regional guidelines on CVD prevention.2 The report notes the importance of a societal approach to CVD prevention, in every country: Prevention of the causes of the risk factors is needed worldwide, and cardiovascular disease specialists are called on to join with epidemiologists and public health officials to address this need. Public health approaches are advocated, including governmental and industrial policy changes, education, and testing of individuals in medical or community settings. But the primary focus of the report is on clinical management of individuals at high risk of cardiovascular events and those with already-established CVD. A fundamental issue is defining high risk. In this as several current recommendations in CVD prevention, “major risk factors” continue to be recognized: high LDL-cholesterol concentration, blood pressure, and blood glucose; low HDL-cholesterol concentration; and tobacco smoking. Age is also included as a surrogate for the degree of underlying atherosclerosis. Two other categories of risk factors are recognized in the Forum report. First, “underlying risk factors” are “overweight/obesity, physical inactivity, atherogenic diet, socioeconomic and psychological stress, family history of premature CVD, and various genetic and racial factors.” Second, “emerging risk factors” are a mix of blood lipid components and nonlipid factors—insulin resistance, prothrombotic and proinflammatory markers, and subclinical atherosclerosis. These emerging factors were found in observational epidemiologic studies to be less strongly
associated with cardiovascular risk or less prevalent than the major risk factors.2, p 3113 Each of these factors is addressed in Part III, described previously. This distinction among categories of associated factors was necessary to define “total CVD risk,” a concept similar to “global risk” or “baseline risk,” discussed in Chapter 18. It is defined in this report as the estimated individual risk of suffering a CVD event on the basis of prediction equations derived from prospective epidemiologic studies. Total CVD risk is stratified as high, intermediate, or low. Factors incorporated in risk estimation are those identified here as major risk factors, including age. Underlying factors are not included in risk-prediction equations. However, they are regarded as important in determining population differences in absolute CVD rates and may be used to adjust or “calibrate” risk estimates between populations. Emerging factors are excluded from risk calculation on grounds of being less important than major risk factors. However, they are regarded as potentially useful when advising individual patients in clinical practice. The several principles proposed in this report are summarized in Table 20-1: Risk factors are the target of recommendations and remain fundamental to concepts of prevention. Guidelines should be evidencebased but at the same time allow for professional judgment. Applying estimates of absolute risk for any population requires data on risk-factor distributions specific to that population. Total risk should be the basis for treatment decisions—the greater the risk, the more stringent are the goal levels of risk reduction. Achieving national perspective on CVD prevention on the part of professional societies and others depends on adequacy of vital statistics and other nationallevel data regarding lifestyles, risk factors, and causes and outcomes of CVD. Engagement of such organizations in policy, education, and training is urged. Finally, prevention must become part of daily clinical practice. From the perspective of this report, recent clinical guidelines can now be reviewed as they have been published by organizations in the United States and Europe and by the World Health Organization (WHO). Typically, these refer to adults and are specific by sex and age. Separate guidelines for CVD prevention in women, the elderly, and children and adolescents are also noted. Some aspects of risk estimation warrant particular comment in relation to clinical guidelines. Community-level guidelines are the public health counterpart to those for patient care and have several distinguishing features. Finally, public policy for CVD prevention will be highlighted by examples that apply on a national, regional, or global scale.
46894_CH20_591_618.qxd
594
2/25/10
4:15 PM
Page 594
CHAPTER 20 RECOMMENDATIONS, GUIDELINES, AND POLICIES
Table 20-1
Strategic Principles for the Development of National Clinical Guidelines
1. Governments, national societies, and foundations should collaborate to develop clinical and public health guidelines for CVD prevention that target risk factors. 2. Evidence-based guidelines should incorporate professional judgment on the translation of such evidence into effective and efficient care addressing all areas of CVD risk. 3. The assessment of total CVD risk should be based on epidemiological risk factor data appropriate to the population to which it is applied. 4. Policy recommendations and guidelines should emphasize a total risk approach for CVD prevention. 5. The intensity of interventions should be a function of the total risk of CVD, with lower treatment thresholds for higher-risk patients. 6. National cardiovascular societies/foundations should promote routine prospective collection of validated national vital statistics on the causes and outcomes of CVD for use in the development of national policies. 7. National professional societies should inform policymakers of risk factor targets and drug therapies for prevention of CVD that are culturally and financially appropriate to their nation and ask the government to incorporate prevention of CVD into legislation whenever relevant. 8. National professional societies/foundations should facilitate CVD prevention through education and training programs for health professionals. 9. National professional societies should assess the achievement of lifestyle, risk factor, and therapeutic targets defined in the national guidelines. 10. Health professionals should include prevention of CVD as an integral part of their daily clinical practice. Source: Adapted with permission from SC Smith Jr, R Jackson, TA Pearson, et al., Circulation, Vol 109, © 2004 American Heart Association, Inc., pp 3120–3121.
CLINICAL GUIDELINES The United States The US National Heart, Lung and Blood Institute (NHLBI) of the National Institutes of Health lists five clinical practice guidelines and reports that are directly relevant to CVD prevention published between 2002 and 2004.3 Guidelines for clinical CVD prevention have proliferated in recent years, as shown by the current listing of American Heart Association (AHA) publications in this area, 29 of them appearing from 2000 through 2007.4 Current clinical preventive guidelines will be illustrated by an example from each of these sources. (The US Preventive Services Task Force [USPSTF] procedure for developing clinical recommendations was also reviewed in Chapter 19. Each of these recommendations generally addresses a single intervention, such as screening for one particular condition.5 Pertinent USPSTF recommendations are noted in the corresponding chapters of Part III.) Prevention of Coronary Heart Disease (CHD) Estimation of cardiovascular risk as a guide to individual-level treatment decisions can be illustrated for the United States by the example of the Third Report on the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) (often referred to
as “ATP III”).6 Although the main focus of this report is on blood cholesterol, its scope is far broader. As with the World Heart and Stroke Forum report, the need for population-wide, public health approaches for CVD prevention is discussed. The importance of lifestyle changes as the first line of prevention among persons free of recognized coronary heart disease is also emphasized. But in preference to the former practice of evaluating risk only in terms of individual risk factors in a categorical approach, ATP III adopted the concept of global risk assessment described previously. ATP III illustrates current guidelines and policies recommended through systematic review and evaluation of evidence. The process described in this report was much as outlined in Chapter 19:6, p I-2 The ATP III panel played four important roles in forging this evidence-based report. First, it systematically reviewed the literature and judged which reports provided relevant information. Second, it synthesized the existing literature into a series of evidence statements. This synthesis also required a judgment as to the category and strength of the evidence. Third, the panel developed recommendations based on the evidence statements; these recommendations represent a consensus judgment about the clinical significance of each evidence statement. Lastly, the panel created an integrated set of recommendations and guidelines based on individual recommendations.
46894_CH20_591_618.qxd
2/25/10
4:15 PM
Page 595
CLINICAL GUIDELINES 595
In addition to demonstrating this process, ATP III reflects a major expansion in scope beyond a focus on high blood cholesterol alone. It addresses for both risk assessment and management the full array of risk factors—here categorized as lipid risk factors, modifiable and nonmodifiable nonlipid risk factors, emerging risk factors, subclinical atherosclerotic disease, and metabolic syndrome. Risk is now global risk, and management is comprehensive. Recommendations of ATP III for intervention to achieve the established goals are described as evidence based, with classification of evidence by type and strength as in the American College of Cardiology/ American Heart Association approach discussed in Chapter 19. Throughout the report, conclusions are presented in the form of a recommendation with an accompanying evidence statement. Evidence types are: major randomized controlled clinical trials (RCTs) (A), smaller RCTs and meta-analyses of other clinical trials (B), and observational and metabolic studies (C). Level 1 strength is “very strong evidence.” For example:6, p II-4 Evidence statement: Multiple lines of evidence from experimental animals, laboratory investigations, epidemiology, genetic forms of hypercholesterolemia, and controlled clinical trials indicate a strong causal relationship between elevated LDL cholesterol and CHD (A1, B1, C1). Recommendation: LDL cholesterol should continue to be the primary target of cholesterollowering therapy. Further recommendations, accompanied in most instances by a similarly graded evidence statement, address a wide range of issues relevant to CHD prevention and reflect the multiple risk-factor approach. These concern: other risk factors and their treatment (triglycerides, remnant lipoproteins, low HDLcholesterol, lifestyle change for atherogenic dyslipidemia, hypertension treatment per JNC guidelines, smoking prevention and cessation, presence of diabetes as a separate risk category, obesity, physical inactivity, atherogenic diet, age, male sex, family history, and metabolic syndrome); goal levels of LDLcholesterol lowering and policies for routine cholesterol testing; clinical noncoronary atherosclerosis and type 2 diabetes as CHD risk equivalents, and use of multidisciplinary teams for their management; cost effectiveness of drugs for lowering cholesterol; dietary aspects of management (weight loss, saturated fatty acids, trans-fatty acids, dietary cholesterol, monounsaturated fatty acids, linoleic acid, unsaturated fatty acids, total fat intake, carbohydrate intake, viscous fiber intake, plant stanol/sterol
esters, soy protein, n-3 fatty acids, folate, dietary antioxidants, alcohol, salt, herbal and botanical dietary supplements, high protein, high total fat, and saturated fat weight loss regimens); lipid-lowering drugs (statins, bile acid sequestrants, nicotinic acid, and fibrates); and hormone replacement therapy. Subsequent sections of the report address issues in detection and evaluation, the general approach to treatment, lifestyle interventions, dietary interventions, drug therapy, management of specific dyslipidemias, special considerations for different population groups, and adherence to drug therapy. Detailed algorithms are presented to guide the physician at each step of decision making in management of the individual patient, reflecting the extensive evidence and numerous recommendations in the report (see also Chapter 11). This report presents detailed discussion of the rationale and methods of global risk estimation, as adopted for ATP III, based on a model from the Framingham Heart Study to predict “hard CHD,” in this case meaning myocardial infarction or CHD death. Table 20-2 presents the scoring weight derived from the model for each level of several contributing variables, separately for men and women. Variables included were total cholesterol concentration and smoking status by age, from 20–39 to 70–79 years; systolic blood pressure, separately for persons treated or untreated at the time of measurement; and HDLcholesterol concentration. The “points” for a given level of a risk factor differ by sex and are typically greater for women. This is because the baseline risk for men is already greater than for women and the effect of higher risk-factor levels in women is the tendency to catch up with the risk for men of the same age. For men, 10-year risks of acute coronary events increase from less than 1% to 30% over a range of total points from 0 to 17 or greater; for women, the same range of risks corresponds to a range of total points from 9 to 25 or greater. The number of points for a given level of total cholesterol or smoking declines with age. This does not mean that these predictors are less important at older age. Rather, it reflects the fact that at older versus younger ages fewer points represent as great or greater increase in absolute 10-year risk. In addition to this categorical approach to risk estimation, ATP III also calls attention to an electronic calculator that more precisely estimates risk by use of continuous values for the risk factors (available at www.nhlbi.nih .gov/guidelines/cholesterol).7 The risk score is used in decision making for treatment in the following way: Persons with existing CHD are considered separately from candidates for
46894_CH20_591_618.qxd
596
2/25/10
4:15 PM
Page 596
CHAPTER 20 RECOMMENDATIONS, GUIDELINES, AND POLICIES
Table 20-2
10-Year Risk Estimates for Men and Women (Framingham Point Scores) Points at Points at Points at Points at Ages 20–39 Ages 40–49 Ages 50–59 Ages 60–69 Total Cholesterol M F M F M F M F 160 0 0 0 0 0 0 0 0 160–199 4 4 3 3 2 2 1 1 200–239 7 8 5 6 3 4 1 2 240–279 9 11 6 8 4 5 2 3 280 11 13 8 10 5 7 3 4 Nonsmoker 0 0 0 0 0 0 0 0 Smoker 8 9 5 7 3 4 1 2 SBP 120 120–129 130–139 140–159 160 Males Point Total 10-Year Risk (%) Point Total 10-Year Risk (%) Females Point Total 10-Year Risk (%) Point Total 10-Year Risk (%)
Points at Ages 70–79 M F 0 0 0 1 0 1 1 2 1 2 0 0 1 1
If untreated 0 0 0 1 1 2 1 3 2 4
If treated 0 1 2 2 3
0 1 9 5
0 1 10 6
1 1 11 8
2 1 12 10
3 1 13 12
4 1 14 16
5 2 15 20
6 2 16 25
7 3 17 30
8 4
9 1 18 6
9 1 19 8
10 1 20 11
11 1 21 14
12 1 22 17
13 2 23 22
14 2 24 27
15 3 25 30
16 4
17 5
0 3 4 5 6
HDL 60 1 50–59 0 40–49 1 40 2
M, male; F, female; HDL, high-density-lipoprotein cholesterol; SBP, systolic blood pressure. See text for explanation. Source: Data from National Cholesterol Education Program, Third Report from the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) Final Report. NIH Publication No. 02-5215, September 2002, pp III-4–III-5.
primary prevention; they are defined a priori as highrisk patients for whom maximum intensity of cholesterol-lowering treatment is indicated. For persons without existing CHD, intensity of treatment is intended to match the severity of risk. The first consideration is the number of risk factors present beside LDL-cholesterol. Included in the risk-factor count are age, family history, smoking, hypertension, and low HDL-cholesterol (high HDL-cholesterol is protective and offsets 1 risk factor if present). Persons with two or more of these risk factors are then classified in one of three categories of 10-year CHD risk, based on the risk score: 20%, 10–20%, or 10%. Depending on the estimated risk, the goal level for LDLcholesterol lowering and the levels at which lifestyle change is instituted and, if necessary, drug therapy is added to lifestyle change are set successively lower. Higher goal levels are set for those with only zero to one other risk factors. Other guidelines developed under the aegis of NHLBI concern blood cholesterol in children and
adolescents; high blood pressure in adults and, separately, in children and adolescents; obesity; and other conditions. A new approach to this process at NHLBI, in some respects anticipated by ATP III, is expected to replace risk-factor-specific guidelines with more integrated recommendations. The NHLBI Web site cited previously will likely be useful as a continuing source for current information. Prevention of Ischemic Stroke A wide range of topics in CVD prevention is also represented in the many Scientific Statements and Guidelines published by the American Heart Association (AHA), whether as sole sponsor or in collaboration with its component organization, the American Stroke Association (ASA), or the American College of Cardiology (ACC). The ACC/AHA procedures for guideline development reviewed in Chapter 19 have been typical of these documents. Examples of topics addressed are diet and lifestyle, primary prevention of
46894_CH20_591_618.qxd
2/25/10
4:15 PM
Page 597
CLINICAL GUIDELINES 597
CVD and stroke, reducing delay in treatment-seeking with impending heart attack or stroke, and cardiac rehabilitation and secondary prevention for survivors with CHD.8–13 These clinical guidelines are updated from time to time, as new evidence dictates. One example is the 2006 AHA/ASA guideline on primary prevention of ischemic stroke, an update of previous reports.11 The process for rating levels of evidence and recommendations and the estimated certainty and size of treatment effects was as in Table 19-3 and Figure 19-3. Risk factors for stroke were reviewed and classified as “nonmodifiable,” “well-documented and modifiable,” or “less welldocumented or potentially modifiable.” Nonmodifiable factors identified were age, sex, race, low birth weight, and family history of stroke or TIA. For each modifiable or potentially modifiable factor, data were presented on its prevalence, populationattributable risk for stroke, relative risk, and risk reduction with treatment if known. A “modified Framingham stroke risk profile” was calculated on the basis of a subset of the welldocumented modifiable factors. Analogous to the coronary risk score used in ATP III, this profile assigned points to specified categories of each factor, for individual patients aged from 54 to 85 years. Assignable points ranged from 0 to 10 for both men and women; blood pressure levels were rated by age category; and diabetes, smoking, cardiovascular disease, atrial fibrillation, and left ventricular hypertrophy on electrocardiogram were rated as present or absent. Numbers of points corresponded to the relative potency of each factor’s contribution to risk. The 10-year probability of stroke ranged from 3% to 88% for men with scores from 1 to 30 points and from 1% to 84% for women with scores from 1 to 27 points. Recommendations for reducing risk were presented regarding for each identified factor. Blood pressure management was to be in accordance with the NHLBI-sponsored 7th Joint National Committee Report (see Chapter 12). Similarly, management of dyslipidemia was in accordance with ATP III. For each of 24 other factors, evidence was summarized and, where it was found to be sufficient, an intervention was recommended. Factors emphasized were atrial fibrillation, cigarette smoking, diabetes, asymptomatic carotid stenosis, sickle cell disease, poor nutrition, alcohol, drug abuse, oral contraceptive use, and sleep-disordered breathing. Europe The European Society of Cardiology (ESC) is extensively involved in development of clinical pre-
ventive guidelines, publishing 33 of them from 2000 through 2007.14 Growth of partnerships in guideline development among key organizations—initially the ESC, European Atherosclerosis Society, and European Society of Hypertension, was outlined in a 2005 report by Graham.15 Joint European recommendations were first developed in 1994. By the time the Third Joint Task Force Recommendations were published in 2003, organizations representing family practice, behavioral medicine, and diabetes were included, as well as the European Heart Network of lay-oriented foundations. As emphasized by Graham:15, p 431 The philosophy behind the European Recommendations is to harmonize the advice from as many major bodies as possible with an interest in prevention. Each individual partner is encouraged to develop their own, more detailed guidelines, but it is hoped that they will be seen as an extension of the basic Joint Recommendations and specifically compatible with them. . . . A key tenet of the Joint European Recommendations from the first has been the concept of total risk. The guidelines recognize that atherosclerotic CVDs are the product of the interaction of multiple causes. This supports the concept of Joint Recommendations, rather than separate expert groups on, for example, blood lipids, hypertension, and smoking. The Third Joint Task Force Recommendations represent further evolution of several underlying concepts. The target of prevention was CVD—not limited to CHD but including ischemic stroke and peripheral arterial disease. However, the European data for risk estimation were based on CVD death, not including non-fatal CVD, from 12 pooled cohort studies. These comprised 200,000 persons and more than 7000 fatal CVD events. The risk threshold for treatment was set at 5% or greater risk in 10 years, a level thought to include the majority of persons at risk of nonfatal CVD within the same time interval. The risk assessment model used was the “Systematic COronary Risk Evaluation (SCORE)” model. This model has flexibility to take into account different levels of population risk among European countries. The score is based on sex, smoking status, systolic blood pressure, and blood cholesterol concentration. The resulting risk scores were distributed as shown in “high-risk” and “low-risk” regions of Europe in Figures 20-1 and 20-2, respectively. Low-risk countries are identified in Table 20-3: Belgium, France, Greece, Italy, Luxembourg, Spain, and Portugal. All
46894_CH20_591_618.qxd
598
2/25/10
4:15 PM
Page 598
CHAPTER 20 RECOMMENDATIONS, GUIDELINES, AND POLICIES
Women Non-Smoker
Systolic Blood Pressure
180 7
8
Men
Age
Smoker
9 10 12
Non-Smoker
Smoker
13 15 17 19 22
14 16 19 22 25
25 30 35 41 47
9 11 13 15 16
18 21 25 29 34
160 6
5
6
7
8
9 10 12 13 16
140 3
3
4
5
6
6
7
8
9 11
6
9
9 11 13
13 15 17 20 24
120 2
2
3
3
4
4
6
6
6
4
5
6
9
9 10 12 14 17
180 4
4
6
6
7
8
9 10 11 13
9 11 13 15 18
18 21 24 28 33
160 3
3
3
4
5
6
6
7
8
9
6
7
9 10 12
12 14 17 20 24
140 2
2
2
3
3
3
4
5
5
6
4
5
6
7
9
8 10 12 14 17
120 1
1
2
2
2
2
3
3
4
4
3
3
4
5
6
6
180 2
2
3
3
4
4
6
6
6
7
6
7
8 10 12
12 13 16 19 22
15% and over
160 1
2
2
2
3
3
3
4
4
6
4
5
6
7
8
8
9 11 13 16
10%–14%
140 1
1
1
1
2
2
2
2
3
3
3
3
4
5
6
5
6
8
9 11
120 1
1
1
1
1
1
1
2
2
2
2
2
3
3
4
4
4
5
6
180 1
1
1
2
2
2
2
3
3
4
4
4
5
6
7
7
8 10 12 14
1%
160 1
1
1
1
1
1
2
2
2
3
2
3
3
4
6
5
6
7
8 10
< 1%
140 0
1
1
1
1
1
1
1
1
2
2
2
2
3
3
3
4
5
6
7
120 0
0
1
1
1
1
1
1
1
1
1
1
2
2
2
2
3
3
4
6
180 0
0
0
0
0
0
0
0
1
1
1
1
1
2
2
2
2
3
3
4
160 0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
2
2
2
3
1
1
1
2
2
65
7
60
55
7
7
8 10 12
8
SCORE 5%–9% 3%–4% 2%
50
40
140 0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
120 0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
4
5
6
7
8
4
5
6
7
8
4
5
6
7
8
4
5
6
7
8
Cholesterol mmol
10-year risk of fatal CVD in populations at high CVD risk
150 200 250 300 mg/dl
Figure 20-1 Ten-Year Risk of Fatal CVD in High-Risk Regions of Europe by Sex, Age, Systolic Blood Pressure, Total Cholesterol, and Smoking Status. Source: Reprinted with permission from IM Graham, Current Opinion in Cardiology, Vol 20, © 2005 Lippincott Williams and Wilkins, p 434.
other countries of Europe were high risk. In both high-risk and low-risk populations, risk at a given level of risk factors was greater for men, smokers, and older persons and increased with systolic blood pressure and total cholesterol concentration. Risks for men were about two times those for women at age 65 years and three times greater at ages 55 years and younger, in both categories of countries; risks for men in low-risk countries were about the same as for women in high-risk countries. Table 20-3 presents instructions on use of the tables in patient education to demonstrate an individual’s risk, the relative risk of different risk-factor combinations, and the effect on risk of reducing one or more factors from the present to a more favorable category. Two recommendations were especially note-
worthy: First, persons at low risk should be advised how to keep their risk low. Second, those whose currently estimated risk would be expected to increase to treatable levels simply by reaching middle age “should be given maximal attention.”15 p 436 In this way, the Joint European Guidelines extend beyond persons already at high short-term risk to reach a lower-risk stratum of the population and to distinguish between persons at higher and lower long-term risk. Worldwide The WHO approach to CVD risk assessment was reviewed in Chapter 19.16 The resulting management recommendations are presented in Table 20-4. The intended reach of these recommendations was world-
46894_CH20_591_618.qxd
2/25/10
4:15 PM
Page 599
CLINICAL GUIDELINES 599
Women
Systolic Blood Pressure
Non-Smoker
Men
Age
Smoker
Non-Smoker
Smoker
8
9 10 12 14
15 17 20 23 26
5
6
7
9 10
10 12 14 16 19
4
4
5
6
7
7
8
9 11 13
4
2
3
3
4
5
5
6
5
7
8
5
6
7
8
9
10 11 13 15 18
4
5
6
3
4
5
5
6
7
8
9 11 13
3
3
4
2
3
3
4
4
5
6
6
7
9
2
2
2
3
2
2
2
3
3
3
4
4
6
6
3
3
3
4
4
3
4
4
5
6
6
7
8 10 12
15% and over
2
2
2
3
3
2
2
3
3
4
4
5
6
7
8
10%–14%
1
1
1
1
2
2
1
2
2
2
3
3
3
4
6
6
5%–9%
1
1
1
1
1
1
1
1
1
1
2
2
2
2
3
3
4
1
1
1
1
1
2
2
2
2
2
3
3
4
4
4
5
6
7
1%
0
1
1
1
1
1
1
1
1
1
1
2
2
2
2
3
3
4
6
< 1%
140 0
0
0
0
0
1
1
1
1
1
1
1
1
1
2
2
2
2
3
3
120 0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
2
2
2
180 0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
2
2
160 0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
140 0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
120 0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
4
5
6
7
8
4
5
6
7
8
4
5
6
7
8
4
5
6
7
8
180 4
5
6
6
7
8
9 11 12 14
160 3
3
4
4
5
5
6
7
8 10
140 2
2
2
3
3
4
4
5
6
7
120 1
1
2
2
2
3
3
3
4
180 3
3
3
4
4
5
6
6
160 2
2
2
2
3
3
4
140 1
1
1
2
2
2
2
120 1
1
1
1
1
1
180 1
1
2
2
2
160 1
1
1
1
1
140 1
1
1
1
120 0
0
1
180 1
1
160 0
65
60
55
8
9
SCORE 3%–4% 2%
50
40
Cholesterol mmol
10-year risk of fatal CVD in populations at low CVD risk
150 200 250 300 mg/dl
Figure 20-2 Ten-Year Risk of Fatal CVD in Low-Risk Regions of Europe by Sex, Age, Systolic Blood Pressure, Total Cholesterol, and Smoking Status. Source: Reprinted with permission from IM Graham, Current Opinion in Cardiology, Vol 20, © 2005 Lippincott Williams and Wilkins, p 435.
wide, as indicated in an accompanying table (Appendix 2, discussed in Chapter 21). This table shows the proportion of population in each category of estimated risk (from 10% to 40%) by sex and age (from 50 to 70 years), for 14 subregions representing all six WHO Regions of the world. Especially in low- and middle-income countries, where CVD was an already-present burden but resources to address it faced competing priorities, it was argued that “it is imperative to target the limited resources on those who are most likely to benefit.”16, p 2 As shown in Table 20-4, each recommendation was annotated as to the underlying evidence rating. Strong recommendations were considered likely to apply everywhere, whereas those that were weak would depend, at the individual level, on patient-specific considera-
tions and, on the population level, on views of a wide range of stakeholders. Recommendations were stratified by 10-year risk of a fatal or nonfatal vascular event ( 30%, 20–30%, 10–20%, and 10%). Persons with established cardiovascular conditions (coronary artery disease, cerebrovascular disease, and peripheral vascular disease) were presumed to require intervention appropriate to the risk factors present so are not included in these stratified recommendations. Policy interventions were considered important to establish conditions conducive to behavior change—such actions would benefit even those at low risk, especially when accompanied by lifestyle interventions. Some recommended interventions would apply regardless of risk score—improving smoking, diet, physical activity, weight, and alcohol
46894_CH20_591_618.qxd
600
2/25/10
4:15 PM
Page 600
CHAPTER 20 RECOMMENDATIONS, GUIDELINES, AND POLICIES
Table 20-3
•
•
•
•
•
•
Instructions on How to Use the Chart (Figures 20-1 and 20-2) The low-risk chart should be used in Belgium, France, Greece, Italy, Luxembourg, Spain, and Portugal; the high-risk chart should be used in all other countries of Europe. To estimate a person’s total 10-year risk of cardiovascular disease death, find the table for their sex, smoking status, and age. Within the table, find the cell nearest the person’s systolic blood pressure (mmHg) and total cholesterol (mmol/l or mg/dl). The effect of lifetime exposure to risk factors can be seen by following the table upward. This can be used when advising younger people. Low-risk individuals should be offered advice to maintain their low-risk status. Those who are at 5% risk or higher or will reach this level in middle age should be given maximal attention. To define a person’s relative risk, compare their risk category with that of a nonsmoking person of the same age and sex; blood pressure 140/90 mmHg and total cholesterol 5 mmol/l (190 mg/dl). The chart can be used to give some indications of the effect of changes from one risk category to another for example, when the subject stops smoking or reduces other risk factors.
Source: Reprinted with permission from IM Graham, Current Opinion in Cardiology, Vol 20, © 2005 Lippincott Williams and Wilkins, p 436.
intake. Other interventions would be tailored to specific risk strata, each with its own evidence rating— drugs for smoking cessation; antihypertensive drugs; lipid-lowering drugs; hypoglycemic drugs; and antiplatelet drugs. These worldwide recommendations were proposed as a universal approach to risk assessment and management for CVD prevention. At the same time, the need to adapt recommendations “to suit different political, economic, social, cultural and medical conditions” was recognized.16, p 3 Differences remain among national, regional, and worldwide guidelines for preventing first cardiovascular events, from ratings of evidence to strength of recommendations and scope of eligibility for intervention. Should the harmonizing advice of the World Heart and Stroke Forum prevail, these differences may be resolved in the future. Secondary Prevention Clinical guidelines are not limited to primary prevention but have also been developed for secondary prevention, or prevention of recurrent cardiovascular events. Secondary prevention includes reducing disability resulting from prior CVD events. Therefore, rehabilitation has been included in some secondary prevention guidelines. One example is the joint
Scientific Statement from AHA and the American Association of Cardiovascular and Pulmonary Rehabilitation (AACVPR) on cardiac rehabilitation and secondary prevention.17 This update to a report in 2000 that addressed core components of cardiac rehabilitation programs was occasioned by the need for consistency with other, more current, recommendations for secondary prevention. This new guideline presented detailed directives as to several aspects of patient care and indicated goals for patient evaluation, interventions, and expected outcomes under seven areas: patient assessment; nutritional counseling; management of weight, blood pressure, lipids, and diabetes; tobacco cessation; psychosocial management; physical activity counseling; and exercise training. The intended impact of these recommendations was to establish in all cardiac rehabilitation/secondary prevention programs “specific core components that aim to optimize cardiovascular risk reduction, foster healthy behaviors and compliance to these behaviors, reduce disability, and promote an active lifestyle for patients with cardiovascular disease.”17, p 2675 Groups of Special Concern Recommendations for individual-level interventions depend heavily on evidence from RCTs. The question of whether findings in the study population of a trial (internal validity) predict the impact of intervention in the population at large (external validity) was discussed in Chapter 19. This question takes a special form when recommendations are desired for one or another subgroup of the general population that is not adequately represented in the relevant trials. This issue was addressed in ATP III, in a section titled “Special Considerations for Different Population Groups”:6, p VIII-1 . . . randomized clinical trials have not been carried out to address all therapeutic questions pertaining to all age groups, both sexes, and different racial/ethnic groups. Consequently, ATP III recommendations must be made by combining what has been learned from clinical trials with other lines of evidence such as epidemiologic findings. . . . No attempt will be made to grade the category and strength of evidence for all recommendations made in this section. Discussion follows on several specific groups: middle-aged men (35–65 years); women; older persons (men 65 years, women 75 years); younger adults (men 20–35 years, women 20–45 years); and racial and ethnic groups—African Americans, Hispanic Americans, Native Americans (American Indians), Asian and Pacific Islanders, and South Asians.
46894_CH20_591_618.qxd
2/25/10
4:15 PM
Page 601
CLINICAL GUIDELINES 601
Prevention of Cardiovascular Disease According to Individual Total Riska 10-Year Risk of 10-Year Risk of 10-Year Risk of Cardiovascular Event Cardiovascular Event Cardiovascular Event 30% 20–30% 10–20% Individuals in this category Individuals in this category Individuals in this category are at moderate risk of fatal are at high risk of fatal or are at very high risk of fatal or nonfatal vascular events. nonfatal vascular events. or nonfatal vascular events. Table 20-4
Monitor risk profile every 3–6 months.
Monitor risk profile every 3–6 months.
Monitor risk profile every 6–12 months.
10-Year Risk of Cardiovascular Event 10% Individuals in this category are at low risk. Low risk does not mean “no” risk. Conservative management focusing on lifestyle interventions is suggested.b
When resources are limited, individual counselling and provision of care may have to be prioritized according to cardiovascular risk. SMOKING CESSATION All nonsmokers should be encouraged not to start smoking. All smokers should be strongly encouraged to quit smoking by a health professional and supported in their efforts to do so. (1, A) It is suggested that those who use other forms of tobacco be advised to stop. (2, C) Nicotine replacement therNicotine replacement apy and/or nortriptyline therapy and/or nortriptyor amfebutamone (buproline or amfebutamone pion) should be given to (bupropion) should be motivated smokers who given to motivated smokfail to quit with couners who fail to quit with selling. (1, B) counselling. (1, B) DIETARY CHANGES All individuals should be strongly encouraged to reduce total fat and saturated fat intake (1+, A). Total fat intake should be reduced to about 30% of calories, saturated fat intake should be limited to less than 10% of calories and trans-fatty acids eliminated. Most dietary fat should be polyunsaturated (up to 10% of calories) or monounsaturated (10–15% of calories). (1, A) All individuals should be strongly encouraged to reduce daily salt intake by at least one-third and, if possible, to 5 g or 90 mmol per day. (1, A) All individuals should be encouraged to eat, at least 400 g a day, of a range of fruits and vegetables, as well as whole grains and pulses. (2, A) PHYSICAL ACTIVITY All individuals should be strongly encouraged to take at least 30 minutes of moderate physical activity (e.g., brisk walking) a day, through leisure time, daily tasks and work-related physical activity. (1, A) WEIGHT CONTROL All individuals who are overweight or obese should be encouraged to lose weight through a combination of a reduced-energy diet (dietary advice) and increased physical activity. (1, A) ALCOHOL INTAKE Individuals who take more than 3 units of alcohol per day should be advised to reduce alcohol consumption. (2, B) c
ANTIHYPERTENSIVE DRUGS ✓ All individuals with blood pressure at or above 160/100 mm Hg, or lesser degree of raised blood pressure with target organ damage should have drug treatment and specific lifestyle advice to lwer their blood pressure and risk of cardiovascular disease. (2, B) Individuals with persistent blood pressure 130/80 mm Hg should be given one of the flowing drugs to reduce blood pressure and risk of cardiovascular disease: thiazide-like diuretic, ACE inhibitor,
Individuals with persistent blood pressure 140/90 mm Hge who are unable to lower blood pressure through life style strategies with professional assistance within 4–6 months, should be
Individuals with persistent blood pressure 140/90 mm Hg,e should continue life style strategies to lower blood pressure and have their blood pressure and total cardiovascular risk reassessed annually
Individuals with persistent blood pressure 140/90 mm Hg,e should continue life style strategies to lower blood pressure and have their blood pressure and total cardiovascular risk reassessed every two continues
46894_CH20_591_618.qxd
602
2/25/10
4:15 PM
Page 602
CHAPTER 20 RECOMMENDATIONS, GUIDELINES, AND POLICIES
Prevention of Cardiovascular Disease According to Individual Total Riska—continued 10-Year Risk of 10-Year Risk of 10-Year Risk of 10-Year Risk of Cardiovascular Event Cardiovascular Event Cardiovascular Event Cardiovascular Event 30% 10% 20–30% 10–20% depending on clinical cirto five years depending considered for one of the calcium-channel blocker, cumstances and resource on clinical circumstances following drugs to reduce beta-blocker.d availability. and resource availability. blood pressure and risk of A low-dose thiazide-like dicardiovascular disease: uretic, ACE inhibitor, or thiazide-like diuretic, ACE calcium-channel blocker is inhibitor, calcium-channel recommended as first-line blocker, beta-blocker.d therapy. (1, A) A low-dose thiazide-like diuretic, ACE inhibitor, or calcium-channel blocker is recommended as firstline therapy. (1, A) Table 20-4
LIPID-LOWERING DRUGS (STATINS) ✓ All individuals with total cholesterol at or above 8 mmol/l (320 mg/dl), should be advised to follow a lipid-lowering diet and given a statin to lower the risk of cardiovascular disease. (2, B) Individuals in this risk category should be advised to follow a lipid-lowering diet and given a statin. (1, A) Serum cholesterol should be reduced to less than 5.0 mmol/l (LDL-cholesterol to below 3.0 mmol/l), or by 25% (30% for LDLcholesterol) which ever is greater.f
Adults over the age of 40 years with persistently high serum cholesterol ( 5.0 mmol/l), and or LDL-cholesterol 3.0 mmol/l, despite a lipidlowering diet, should be given a statin. (1, A)
Should be advised to follow a lipid lowering diet.g
HYPOGLYCEMIC DRUGS ✓ Individuals with persistent fasting blood glucose 6 mmol/l despite diet control should be given metformin. (1, A)
Recommendations as for moderate risk, as resources permit.
ANTIPLATELET DRUGS ✓ Individuals in this risk category should be given lowdose aspirin. (1, A)
For individuals in this risk category cardiovascular risk, the balance of benefits and harms from aspirin treatment is not clear.h Aspirin should probably not be given to individuals in this risk category. (1, A)
For individuals in this risk category, the benefits of aspirin treatment are balanced by the harm caused. Aspirin should not be given to. (1, A)
For individuals in this risk category, the harm caused by aspirin treatment outweighs the benefits. Aspirin should not be given to individuals in this lowrisk category. (1, A)
DRUGS THAT ARE NOT RECOMMENDED Hormone replacement, vitamin B, C, E and folic acid supplements, are not recommended for reduction of cardiovascular risk. a
Excluding people with established coronary artery disease, cerebrovascular disease, and peripheral vascular disease. Policy measures that create conducive environments for quitting tobacco, engaging in physical activity, and consuming healthy diets are necessary to promote behavioral change. They will benefit the whole population. For individuals in low risk categories, they can have a health impact at lower cost, compared to individual counselling and therapeutic approaches. c One unit (drink) half pint of beer/lager (5% alcohol), 100 ml of wine (10% alcohol), spirits 25 ml (40% alcohol). b
continues
46894_CH20_591_618.qxd
2/25/10
4:15 PM
Page 603
CLINICAL GUIDELINES 603
Table 20-4
Prevention of Cardiovascular Disease According to Individual Total Riska—continued
d
Evidence from two recent meta-analyses indicates that beta-blockers are inferior to calcium-channel blockers and ACE inhibitors in reducing the frequency of hard endpoints. In addition, beta-blockers are less well tolerated than diuretics (see Part III, section 4). Most of this evidence comes from trials where atenolol was the beta-blocker used. e Reducing blood pressure by 10–15/5–8 mm Hg with drug treatment reduces combined CVD mortality and morbidity by about one-third, whatever the pretreatment absolute risk. However, applying this recommendation will lead to a large proportion of the adult population receiving antihypertensive drugs. Even in some high-resource settings, current practice is to recommend drugs for this group only if the blood pressure is at or above 160/100 mm Hg. f Reducing cholesterol level by 20% (approximately 1 mmol/l) with statin treatment would be expected to yield a coronary heart disease mortality benefit of 30%, whatever the pretreatment absolute risk. However, applying this to the general population may not be cost effective. It will lead to a large proportion of the adult population receiving statins. Even in some high-resource settings, current practice is to recommend drugs for this group only if serum cholesterol is above 8 mmol/l (320 mg/dl). g There are no clinical trials that have evaluated the absolute and relative benefits of cholesterol lowering to different cholesterol targets in relation to clinical events. h Consider aspirin in areas where coronary heart disease rates exceed stroke rates. ✓Best Practice point: Unless there are compelling indications to use a specific drug, the least expensive preparation of the above classes of drugs should be used. Good quality generic preparations of medicines listed in WHO essential medicines list are recommended. Source: Adapted with permission from World Health Organization, Prevention of Cardiovascular Disease. Guidelines for Assessment and Management of Cardiovascular Risk, © World Health Organization 2007, pp 22–26.
Details of special considerations for cholesterol management are tabulated for each of the identified agesex subgroups and for African Americans. Several points are noteworthy regarding these groups. For middle aged men, recommendations for cholesterol-lowering interventions were given separately according to risk level and emphasize lifestyle intervention at lower risk (10-year risk 10%). For women, the 10- to 15-year relative delay in onset of clinical CHD was the basis for providing recommendations at ages 45–75 years, in contrast to ages 35–65 years for men; there were no other substantial differences for women and for men. For older persons, the reliability of risk prediction is less certain and subclinical atherosclerosis may increase risk even though it is unapparent. Therefore noninvasive assessment of atherosclerosis was considered potentially useful in judging the appropriate intensity of cholesterol-lowering intervention. For younger adults, beginning from age 20 years for both men and women, testing for lipids and lipoproteins was recommended, with drug therapy mainly limited to those with LDL-cholesterol concentrations of 190 mg/dl or greater. For African Americans, it was recognized that the relative excess of high blood pressure and left ventricular hypertrophy (LVH) may not be accounted for adequately by the Framingham risk score, which does not include LVH as a predictor. Regardless of this limitation, the ATP III guidelines were treated as equally applicable to African Americans and Whites. For the remaining groups, accuracy of Framingham risk predictions has been questioned and some special considerations may apply—especially for South Asians, with a relatively high prevalence of CHD at earlier ages than Whites—but these and other differences were judged to be insufficient to warrant separate algorithms for cholesterol management for any of these groups.
Women An AHA Guideline for primary and secondary CVD prevention in women, published in 2007, updated an evidence-based review reported in 2004 and a still earlier guide from 1997.18 The most recent report was based on 246 accepted articles in which women were included—RCTs, large prospective cohort studies, and surrogate endpoint studies. Women were to be classified as “high risk,” “at risk,” or “optimal risk.” Lifetime risk rather than short-term (10-year) risk was emphasized, on grounds of overall high CVD risk in women and an otherwise unduly narrow focus on only 10-year risk. The Framingham risk score was less heavily relied on by this review group than by others, on three bases in the group’s assessment: omission of family history; understatement of risk in women with subclinical disease; and unreliability as a predictor in non-White populations. Twenty-eight clinical recommendations were presented, with the corresponding classification and level of evidence for each, in the areas of lifestyle interventions, major risk-factor interventions, preventive drug interventions, and Class III interventions (“not useful or effective” or “may be harmful”). Use of aspirin for primary prevention of heart disease was one of few examples of differences in recommendations between women and men: Its use was not recommended to prevent MI in women 65 years of age or in women aged 65 years if blood pressure was not controlled or risks outweighed benefits. A detailed algorithm outlined the process of stratification on presence or absence of high risk (presence of CVD or a global 10-year risk 20%). Lifestyle recommendations applied to all women regardless of risk level. Those classified as high risk and having a recent CVD event would be referred for rehabilitation. Research was proposed on several topics including the impact of the guidelines and effectiveness of their implementation in various settings.
46894_CH20_591_618.qxd
604
2/25/10
4:15 PM
Page 604
CHAPTER 20 RECOMMENDATIONS, GUIDELINES, AND POLICIES
The Elderly An AHA Scientific Statement addressed secondary prevention of CHD in the elderly, especially persons 75 years of age.19 (No reference was made to primary prevention in this group. But, while obstructive coronary artery disease is found at autopsy in 70% to 80% in elderly men and 50% in elderly women—half of elderly women and one in four elderly men might still be candidates for primary prevention.) Recommendations in this report addressed clinical intervention for management of hypertension, abnormal blood lipids, obesity, diabetes, psychosocial factors, physical inactivity, and cardiac rehabilitation. Unlike the more formal evidence synthesis and evaluation that underlies the previously discussed recommendations, this report was based on expert opinion, with no attempt to rate the evidence or recommendations. Children and Adolescents Recommendations for CVD prevention at the individual level in childhood and adolescence have been published by AHA, NHLBI, and others over the past 20 years. Reports specific to one risk factor—such as blood pressure, cholesterol, smoking, or obesity—
Table 20-5 Cardiovascular Health Schedule Birth • Family history for early coronary heart disease, hyperlipidemia → if positive, introduce risk factors; parental referral • Start growth chart • Parental smoking history → smoking cessation referral 0–2 Years • Update family history, growth chart • With introduction of solids, begin teaching about healthy diet (nutritionally adequate, low in salt, low in saturated fats) • Recommend healthy snacks as finger foods • Change to whole milk from formula or breast feeding at approximately 1 year of age 2–6 Years • Update family history, growth chart → review growth charta with family (concept of weight for height) • Introduce prudent diet ( 30% of calories from fat) • Change to low-fat milk • Start blood pressure chart at approximately 3 years of age;b review for concept of lower salt intake • Encourage active parent-child play
are addressed in Part III, presented previously. Perhaps the first recommendations for comprehensive clinical preventive measures against CVD in childhood appeared in a Special Report of the AHA in 1992.20 This report incorporated guidelines for a “cardiovascular health schedule” and accompanying action by healthcare providers for all children—irrespective of individual risk—from birth to ages over 10 years, presumably to age 17 or 18 (Table 20-5). At birth, the schedule called for ascertainment of the family history of early CHD or hyperlipidemia. Because this is a dynamic characteristic of the family history, it was to be updated throughout childhood. According to this schedule, the child’s growth chart would be initiated at birth and updated periodically. The parental smoking history would also be ascertained at this time, with referral for smoking cessation if positive for either parent. From birth to 2 years of age, dietary recommendations were the focus of intervention. Foods would be introduced to establish a dietary pattern free of high fat and salt content, and whole milk would be substituted for breast milk beginning at age 1 year. At ages 2 to 6 years, the growth chart was to be reviewed to introduce to parents the concept of the child’s
• Lipid determination in children with positive family history or with parental cholesterol 240 mg/dl (obtain parental lipid levels if necessary) → if abnormal, initiate nutrition counseling 6–10 Years • Update family history, blood pressure and growth charts • Complete cardiovascular health profile with child; determine family history, smoking history, blood pressure percentile, weight for height, fingerstick cholesterol, and level of activity and fitness • Reinforce prudent diet • Begin active antismoking counseling • Introduce fitness for health → life sport activities for child and family • Discuss role of watching television in sedentary lifestyle and obesity 10 Years • Update family history, blood pressure and growth charts annually • Review prudent diet, risks of smoking, fitness benefits whenever possible • Consider lipid profile in all patients • Final review of personal cardiovascular health status
If weight 120% of normal for height, diagnosis of obesity should be considered and the subject addressed with child and family. If three consecutive interval blood pressure measurements exceed the 90th percentile and blood pressure is not explained by height or weight, diagnosis of hypertension should be made and appropriate evaluation considered.
a
b
Source: Reprinted with permission from Circulation, Vol 85, p 1648, Copyright 1992, American Heart Association.
46894_CH20_591_618.qxd
2/25/10
4:15 PM
Page 605
CLINICAL GUIDELINES 605
weight in relation to height. The “prudent diet” would be introduced, with the goals of less than 30% of calories from fat and replacement of whole milk by low-fat milk. Active parent-child play would be encouraged to involve mutual participation in physical activity. Monitoring of blood pressure would be undertaken from age 3 years, with intervention if readings were persistently high and advice for all children to maintain low salt intake. Blood lipids would be evaluated for those children whose family history indicated early CHD or hypercholesterolemia and then dietary advice would be provided if indicated. (Parental lipid levels were to be determined if necessary.) At ages 6 to 10 years, in addition to updating the assessments of family history, blood pressure, and growth, the cardiovascular profile would be completed. This included the parent’s smoking history, weight for height, finger-stick cholesterol determination, and level of physical fitness and activity. Counseling was to be provided on diet, physical activity, tobacco, and the role of television watching in leading to sedentary habits and obesity. Continuing from age 10, annual updates of family history, growth charts, and blood pressure were recommended. Counseling on the prudent diet, risks of smoking, and benefits of fitness would occur “whenever possible.” Laboratory determination of the lipid profile would be considered in all patients. The “final review” of cardiovascular health status presumed a child’s valedictory physician visit for a precollege or preemployment final examination. Throughout this period, reliance on reference charts for blood pressure measurements would continue. A decade later, AHA recommendations for CVD prevention in children and adolescents adopted the concept of identifying high-risk individuals.21 Guidelines in this 2003 update presented health promotion goals for all children and adolescents with respect to diet, smoking, and physical activity. Individual assessment was proposed for risk stratification with respect to lipids and lipoproteins, blood pressure, and body size. Goals were specified for each of several risk factors—blood cholesterol, other lipids and lipoproteins, blood pressure, weight, diabetes, and cigarette smoking. For the child or adolescent whose assessment indicated that he or she was not at goal for a given risk factor, specific interventions were recommended with the intent to reach that goal. Evidence was cited in general support of those recommendations and included studies of pathology of atherosclerosis in the young and the impact of risk factors in this period of life; prevalence in youth of obesity and type 2 diabetes; tracking of risk factors from childhood into adult life; acquisition of risk be-
haviors in childhood; and intervention trials (four being listed) on dietary lowering of cholesterol, smoking prevention, and influencing school meal service and physical education. No formal assessment of this evidence was presented, and as in the case of recommendations in the elderly, noted previously, the recommendations were based on expert opinion regarding the available science. Further examples of organizations whose interest focuses on specific population groups in connection with CVD prevention include the American Diabetes Association and National Kidney Foundation, among others. A Note on Risk Scores Defining a high-risk stratum of the population for multifactor intervention, by addressing some equivalent of “total CVD risk,” is fundamental to current clinical guidelines for CVD.2 Clearly, risk stratification of individuals by use of predictive models derived from prospective epidemiologic studies has become a widely accepted convention. But this approach is not without shortcomings, and a number of concerns about cardiovascular risk scoring warrant comment. The most prominent issues are the wide range of predictive performance between models or across populations; limited predictive accuracy specifically for persons at intermediate risk, for women, and for lower socioeconomic strata of a population; resulting misclassification of individuals as being eligible or ineligible for intervention; and the narrow sense of “risk” when only fatal events, only CHD, or only “hard CHD” is the predicted outcome. A brief historical account of efforts to estimate absolute risk of cardiovascular events began with an AHA committee report in 1973.22 Data from the Framingham Heart Study have been used for this purpose to the present, and advances in analytic methods have yielded improved risk estimates. Availability of data from several other cohort studies, often pooled for increased generality and power of estimation, has broadened interest in application of this approach to European and other populations. Organizations developing guidelines increasingly rely on new prediction models. In 2003, a systematic review of risk prediction tools (tables and charts) used by clinicians described 16 of them based on the Framingham model alone.23 In addition, six studies were summarized in which one or more of these tools was compared with use of the full Framingham equation. In 2005, at least six epidemiologic data sets from studies other than the Framingham Heart Study were in use for building risk scores (continuous estimates of risk) and charts
46894_CH20_591_618.qxd
606
2/25/10
4:15 PM
Page 606
CHAPTER 20 RECOMMENDATIONS, GUIDELINES, AND POLICIES
(simplified categorical representations of risk as in Figures 20-1 and 20-2). Giampaoli and colleagues noted that risk predicted in a single population varied widely with use of different models.24 This point was further emphasized by Brindle and others.25 They used recent Framingham models to predict CHD or CVD in 27 study populations representing more than 71,000 participants. Across these 27 studies, predicted-to-observed ratios for combined fatal and nonfatal CHD events ranged from underprediction (relative risk 0.43) to overprediction (relative risk 2.87). Overprediction was greatest in low-risk populations, and underprediction was greatest in high-risk populations. A similar, though less marked, gradient was found for prediction of combined fatal and nonfatal CVD. The suggested consequence of these systematic variations was that in low-risk populations, overestimation of risk would lead to unnecessary treatment for many, whereas the opposite would be true in high-risk populations. These investigators also sought evidence on effectiveness of use of these predictive tools in physician practice but found few relevant trials and no supportive data. Further studies have assessed performance of multiple predictive models in specific populations and concluded, for example, that the risk equations derived from a set of 11 Italian cohorts in the CUORE Study performed better for the Italian population than those from either the Framingham or PROCAM (Münster, Germany) Studies.26 Similarly, neither of these risk functions performed acceptably for the lowrisk populations of Belgium or France.27 Attention has focused on ways to improve risk scoring and its application in diverse populations. D’Agostino and others evaluated applicability of the sex-specific Framingham coronary heart disease prediction scores in multiple ethnic groups—US Blacks, Whites, and Native Americans, and in Hispanic and Japanese American men.28 They found good performance of the scores for Blacks and Whites and, after recalibration for population-specific risk factor distributions and coronary heart disease mortality rates, acceptable applicability for the other groups as well. Ridker and others proposed inclusion of additional risk markers to improve prediction among women, by using data from the Women’s Health Initiative study to derive and test two models.29 The best-fitting model included diabetes determined by HbA1c, systolic blood pressure, current smoking, three lipoprotein measures, and high-sensitivity CRP. They reported improved performance of these models over the original one with reclassification of 40–50% of women into higher- or lower-risk cate-
gories. Thompson and others proposed use of coronary artery calcium scoring, high-sensitivity CRP, heart rate recovery, and exercise tolerance during exercise stress testing to improve stratification of intermediate-risk patients, but without incorporating these measures in the Framingham Risk Score.30 Brindle and others, noted previously, and TunstallPedoe and colleagues argued for taking social deprivation into account in improving the performance of cardiovascular risk scoring.25,31 The case was made that populations, or within-population strata, at highest risk are those with greatest social deprivation; if risk were systematically underestimated, these populations would be undertreated and the disparity in health associated with poverty would be widened, counter to widely advocated health policy. Grundy called attention to the scope of defined outcomes to be predicted, suggesting that restriction to hard CHD is unduly narrow relative to a broader category of atherosclerotic cardiovascular disease (ASCVD) (acute coronary syndromes, coronary artery procedures, coronary deaths, and fatal and nonfatal strokes).32 In his view, social and economic costs of ASCVD, beyond hard CHD and fatal CVD events alone, were increasingly recognized. Therefore, prevention of CVD more generally should be considered in risk prediction. In further work on risk scores, D’Agostino and colleagues developed a general sexspecific cardiovascular risk profile intended for use in primary care to estimate risk of coronary heart disease, cerebrovascular diseases, peripheral vascular disease, and heart failure, both separately and as a combined outcome.33 In Grundy’s view, strata of lower to moderate, moderately high or intermediate, and high risk should be retained, being defined by scores of 10%, 10–20%, and 20% 10-year risk, respectively. This approach was proposed by Hense as well.22 He also urged improved calibration of risk functions to regional or local populations and, in view of the necessary cautions about the current status of risk estimation, suggested that risk communication with patients may be the most effective use of these tools in CVD prevention. (Readers interested in calculating their own risk score may do so by going to http://hp2010.nhlbihin.net/atpiii/evalData.asp.)34 Further, in an earlier report, Grundy and others had drawn an important distinction between longterm and short-term prevention, in the context of global risk assessment.35 Ten-year risk is typically low in younger adults, who are therefore less likely than older people to be considered eligible for drug therapy to manage risk factors. But from a longerterm perspective, documentation of low risk may be important for patient education, much as Hense later
46894_CH20_591_618.qxd
2/25/10
4:15 PM
Page 607
COMMUNITY GUIDELINES 607
suggested.22 Although less attention has been given to this aspect of risk assessment, it has potential to stimulate nonpharmacologic preventive measures that maintain low risk, or reduce risk, throughout adult life. Grundy also called for periodic risk assessment in early adulthood and early middle age as a measure to reduce long-term risk. Implicitly the results could reinforce successful preservation of low risk or identify early adverse changes that might still be reversed with lifestyle changes alone. In summary, risk prediction to support clinical CVD prevention is under active development, with several fundamental questions being addressed in the current research. Substantial progress over the past decade has set the stage for new integrated CVD prevention guidelines anticipated from NHLBI in the near future.
COMMUNITY GUIDELINES The following discussion turns from high-risk, individual-level guidelines to those for communityor population-wide implementation. Diet, physical activity, and tobacco have long been the focus of CVD prevention at the population level. These are the keys to prevention and control of obesity, adverse blood lipid profiles, high blood pressure, and glucose intolerance, insulin resistance, and diabetes. Part III indicates the potential for population-wide measures to reduce incidence and prevalence of specific risk factors and mitigate their contributions to risks of atherosclerotic and hypertensive diseases. Similarly, as noted in Chapter 19, the Guide to Community Preventive Services presents recommendations for intervention on several specific conditions, some of which are directly related to CVD prevention, but it does not currently provide comprehensive recommendations for prevention of CHD or stroke.36 This gap is addressed in a 2003 AHA Scientific Statement, American Heart Association Guide for Improving Cardiovascular Health at the Community Level.37 This report, cited in Chapter 18, underscored the potential link between community-level intervention and recommendations in four previous AHA or AHA/ACC clinical prevention guidelines:37, p 645 This Guide differs from these four clinical guidelines because it provides a comprehensive approach to reducing the burden of cardiovascular diseases (CVD) through improving the local policies and environment as a means to promote cardiovascular health. Changes toward a healthier environment could be expected to en-
hance the clinically oriented guidelines because both the primary and secondary prevention guidelines recommend that healthcare providers encourage behavior change in individual patients. Improvements in facilities and resources in the places where people work and live should enhance the achievement of many goals, including: cessation of tobacco use and avoidance of environmental tobacco smoke; reduction in dietary saturated fat, cholesterol, sodium, and calories; increased plant-based food intake; increased physical activity; access to preventive healthcare services; and early recognition of symptoms of heart attack and stroke. Healthcare providers and their patients have better opportunities for successfully implementing the clinical guidelines when they live in such communities. The Guide was developed on the basis of expert opinion, an extensive body of scientific literature, and an evolving policy framework supporting community intervention (see Policy, as follows). Its recommendations were presented under six broad strategies: Assessment, Education, Community Organization and Partnering, Assuring Personal Health Services, Environmental Change, and Policy Change. Education was further categorized as General Health Education, School and Youth Education, Work Site Education, and Healthcare Facility Education. These six strategies, 19 goals, and 59 recommendations for comprehensive community intervention are of sufficient interest to warrant reproducing them in full (Table 20-6). Recognizable in this Guide are the core functions public health—assessment, policy development, and assurance—as well as the fundamental public health goal “to prevent the onset of risk factors in the first place, referred to as ‘primordial prevention’ or health promotion.” 37, p 645 How to implement this Guide was addressed in a subsequent AHA Scientific Statement, Taking the Initiative, whose purpose was to enable communities and local leaders to adopt the Guide, setting priorities on those recommendations that are most promising and amenable to action under local circumstances.38 This process would begin with community mobilization and would progress through assessment of local needs and resources and community-based planning to the stage of widespread and sustained implementation. Evaluation of the process and impact would then lead to further, continuing community efforts for improved cardiovascular health. Materials were identified and described regarding related guidelines, data sources, and tools for health promotion initiatives, partnership devel-
46894_CH20_591_618.qxd
608
2/25/10
4:15 PM
Page 608
CHAPTER 20 RECOMMENDATIONS, GUIDELINES, AND POLICIES
Table 20-6
Guide to Improving Cardiovascular Health at the Community Level Strategies and Goals Recommendations Assessment Goal: All persons and communities should know that CVD and stroke are the leading causes of death and disability in men and women.
• Determine and make available data on the burden of CVD and stroke mortality at the local level (city or county). • Identify groups defined by sex, race/ethnicity, socioeconomic status, or geographic location that are at especially high risk of CVD and stroke within each community. • Assess the levels of major preventable causes of CVD and stroke in the community, including lifestyle behaviors (e.g., adverse nutrition, cigarette smoking, sedentary lifestyle) and risk factors (hypertension, atrial fibrillation, diabetes, elevated blood cholesterol, and obesity).
Education General Health Education Goal: All communities should provide information to their members about the burden, causes, and early symptoms of CVD and stroke.
• Mass media (television, radio, newspapers) should disseminate results of surveillance about the burden of CVD and stroke in the community. • Mass media and local media (e.g., pamphlets, brochures) should emphasize the importance of lifestyle behaviors and risk factors on cardiovascular health. • Public education campaigns should make the community aware of guidelines for primary and secondary prevention of CVD and stroke. • Mass and local media should emphasize the early warning signs of myocardial infarction and stroke. • Ongoing education programs should provide training of lay members in cardiopulmonary resuscitation. • All citizens should know how to access the emergency medical care system.
Goal: Communities should provide materials and programs to motivate and teach skills for changing risk behaviors that will target multiple population subgroups.
• A guide to community resources (services and programs) for prevention, diagnosis, and treatment of CVD and stroke should be available. • Communities should support and publicize research-based programs for CVD risk reduction that are targeted to key population subgroups, especially disadvantaged groups and people at all levels of readiness to change. • Communities should promote the use of web site programs for risk reduction by making web site access to such programs available in public libraries and schools. • Food advertising directed to youth should be limited to foods that meet health guidelines. • TV shows for children should promote physical activity during commercial breaks.
School and Youth Education Goal: All schools should have researchbased, comprehensive, and ageappropriate curricula about cardiovascular health and ways to improve health behaviors and reduce CVD risk.
• School curricula should include lessons about risk factors for CVD and stroke and the extent of heart disease and stroke in the community. • Research-based curricula about effective methods of changing health behaviors should be implemented. • Students should learn skills needed to achieve regular practice of healthful behaviors, and parents should learn how to support their children’s healthful behaviors. • Specific curricular materials for healthy nutrition and physical activity should be offered. • Physical education should be required at least three times a week in grades K–12, with an increasing emphasis on lifetime sports/activities. Implementation of research-based curricula is recommended. • Meals provided at schools should include alternatives conducive to cardiovascular health.
Goal: All schools should implement ageappropriate curricula on changing dietary, physical activity, and smoking behaviors.
Goal: All schools should provide teaching of early warning signs of myocardial infarction and stroke and appropriate initial steps of emergency care.
• Students should know how to activate the emergency medical system. • Cardiopulmonary resuscitation instruction should be provided to students at appropriate ages
continues
46894_CH20_591_618.qxd
2/25/10
4:15 PM
Page 609
PUBLIC POLICIES 609
Table 20-6
Guide to Improving Cardiovascular Health at the Community Level—continued Strategies and Goals Recommendations Work Site Education Goal: All work sites should provide materials and services to motivate and assist employees to adopt and maintain heart-healthy behaviors. Goal: All work sites should provide instruction in early warning signs of myocardial infarction and stroke and appropriate initial steps of emergency care.
• Work sites should promote increased physical activity in the day’s work (e.g., stair climbing). • Workers should have access to research-based effective materials and services to help them adopt and maintain heart-healthy behaviors. • Workers should know how to activate the emergency medical system. • Cardiopulmonary resuscitation instruction should be available to all workers.
Healthcare Facility Education Goal: All healthcare facilities should make available research-based, effective educational materials and programs about changing and maintaining risk factors/risk behaviors, ways to prevent CVD and stroke, and early warning signs of CVD and stroke.
• Print and other media should be available in healthcare facilities to describe CVD and stroke risk factors and their early warning signs. • Guides for primary and secondary prevention should be made available for all patients. • Educational materials should be modified to accommodate for limited literacy, cultural and language diversity, sex differences, and dissemination flexibility.
Source: Reprinted with permission from TA Pearson, TL Bazzarre, SR Daniels, et al., Circulation, Vol 107, © 2003 American Heart Association Inc., pp 648–649
opment and operation, evaluation, and education. Community-level indicators were listed for monitoring underlying social and environmental conditions, risk behaviors and biological risk factors, and CVD outcomes. Another sequel to the Guide illustrates targeting of intervention specifically to the school setting. Hayman and others, in Cardiovascular Health Promotion in the Schools, updated the guidelines for children and adolescents cited previously20,21 to place recommendations for school-based intervention in the context of community-level CVD prevention.39 Together, these several reports exemplify the community component of CVD prevention and highlight the importance of this public health component of preventive efforts, its complementary relation to clinical preventive guidelines, and the distinct character of community-wide intervention in contrast to physician- and patient-oriented recommendations.
PUBLIC POLICIES Background Policy interventions were discussed among other strategies of prevention in Chapter 18. Here, selected examples will illustrate policy for CVD prevention in the United States and its development to date on a
worldwide scale, especially as it relates to developing countries. The 1988 report, The Future of Public Health, identified three core functions of public health: assessment, policy development, and assurance. Policy development was characterized in this way:40, p 8 The Committee recommends that every public health agency exercise its responsibility to serve the public interest in the development of comprehensive public health policies by promoting use of the scientific knowledge base in decisionmaking about public health and by leading in developing public health policy. Agencies must take a strategic approach, developed on the basis of a positive appreciation for the democratic political process. Public policy in its fullest form influences whole populations or societies. However, multiple levels of society may be considered as potential beneficiaries of policy change. In their conceptualization of determinants of health, Dahlgren and Whitehead constructed the original version of Figure 18-3, now commonly referred to as the “socioecological model” of health.41 In parallel with the model, four levels of policy intervention were ordered from the outermost to the innermost regions of the model:41, p 12 Policy Level 1––aimed at structural changes
46894_CH20_591_618.qxd
610
2/25/10
4:15 PM
Page 610
CHAPTER 20 RECOMMENDATIONS, GUIDELINES, AND POLICIES
Policy Level 2––aimed at improving living and working conditions through healthy public or business strategies Policy Level 3––aimed at strengthening social and community support Policy Level 4––aimed at influencing individual lifestyles and attitudes They noted that for any particular health policy goal, strategies could be devised at any of these four levels and emphasized the potential synergistic effect of multilevel policy interventions. Another kind of distinction has been drawn between two types of public health interventions, each of which would be included in Dahlgren’s and Whitehead’s four-level concept of policy:42, p 15 Policies, which include laws, regulations, and rules (both formal and informal). Environmental interventions, which include changes to the economic, social, or physical environments. Brownson and others concluded in a 2006 review that “a considerable body of evidence” shows effectiveness of environmental and policy interventions in risk-factor prevention (especially reduction in tobacco use). They admonished practitioners to “consider the power of environmental and policy approaches to set the stage for other interventions” before implementing individual-level chronic disease prevention programs.43, p 363 A further aspect of “policy instruments,” as termed by Gaziano and others and noted in Chapter 18, is their relative cost-effectiveness as a criterion of acceptability.44 The point was also made in Chapter 18 that the agent of policy interventions, although typically a governmental health agency, might in some instances—described as “whole-ofgovernment” interventions—involve a broad spectrum of interests well beyond the health sector alone.45 Taking these qualities together, ideal public policies for CVD prevention would be expected to have broad reach, multilevel components, multisector engagement, evidence of cost-effectiveness, and a facilitating effect in support of family or individuallevel interventions. The United States The Inter-Society Commission for Heart Disease Resources Early policy development for CVD prevention as a national-level governmental activity in the United States is illustrated by the reports of the Inter-Society Commission for Heart Disease Resources, published
in the 1970s.46 The Commission was created to fulfill congressional intent in establishing the Regional Medical Programs Service in 1965. It was charged to develop guidelines regarding “optimum medical resources” for prevention and treatment of cardiovascular diseases. With 29 participating organizations—chiefly representing national professional associations—and four federal agencies as advisory organizations, the Commission formed seven categorical and seven general study groups. Primary Prevention of the Atherosclerotic Diseases was a joint report of the Atherosclerosis Study Group (Jeremiah Stamler, Chair) and the Epidemiology Study Group (Abraham Lilienfeld, Chair). (The report on hypertension is noted in Chapter 12.) The report pointed to the need for a national commitment to primary prevention of CHD and other atherosclerotic diseases, outlined the rationale for prevention, and estimated the potential impact of primary prevention of CHD in terms of cases and deaths prevented among white males aged 35–64 years. The Commission called for “immediate and concurrent implementation” of its recommendations:47, p 44 The Commission recommends that a strategy of primary prevention of premature atherosclerotic diseases be adopted as long-term national policy for the United States and to implement this strategy that adequate resources of money and manpower be committed to accomplish: Changes in diet to prevent or control hyperlipidemia, obesity, hypertension and diabetes Elimination of cigarette smoking Pharmacologic control of elevated blood pressure. Further, among detailed recommendations within each of these three target areas, community prevention programs were proposed:47, p 50 “The Commission recommends that community programs be developed and expanded for the detection and treatment of persons of all ages who are very susceptible to premature atherosclerotic diseases due to combinations of the major risk factors.” Finally, the Commission addressed as a matter of national policy the relation between public health action and continuing research to resolve still unanswered questions:47, p 44 The Commission recommends that a Special Committee be established at a high level of the Federal Government to develop coordinated plans for large-scale, long-term trials to determine the effect of various interventions, particularly diet modification, on the rates of premature atherosclerotic diseases in the United
46894_CH20_591_618.qxd
2/25/10
4:15 PM
Page 611
PUBLIC POLICIES 611
States. . . . The public health importance of CHD makes it mandatory to conduct such trials. . . . At times urgent public health decisions must be made on the soundest evaluation and best judgment of available incomplete evidence. The report thus contemplated immediate public health measures to address diet, smoking, and blood pressure, concurrent with a program of major trials to strengthen evidence supporting primary prevention of the atherosclerotic diseases. How directly this recommendation contributed to the subsequent large-scale trials in the United States is difficult to judge, but the conceptual foundation for prevention policy was clearly evident from the date of this report. Healthy People 2010 Following shortly on the work of the Inter-Society Commission was the federal-level Healthy People initiative, described in Chapter 18. Reference to reducing death rates from heart attacks and strokes as a subgoal and high blood pressure control as a target for preventive health services was present in the 1979 Surgeon General’s Report.48 However, it was not until release of Healthy People 2010 early in the year 2000 that heart disease and stroke became a distinct focus area.49 Focus Area 12 in Healthy People 2010 can be seen as a policy statement for CVD prevention. It directs the two co-lead agencies of the Department of Health and Human Services, the Centers for Disease Control and Prevention (CDC) and the National Institutes of Health (NIH), to lead the nation’s efforts toward achievement of the four-part goal, discussed in Chapter 18: prevention of risk factors, detection and treatment of risk factors, early identification and treatment of heart attacks and strokes, and prevention of recurrent cardiovascular events. One policy implication for heart disease and stroke prevention is the implied requirement for intervention in each of these goal areas that span the full range of approaches indicated in Figure 18-12. A second policy implication is reflected in the objectives and targets for improving cardiovascular health. Sixteen measurable objectives are included that address blood pressure, cholesterol, heart disease, and stroke. Forty-six related objectives are presented under other behavioral or disease conditions within Healthy People 2010. Baseline data for each objective are taken from the most recent sources as of 2000, and targets are set usually as a percentage change from baseline by 2010. For several objectives, baseline data are given specifically for one or more population subgroups, variously by race and ethnicity, gender, education level, family in-
come level, disability status, presence or absence of diabetes, or urban or rural residence. Sometimes marked disparities in disease burden are evident in the baseline data, but the target levels are generally set at the same level for all groups. This reflects the overarching goal to eliminate disparities in health, a major policy provision of Healthy People 2010. It also implies especially intensified intervention for groups at a relative health disadvantage at baseline. An example is the target to bring African Americans and non-Hispanic Whites to the same greatly reduced prevalence of high blood pressure, in the face of a wide disparity in this condition. The National Forum for Heart Disease and Stroke Prevention Following its mandate under Healthy People 2010, the CDC brought a number of partner organizations and individual leaders together to develop a long-range strategic plan. The result was A Public Health Action Plan to Prevent Heart Disease and Stroke, released in 2003 and updated in 2008.50,51 The Plan itself comprised 24 recommendations and nearly 70 proposed action steps in seven thematic areas of communication, public health leadership, policy implementation, capacity development, monitoring and evaluation, prevention research, and regional and global collaboration. A new national organization was established to put the Plan to work, the National Forum for Heart Disease and Stroke Prevention. For each thematic area, an Implementation Group was established— including the Action Priorities group for policy implementation and the Policy Research group for fostering research on scientific bases for policy development and implementation. In accordance with their respective missions, these groups are expected to advance development and translation into action of public health policies for heart disease and stroke prevention:51, pp 50,64 Action Priorities Implementation Group: To identify effective policies in cardiovascular health promotion and cardiovascular disease prevention at the national, state, and local levels to ensure effective public health action against heart disease and stroke. Policy Research Implementation Group: To develop a comprehensive policy research agenda, foster translating this research into practice, and investigate relevant economic models. A third group, on Regional and Global Collaboration, searched for an authoritative statement of US policy regarding global dimensions of CVD prevention. Finding none, the group proposed and the National Forum adopted a Policy Framework Statement for Regional and Global Partnerships.52 The
46894_CH20_591_618.qxd
612
2/25/10
4:15 PM
Page 612
CHAPTER 20 RECOMMENDATIONS, GUIDELINES, AND POLICIES
statement is consonant with CDC’s Global Health Promotion Goal and Global Health Diplomacy Goal,53 and it proposes as core values that “Health is a fundamental right of all the world’s people” and that “Everyone should have a fair and equitable opportunity to attain his/her full health potential.” Guiding principles address solidarity, community participation, accountability, sustainability, advocacy, and promotion of programs and actions meeting cardiovascular health needs. Implementation strategies support and promote tobacco control, healthier foods, everyday physical activity, and essential drugs, lifestyle counseling, and basic services, as needed—both to control risk factors prior to CVD onset and during acute CVD events and to prevent recurrence or progression of CVD through long-term management. State Policies In the United States, legal authority for public health resides primarily in the states. Accordingly, policies bearing on cardiovascular health at the state level are of particular interest. However, taking an inventory of health policies for 50 states and the District of Columbia in a comprehensive way is a complex task. The National Council of State Legislatures, for example, maintains a database of state laws, but health is only one of 20 topic areas, public health is one of 10 health subtopics, and only laws related to heart attacks, cardiac arrest, and defibrillators are currently included [http://www.ncsl.org/programs/health/aed .htm, accessed April 18, 2008].54 To address this interest, the Division for Heart Disease and Stroke Prevention at CDC hosts the HDSP Policy Project interactive Web site that lists 207 state-level policies that were in force at any time between 1978 (earliest date searched) and 2005.55 Users can search and study these policies, assisted by an annotated bibliography and spreadsheet display of policy details (state, bill number, status, year passed, topic area, policy abstract, and pdf file name of the policy). In addition, the site offers a mapping function that permits display of states with specified policies, a guide to policy making, and a handbook on policy assessment. These materials are available at www .cdc.gov/dhdsp/dhdspleg56 and at the Division Web site, www.cdc.gov/dhdsp.57 Worldwide/Developing Countries The World Health Organization (WHO) in 1955 convened a Study Group on Atherosclerosis and Ischaemic Heart Disease “to discuss the present status of knowledge of the etiology and pathogenesis of these diseases and to advise the Organization on
means of furthering knowledge which might lead to effective preventive programmes.”58, p 3 The focus of the Study Group’s report was research needs throughout the world. Annex 2 to the report commented on the state of knowledge regarding prevention: “Various preventive measures for the avoidance of ischaemic heart disease have been suggested but, as the Group’s report points out, adequate evidence does not yet exist to give scientific support to the premise that application of specific measures will prevent or delay the onset of manifest ischaemic heart disease.”58, p 37 No global policy for prevention was yet available. Stamler traced policy development for CVD prevention through the 1960s and 1970s.59 He cited several prominent landmarks in the United States indicating participation by multiple government agencies and nongovernmental organizations. Those reports addressed prevention of CVD either very broadly or in terms specific to nutrition or smoking. Additional policy reports were published in this period by Scandinavian countries, Germany, Australia, New Zealand, the United Kingdom, WHO, and the International Society and Federation of Cardiology. Clearly, wider recognition of the problem and the potential for prevention were emerging. In the 1980s, three prominent reports from WHO Expert Committees addressed prevention of CHD. The 29th World Health Assembly in 1976 had called for preparation of a long-term WHO program to promote research on CVD prevention and coordinate international cooperation in the cardiovascular area. WHO Member States were urged to implement prevention programs. That action led to convening the first of these three Expert Committees in late 1981. The resulting report, Prevention of Coronary Heart Disease, appeared in 1982.60 It adopted the dual population-wide and high-risk strategies and incorporated primordial prevention as the strategic foundation for developing countries. The report suggested that primordial prevention should be closely linked with primary healthcare resources, whose strengthening was already being advocated by WHO. Community Prevention and Control of Cardiovascular Diseases followed in 1986.61 This Expert Committee Report embraced the recommendations of the 1982 report, addressed means of implementation at the community level, and presented a model regional plan for CVD prevention. Prevention in Childhood and Youth of Adult Cardiovascular Diseases: Time for Action resulted from an Expert Committee meeting in 1988 and appeared in 1990.62 The focus was on early intervention to avert both the risk factors themselves and
46894_CH20_591_618.qxd
2/25/10
4:15 PM
Page 613
CURRENT ISSUES 613
later complications of advanced atherosclerosis and hypertension, as well as other major cardiovascular conditions:62, p 83 It is therefore recommended that countries develop and pursue a comprehensive population strategy for the primary prevention of these diseases as part of their long-term national health development plan. This strategy should emphasize primary prevention beginning in early childhood and youth, in order to avoid the emergence of the established major risk factors for adult CVD and prevent their persistence on a mass scale in the community. Through improved nutrition (including reduced salt intake), increased physical activity, and elimination of smoking, risk factors including overweight and obesity would be prevented. Health promotion activities in schools and community organizations were seen as fundamental to this effort. These examples illustrate the seminal work of the WHO CVD Unit throughout its existence from 1960 until the mid-1990s. Subsequently, WHO has addressed CVD prevention mainly in conjunction with other chronic, or noncommunicable, diseases, and often with special reference to developing countries.
WHO Global Strategy on Diet, Physical Activity and Health Reflecting this broader context, the 2004 WHO report, Global Strategy on Diet, Physical Activity and Health discussed the global burden of these conditions and two main risk factors, diet and physical activity.63 It presented the challenge and opportunity for preventing these conditions, provided explicit goals and objectives, and outlined evidence, principles, and responsibilities for action. It described the roles of Member States, WHO, international partners, civil society and nongovernmental organizations, and the private sector. In particular, Member States were asked to develop national strategies, policies, and action plans; address national food and agricultural policies; engage multisectoral partners and especially schools; consult with policy stakeholders; link with health services; and invest in surveillance, research, and evaluation. Policy for CVD prevention now concerns a broad spectrum of issues, engages multiple sectors of society, and has potential impact on a wide range of health outcomes. (The landmark global treaty on tobacco, the Framework Convention on Tobacco Control, is discussed in Chapter 14, “Smoking and Other Tobacco Use.”)
Disease Control Priorities in Developing Countries Finally, in the global perspective, the World Bank report Disease Control Priorities in Developing Countries first appeared in 1993.64 Efficacy and cost of both community-based and clinic-based preventive strategies for CVD were reviewed. The second edition of this report similarly addressed both “population” and “personal” interventions.65 In the former category were policy-level interventions that included legislation to substitute trans-fat with polyunsaturated fat, reduce salt content of the diet, increase tobacco price, and implement nonprice interventions against tobacco use. Extensive detail on the rationale and projected cost and impact of these interventions was provided in discussion of priority setting for health interventions in low- and middleincome countries.
CURRENT ISSUES Current issues in this area concern, first, the extent to which existing recommendations, guidelines, and policies are actually implemented and their impact is evaluated and, second, the need for further research to identify and overcome barriers to advances in policy development and implementation. Effective action in both respects is required to assure that the intended benefits of CVD prevention are realized for the whole population, including individuals at high risk. The implementation and impact of clinical guidelines have been assessed in numerous studies, among them three sequential surveys of patients in preventive cardiology practices in Europe.66,67 The first survey, in 1995–1996, included more than 3500 patients in 9 countries; the second, in 1999–2000, included nearly 4000 patients in 15 countries. Need for improved lifestyle factors and use of recommended treatments were documented at the baseline survey and comparisons were made on the basis of the second one. Smoking and obesity had both increased in the interval; prevalence of high blood pressure showed no improvement, but total cholesterol concentrations were reduced and several medications were increased in use. The third survey was launched in 2006, but results were still awaited.67 The principal findings of this and similar reports are that clinical recommendations and guidelines are insufficient in themselves to change practice. Increasing adoption of performance measures to guide monitoring, evaluation, and in some circumstances reimbursement for clinical services may accelerate this
46894_CH20_591_618.qxd
614
2/25/10
4:15 PM
Page 614
CHAPTER 20 RECOMMENDATIONS, GUIDELINES, AND POLICIES
process. Although population-wide policies for CVD prevention have less often been evaluated for their adoption and impact, a widely held view is that they, too, are seldom implemented on a scale and with sustained support to achieve the intended benefit to the population. The following chapters address what has been learned from experience with intervention, what are currently proposed strategies for more effective action, and what broad research agenda can guide further development, implementation, and evaluation of recommendations, guidelines, and policies for CVD prevention. REFERENCES 1. White PD, Wright IS, Sprague HB, et al. A Statement on Arteriosclerosis: Main Cause of “Heart Attacks” and “Strokes.” New York: National Health Education Committee, Inc; 1959. 2. Smith SC Jr, Jackson R, Pearson TA, et al. Principles for national and regional guidelines on cardiovascular disease prevention. A scientific statement from the World Heart and Stroke Forum. Circulation. 2004;109:3 3112–3121. 3. US Department of Health and Human Services, National Institutes of Health, National Heart, Lung and Blood Institute. Current clinical practice guidelines and reports. Available at: http://www.nhlbi.nih.gov/guidelines/current/ htm. Accessed April 8, 2008. 4. American Heart Association. Prevention. Available at: http://www.americanheart.org/ presenter.jhtml?identifier=3004572. Accessed April 8, 2008. 5. US Department of Health and Human Services. Agency for Healthcare Research and Quality. US Preventive Services Task Force. The Guide to Clinical Preventive Services 2006. Recommendations of the US Preventive Services Task Force. AHRQ Publication No. 06-0588. Washington, DC: Agency for Healthcare Research and Quality; 2006. Available at: http://www .ahrq.gov/clinic/uspstf/uspstbac.htm. Accessed October 14, 2007. 6. US Department of Health and Human Services, National Institutes of Health, National Heart,
Lung and Blood Institute. Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). Final Report. NIH Publication No. 02-5215; September 2002. 7. US Department of Health and Human Services, National Institutes of Health, National Heart, Lung and Blood Institute. Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). Available at: http://www.nhlbi.nih.gov/guidelines/ cholesterol. Accessed May 12, 2008. 8. Lichtenstein AH, Appel LJ, Brands M, et al. Diet and lifestyle recommendations revision 2006. A scientific statement from the American Heart Association Nutrition Committee. Circulation. 2006;114:82–96. 9. Pearson TA, Blair SN, Daniels SR, et al. AHA guidelines for primary prevention of cardiovascular disease and stroke: 2002 update: consensus panel guide to comprehensive risk reduction for adult patients without coronary or other atherosclerotic vascular diseases. Circulation. 2002;106:388–391. 10. Goldstein LB, Adams R, Alberts MJ, et al. Primary prevention of ischemic stroke. A guideline from the American Heart Association/American Stroke Association Stroke Council: Cosponsored by the Atherosclerotic Peripheral Vascular Disease Interdisciplinary Working Group; Cardiovascular Nursing Council; Clinical Cardiology Council; Nutrition, Physical Activity, and Metabolism Council; and the Quality of Care and Outcomes Research Interdisciplinary Working Group. Circulation. 2006;113: e873–e923. 11. Moser DK, Kimble LP, Alberts MJ, et al. Reducing delay in seeking treatment by patients with acute coronary syndrome and stroke. A scientific statement from the American Heart Association Council on Cardiovascular Nursing and Stroke Council. Circulation. 2006;114:168–182.
46894_CH20_591_618.qxd
2/25/10
4:15 PM
Page 615
REFERENCES 615
12. Leon AS, Franklin BA, Costa F, et al. Cardiac rehabilitation and secondary prevention of coronary heart disease. An American Heart Association scientific statement from the Council on Clinical Cardiology (Subcommittee on Exercise, Cardiac Rehabilitation, and Prevention) and the Council on Nutrition, Physical Activity, and Metabolism (Subcommittee on Physical Activity), in Collaboration with the American Association of Cardiovascular and Pulmonary Rehabilitation. Circulation. 2005;111:369–376. 13. Smith SC Jr, Allen J, Blair SN, et al. AHA/ACC guidelines for secondary prevention for patients with coronary and other atherosclerotic vascular disease: 2006 update. Circulation. 2006;113:2363–2372.
19. Williams MA, Fleg JL, Ades PA, et al. Secondary prevention of coronary heart disease in the elderly (with emphasis on patients 75 years of age). An American Heart Association scientific statement from the Council on Clinical Cardiology Subcommittee on Exercise, Cardiac Rehabilitation, and Prevention. Circulation. 2002;105:1735–1743. 20. Strong WB, Deckelbaum RJ, Gidding SS, et al. Special report: integrated cardiovascular health promotion in childhood: a statement for health professionals from the Subcommittee on Atherosclerosis and Hypertension in Childhood of the Council on Cardiovascular Disease in the Young, American Heart Association. Circulation. 1992;85:1638–1644.
14. European Society of Cardiology. Welcome to the ESC Guidelines Section. Available at: http:// www.escardio.org/knowledge/guidelines/ Guidelines_list.htm?hit=quick. Accessed January 13, 2008.
21. Kavey R-EW, Daniels SR, Lauer RM, Atkins DL, Hayman LL, Taubert K. American Heart Association guidelines for primary prevention of atherosclerotic cardiovascular disease beginning in childhood. Circulation. 2003;107: 1562–1566.
15. Graham IM. Guidelines on cardiovascular disease prevention in clinical practice: the European perspective. Curr Opin Cardiol. 2005;20:430–439.
22. Hense H-W. Observations, predictions and decision––assessing cardiovascular risk assessment. Int J Epidemiology. 2004;33:235–239.
16. World Health Organization. Prevention of Cardiovascular Disease. Guidelines for Assessment and Management of Cardiovascular Risk. Geneva: World Health Organization; 2007. 17. Balady GJ, Williams MA, Ades PA, et al. Core components of cardiac rehabilitation/secondary prevention programs: 2007 update. A scientific statement from the American Heart Association Exercise, Cardiac Rehabilitation, and Prevention Committee, the Council on Clinical Cardiology; the Councils on Cardiovascular Nursing, Epidemiology and Prevention, and Nutrition, Physical Activity, and Metabolism; and the American Association of Cardiovascular and Pulmonary Rehabilitation. Circulation. 2007;115:2657–2682. 18. Mosca L, Banka CL, Benjamin EJ, et al. Evidence-based guidelines for cardiovascular disease prevention in women: 2007 update. Circulation. 2007;115:1481–1501.
23. Sheridan S, Pgnone M, Mulrow C. Framinghambased tools to calculate the global risk of coronary heart disease. J General Internal Medicine. 2003;18:1039–1052. 24. Giampaoli S, Palmieri L, Mattiello A, Panicao S. Definition of high risk individuals to optimize strategies for primary prevention of cardiovascular diseases. Nutrition, Metabolism & Cardiovascular Diseases. 2005;15:79–85. 25. Brindle P, Beswick A, Fahey T, Ebrahim S. Accuracy and impact of risk assessment in the primary prevention of cardiovascular disease: a systematic review. Heart. 2006;92: 1752–1759. 26. Ferrario M, Chiodini P, Chambless LE, et al. Prediction of coronary events in a low incidence population. Assessing accuracy of the CUORE Cohort Study prediction equation. Int J Epidemiology. 2005;34:413–421. 27. Empana JP, Ducimetière P, Arveiler D, et al. Are the Framingham and PROCAM coronary
46894_CH20_591_618.qxd
616
2/25/10
4:15 PM
Page 616
CHAPTER 20 RECOMMENDATIONS, GUIDELINES, AND POLICIES
heart disease risk functions applicable to different European populations? European Heart J. 2003;24:1903–1911. 28. D’Agostino Sr RB, Grundy S, Sullivan LM, et al. Validation of the Framingham coronary heart disease prediction scores. Results of a multiple ethnic groups investigation. J American Med Assoc. 2001;286:180–187. 29. Ridker PM, Buring JE, Rifai N, Cook NR. Development and validation of improved algorithms for the assessment of global cardiovascular risk in women. The Reynolds Risk Score. J American Med Assoc. 2007;297:611–619. 30. Thompson JB, Rivera JJ, Blumenthal RS, Danyi P. Primary prevention or patients with intermediate Framingham scores. Current Cardiol Reports. 2006;8:261–266. 31. Tunstall-Pedoe H, Woodward M, for the SIGN Group on Risk Estimation. By neglecting deprivation, cardiovascular risk scoring will exacerbate social gradients in disease. Heart. 2006;92:307–310. 32. Grundy SM. The changing face of cardiovascular risk. J American College of Cardiol. 2005; 46:173–175. 33. D’Agostino Sr RB, Vasan RS, Pencina MJ, et al. General cardiovascular risk profile for use in primary care. The Framingham Heart Study. Circulation. 2008;117:743–753. 34. US Department of Health and Human Services, National Institutes of Health, National Heart, Lung and Blood Institute. Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). 10-year risk calculator results. Available at: http:// hp2010.nhlbihin.net/atpiii/evalData.asp. Accessed July 6, 2007. 35. Grundy SM, Pasternak R, Greenland P, Smith S Jr, Fuster V. Asessment of cardiovascular risk by use of multiple-risk-factor assessment equations. A statement for healthcare professionals from the American Heart Association and the American College of Cardiology. Circulation. 1999;100:1481–1492.
36. Guide to Community Preventive Services. The Community Guide. Available at: http://www .thecommunityguide.org/index.html. Accessed September 25, 2007. 37. Pearson TA, Bazzarre TL, Daniels SR, et al. American Heart Association guide for improving cardiovascular health at the community level: a statement for public health practitioners, healthcare providers, and health policy makers from the American Heart Association Expert Panel on Population and Prevention Science. Circulation. 2003;107:645–651. 38. Veazie MA, Galloway JM, Matson-Koffman D, et al. Taking the initiative. Implementing the American Heart Association guide for improving cardiovascular health at the community level. Circulation. 205;112:2538–2554. 39. Hayman LL, Williams CL, Daniels SR, et al. Cardiovascular health promotion in the schools. A statement for health and education professionals and child health advocates from the Committee on Atherosclerosis, Hypertension, and Obesity in Youth (AHOY) of the Council on Cardiovascular Disease in the Young, American Heart Association. Circulation. 2004;110:2266–2275. 40. Committee for the Study of the Future of Public Health. The Future of Public Health. Washington, DC: Division of Health Care Services, Institute of Medicine. National Academy Press; 1988. 41. Dahlgren G, Whitehead M. Policies and Strategies to Promote Social Equity in Health. Stockholm: Institute for the Futures Studies; 1991. 42. Association of State and Territorial Directors of Health Promotion and Public Health Education, Centers for Disease Control and Prevention. Policy and Environmental Change. New Directions in Public Health. Final Report. Atlanta, GA: United States Department of Health and Human Services, Centers for Disease Control and Prevention; 2001. 43. Brownson RC, Haire-Joshu D, Luke DA. Shaping the context of health: a review of environmental and policy approaches in the prevention of chronic diseases. Ann Rev Public Health. 2006;27:341–370.
46894_CH20_591_618.qxd
2/25/10
4:15 PM
Page 617
REFERENCES 617
44. Jamison DT, Breman JG, Measham AR et al., eds. Disease Control Priorities in Developing Countries. 2nd ed. Washington, DC: International Bank for Reconstruction and Development/The World Bank; 2006.
52. National Forum for Heart Disease and Stroke Prevention. Policy Framework. Statement for Regional and Global Partnerships. Available at: http://www.nationalforum.org. Accessed May 8, 2008.
45. Leeder S, Raymond S, Greenberg H. A Race Against Time: The Challenge of Cardiovascular Disease in Developing Countries. New York: The Trustees of Columbia University in the City of New York; 2004.
53. US Department of Health and Human Services, Centers for Disease Control and Prevention. Healthy People in a Healthy World. A Snapshot of CDC Efforts. http://intranet.cdc.gov/ cogh/goals/pdf. Accessed May 13, 2008.
46. Wright IS, Frederickson DT, eds. Cardiovascular Diseases. Guidelines for Prevention and Care. Reports of the Inter-Society Commission. Washington, DC: US Government Printing Office; 1973.
54. National Conference of State Legislatures. Health Program. State Laws on Heart Attacks, Cardiac Arrests, & Defibrillators. Encouraging or requiring community access and use. http:// www.ncsl.org/programs/health/aed.htm. Accessed April 18, 2008.
47. Stamler J, Lilienfeld AM, Chairmen. Primary prevention of the atherosclerotic diseases. In: Wright IS, Frederickson DT, eds. Cardiovascular Diseases. Guidelines for Prevention and Care. Reports of the Inter-Society Commission. Washington, DC: US Government Printing Office; 1973:11–58. 48. US Department of Health, Education, and Welfare. Healthy People. The Surgeon General’s Report on Health Promotion and Disease Prevention. Public Health Service, Office of the Assistant Secretary for Health and Surgeon General. DHEW (PHS) Publication No. 79-55071. Washington, DC: US Government Printing Office; 1979. 49. US Department of Health and Human Services. Healthy People 2010. 2nd ed. With Understanding and Improving Health and Objectives for Improving Health. 2 vols. Washington, DC: US Government Printing Office; 2000. 50. US Department of Health and Human Services. A Public Health Action Plan to Prevent Heart Disease and Stroke. Atlanta, GA: US Department of Health and Human Services, Centers for Disease Control and Prevention; 2003. 51. US Department of Health and Human Services. Update to a Public Health Action Plan to Prevent Heart Disease and Stroke. Atlanta, GA: US Department of Health and Human Services, Centers for Disease Control and Prevention; 2008.
55. Ford Lattimore B, O’Neil S, Besculides M. Tools for developing, implementing, and evaluating state policy. Prev Chron Dis. 2008;5(2). http://www.cdc.gov/pcd/issues/2008/apr/07_02 10.htm. Accessed 13 May 2008. 56. US Department of Health and Human Services, Centers for Disease Control and Prevention. Division for Heart Disease and Stroke Prevention. Heart Disease and Stroke Prevention Legislative Database. Available at: www.cdc .gov/dhdsp/dhdspleg. Accessed April 18, 2008. 57. US Department of Health and Human Services, Centers for Disease Control and Prevention. Division for Heart Disease and Stroke Prevention. Available at: http://www.cdc.gov/ dhdsp. Accessed April 18, 2008. 58. World Health Organization. Study Group on Atherosclerosis and Ischaemic Heart Disease. World Health Organization Technical Report Series No. 117. Geneva: World Health Organization; 1957. 59. Stamler J. Primary prevention of coronary heart disease: the last 20 years. Am J Cardiol. 1981;47:722–735. 60. World Health Organization Expert Committee. Prevention of Coronary Heart Disease. Technical Report Series 679. Geneva: World Health Organization; 1982.
46894_CH20_591_618.qxd
618
2/25/10
4:15 PM
Page 618
CHAPTER 20 RECOMMENDATIONS, GUIDELINES, AND POLICIES
61. World Health Organization Expert Committee. Community Prevention and Control of Cardiovascular Diseases. WHO Technical Report Series 732. Geneva: World Health Organization; 1986. 62. World Health Organization Expert Committee. Prevention in Childhood and Youth of Adult Cardiovascular Diseases: Time for Action. WHO Technical Report Series 792. Geneva: World Health Organization; 1990. 63. World Health Organization. Global Strategy on Diet, Physical Activity and Health. Geneva: World Health Organization; 2004. 64. Jamison DT, Mosely WH, Measham AR, Bobadilla JL, eds. Disease Control Priorities in Developing Countries. Oxford (England): Oxford University Press; 1993.
65. Jamison DT, Breman JG, Measham AR, Alleyne G, Claeson M, Evans DB, et al., eds. Disease Control Priorities in Developing Countries. 2nd ed. Washington, DC: International Bank for Reconstruction and Development/The World Bank; 2006. 66. EUROASPIRE I and II Group. Clinical reality of coronary prevention guidelines: a comparison of EUROASPIRE I and II in nine countries. Lancet. 2001;357:995–1001. 67. European Society of Cardiology. Scheduled and Ongoing Surveys. Euroaspire III Euro Heart Survey on Secondary and Primary Prevention of Coronary Heart Disease. Available at: http://www.escardio.org/knowledge/ ehs/survey/scheduled-surveys/Euroaspire_III .htm. Accessed January 15, 2008.
46894_CH21_619_656.qxd
2/25/10
4:16 PM
Page 619
C H A P T E R
21 The Case for Prevention factors fail to account sufficiently for causation; interventions to change them are of unknown or uncertain effectiveness; and more immediate public health priorities already consume the available resources. Gaps in evidence are sometimes cited to question whether public health action is justified until further research has been completed. Current issues regarding the case for CVD and chronic disease prevention are, first, to communicate the urgency, potential for prevention, and consequences of inaction to reach the key audiences effectively—the public at large, local, national and global leaders in health and other sectors and policy-makers at all levels; and, second, to take action on the basis of current knowledge.
SUMMARY What is the case for full-scale implementation of public health approaches to cardiovascular disease (CVD) prevention on local, national, and global levels? Arguments specific to CHD, stroke, or other cardiovascular conditions, or CVD more generally, are complemented by those embracing a broader set of major chronic or noncommunicable diseases, including cancer and chronic respiratory diseases as well as CVDrelated conditions. Chronic disease prevention is therefore directly germane to CVD prevention, especially in the arena of national, regional, and global health policy. Several elements of the case for CVD prevention are: First, effectiveness of intervention has been reported from the experience of many clinical and community trials. Second, the growing burden of CVD risk factors for populations throughout the world can be described on the basis of a large body of epidemiologic data. Third, current economic assessments regarding CVD address both the impact of disease and disability and the cost-effectiveness of particular interventions, with special reference to low- and middle-income countries. Fourth, explanation of recent trends in CVD mortality and projections of the future course of the epidemic are now available through several approaches to statistical modeling. Fifth, a number of visionary statements express reasoned expectations of what public health might achieve through effective CVD or chronic disease prevention on a global scale but also propose means to attain such alternative futures. What counter-arguments are raised against largescale public health efforts in CVD prevention? Prominent themes include claims that established risk
INTRODUCTION As concepts of prevention have evolved and understanding of risk factors has grown over the past several decades, the accumulating evidence has been assembled, evaluated, and translated into recommendations, guidelines, and policies for CVD prevention at both individual and population levels. Despite limited implementation of clinical guidelines as discussed in Chapter 20, important progress in CVD prevention has been made over these several decades, accompanied by marked declines in agestandardized CVD mortality in many industrialized countries. However, burdens of CVD are substantial and increasing in low- and middle-income countries, and progress has slowed or reversed in others. These observations add urgency to the fact that strategies of prevention are having much less than their full potential public health impact. In terms of the three
619
46894_CH21_619_656.qxd
620
2/25/10
4:16 PM
Page 620
CHAPTER 21 THE CASE FOR PREVENTION
core functions of public health, assessment and policy development in CVD prevention have advanced substantially, but assurance lags far behind.1 The concepts, strategies, and goals of CVD prevention extend across the full spectrum, illustrated in Figure 18-1, from preventing risk factors to detecting and managing risk, responding to acute events, and reducing risk of recurrent CVD events for those who survive a first event. Public health aspects of prevention across this spectrum range from policy and environmental change to foster maintaining low risk to supporting community and clinical approaches to reduce risk for those in need; to increase public awareness of signs and symptoms of impending CVD events, use of appropriate emergency medical services, and availability of qualified hospital emergency departments; and to assure access to long-term care for rehabilitation, risk reduction, and end-of-life care. The whole array of approaches to intervention is of public health concern. However, those furthest “upstream” are most prominently dependent on public health action. Public health is therefore especially focused on interventions with the earliest potential impact and broadest reach— those aimed at both prevention of risk factors in the first place and detection and treatment of risk factors prior to first CVD events—or, simply, primary prevention. From this perspective, although public health must in principle advocate and pursue a comprehen-
sive strategic approach to CVD prevention, primary prevention is the focus of the present chapter. To put the case for CVD prevention in context, it is important to embrace the other chronic diseases that are of major public health concern and that are likely to be impacted by the same policy interventions. Thus the World Health Organization (WHO) brings together, under the rubric of chronic diseases, heart disease, stroke, cancer, chronic respiratory diseases, and diabetes.2 WHO’s 2005 report, Preventing Chronic Diseases: A Vital Investment, was the forerunner of two series of reports in The Lancet, one appearing in 2005 and the other in 2007, that contribute substantially to the case for preventing chronic diseases—and prominently including CVD.3 Credit for the mounting impetus behind chronic disease prevention as a global health priority is also due to the Disease Control Priorities in Developing Countries Project and the Global Burden of Disease Study.4,5 These and other sources regarding chronic disease prevention are noted here as they bear on the case for CVD prevention. It has long been recognized that factors contributing to cardiovascular risk are related to several other chronic conditions as well. It is therefore expected that effective intervention, through whatever means, will confer multiple benefits. This point was illustrated by Shigan as in Figure 21-1.6 (The original figure has been modified here to include links from ex-
Smoking
Coronary heart disease
Diet
Stroke
Excessive alcohol consumption
Diabetes
Physical inactivity
Lung cancer
Environmental pollution
Chronic bronchitis
Raised blood pressure
Liver cirrhosis
Hyperlipidemia
Gallbladder disease
Blood glucose
Peptic ulcer
Figure 21-1 Risk Factors and Non-Communicable Diseases—Main Links. Source: Reprinted with permission from E Shigan, World Statistics Quarterly, Vol 41, p 268, © 1988, World.
46894_CH21_619_656.qxd
2/25/10
4:16 PM
Page 621
EXPERIENCE WITH MULTIFACTOR PRIMARY PREVENTION 621
cessive alcohol consumption to stroke and from hyperlipidemia to CHD, which were absent in the original. A connection from physical inactivity to diabetes could be added as well.) On the basis of this scheme, strategies that combine efforts to improve all eight factors would be expected to have significant impact not only on CHD and stroke but also on diabetes, lung cancer, chronic bronchitis, liver cirrhosis, gallbladder disease, and peptic ulcer. Even this extended list, however, fails to recognize the role of high blood pressure on heart failure, dementia, and other cardiovascular outcomes of growing importance. These are the main points in the case for CVD prevention: • Experience with multifactor primary prevention has accrued from a large number of studies in the United States and much of the world. What can be learned from this body of work is the cornerstone of the case for prevention. Widespread interest in community approaches suggests increasing readiness over the past decade to take further action. Lessons of experience indicate foremost a need to implement the most promising and comprehensive interventions, in multiple populations, on a large enough scale and with sufficient duration to permit rigorous evaluation. This would offer the greatest opportunity to identify intervention approaches with potential for widespread dissemination and adaptation to local needs and resources. • The fact of global occurrence of CVD on an epidemic level calls for application of current knowledge on a corresponding scale. The burden of risk is global in extent, and every region of the world is experiencing CVD on an epidemic scale. Distributions of particular risk factors vary among populations as do, therefore, their relative and respective populationattributable fractions for CHD and stroke. But the same factors are accountable everywhere. • At a macroeconomic level, CVD and other chronic diseases demand a level of attention and urgency of action that have been seriously underappreciated until quite recently. The economic and social impact of lost productivity, especially—but not exclusively—in low- and middle-income countries, would seem to compel action, the cost of continued inaction being unacceptable from the perspective of public health accountability. On the basis of costeffectiveness analysis, substantial progress could be made by implementing presently affordable preventive measures today.
• Modeling contributes importantly to explanation, description, and prediction of past, present, and future occurrence of CVD and other chronic diseases. Extending beyond the sometimes quite limited observations available, modeling offers insights that can influence decision making about health policy in positive ways. For example, reduction of population-wide risk factors has contributed to approximately half of the gain in coronary heart disease (CHD) mortality in high-income countries in recent decades. This strategy is projected to make continuing major contributions in low- and middleincome countries in the future. Wider interest in modeling can also stimulate strengthening of data sources for future analyses. • The visions expressed in several published statements from responsible organizations represent judgments that go beyond systematic review of evidence on a specific intervention. They reflect not only a sense of what such evidence says, but what it means in terms of societal interests and values. That such belief in the potential for CVD prevention is expressed strongly by many authoritative sources contributes significantly to the case for prevention. • That counter-arguments regarding the case for CVD prevention continue to be raised should not be surprising, given competing interests, priorities, or interpretations of the evidence. Weighed against the elements of the argument in favor of CVD prevention, however, they are not persuasive to many in positions of accountability for the public’s health.
EXPERIENCE WITH MULTIFACTOR PRIMARY PREVENTION The view that the major CVD risk factors and their underlying behavior patterns should be considered together rather than separately is well established. This is reflected in the recent practice of estimating absolute 10-year risk of CHD or stroke for individual patients from multivariate risk models, as discussed in Chapter 20. But already by the early 1970s, the potential for multifactor interventions was a topic of research interest. For example, one proposal for a multifactor trial in the United States was to assess independent and joint contributions of behavior changes in diet, physical activity, and smoking habits in primary prevention of CHD. However, the proposed trial, dubbed “Jumbo” by its designers, was considered by the government to be prohibitively
46894_CH21_619_656.qxd
622
2/25/10
4:16 PM
Page 622
CHAPTER 21 THE CASE FOR PREVENTION
expensive and was never implemented. Instead, such single-factor treatment trials as the Coronary Primary Prevention Trial of cholesterol reduction and Hypertension Detection and Follow-up Program (see Chapters 11 and 12, respectively) were undertaken. The joint impact of such measures for singlerisk-factor change would remain unknown until multifactor trials could be accomplished, however. The need for this evidence became compelling. Studies were initiated that represented both the high-risk strategy and the population strategy of prevention. Clinical trials were undertaken in groups selected for having already-established risk factors, for whom intervention would be applied individually, with spouses, or in groups. Community intervention trials tested population-wide interventions, especially those involving community organization and education, sometimes including risk-factor screening with targeted intervention for high-risk individuals.7 (Multifactor trials of secondary prevention were also conducted and showed benefit among persons with recognized CHD, but the present focus is on primary prevention.) The ultimate goal of these studies was to test the degree to which multifactor intervention could reduce the frequency of CHD. Adults were therefore the principal target for assessing effectiveness of both clinical and community interventions. However, studies in school-age populations were also initiated, usually separately, often with change in risk-factor levels as the primary outcome. As intermediate objectives, most of these studies also assessed success in delivery of interventions and ability to bring about relevant behavior change. It was expected that the lessons learned would guide further development of policies and practices for CVD prevention and strengthen the scientific basis for their implementation. Several examples illustrate the methods and results of such multifactor intervention trials. Intervention in Individuals The Multiple Risk Factor Intervention Trial (MRFIT) It was recommended in 1971 that a trial be undertaken of the combined effects of intervention to reduce high total cholesterol, high blood pressure, and smoking among high-risk men.8 Independent effects of the interventions could not be evaluated statistically, as Jumbo would have allowed, but study size and cost would be acceptable with the proposed design. High risk was defined in relation to a three-variable risk score from the Framingham Study based on the targeted risk factors. Initially the scores at or above the 85th percentile defined eligibility, but to increase the
average risk level of participants, this was raised to the 90th percentile after one-third of the screening had been completed. Men aged 35–57 years were invited to participate. Persons with sustained qualifying riskfactor levels after the third visit were randomly allocated to intervention (special intervention, SI) or control (usual care, UC). The UC group members were expected to receive risk-factor intervention as customarily delivered by their own sources of medical care. The SI group members received interventions for their respective risk factors, beginning with intensive group counseling and reinforced with individual counseling and therapy by a multidisciplinary team throughout the course of the trial. The primary end point of the trial was CHD mortality. As a result of screening 361,662 men in 22 clinical centers, 12,866 men were randomized into the trial. Results reported at the planned close of the intervention period were based on an average followup of 7.0 years.9 Cumulative coronary mortality was 17.9/1000 in SI and 19.3/1000 in UC, a nonsignificant difference of 7.1% with a 90% confidence interval of 15% to 25%. A difference of 26.6% had been projected in the design. Total mortality was slightly greater in SI than UC (41.2 versus 40.4/1000). Failure of intervention to bring about the expected relative reduction in CHD mortality was due, in part, to unexpected improvement in risk factors in UC and occurrence of only approximately two-thirds of the expected coronary mortality in this group. A consequence was to make the difference between groups more difficult to detect. In addition, relatively adverse experience of some participants receiving antihypertensive drug therapy offset the benefits of reductions in cholesterol concentration and smoking in participants without hypertension. This may have resulted from an anomalous deficit of deaths among UC hypertensives. Through continuing mortality follow-up of study participants after the close of the intervention period, MRFIT extended the analysis of intervention effects through 10.5 years and 16 years in subsequent reports.10,11 Ultimately, after 16 years, total deaths had reached three times the number in the initial analysis, with total mortality 5.7% less in SI than in UC (Table 21-1). For fatal acute myocardial infarction the SI-UC difference, 20.4%, was statistically significant. Because risk-factor status was not being monitored in most centers in the posttrial period, interpretation of this result is unclear. However, during the intervention period, fewer nonfatal coronary events had occurred in SI. On this basis, it was predicted that mortality would be reduced in SI subsequently, and the 16-year findings confirmed this prediction. These later results
46894_CH21_619_656.qxd
2/25/10
4:16 PM
Page 623
EXPERIENCE WITH MULTIFACTOR PRIMARY PREVENTION 623
Table 21-1
Cause of Death for MRFIT SI and UC Men Through December 1990 Cause of Death ICD-9 Code No. SI, n UC, n All cardiovascular 390–459 507 550 Acute myocardial infarction 410 185 232 Other ischemic (coronary) heart disease 411–414, 429.2‡ 185 185 Cardiac dysrhythmias 427 15 21 Hypertensive heart disease 402 10 12 Other hypertensive 401, 403–405 6 2 Cerebrovascular 430–438 46 44 Other cardiovascular disease ... 60 54 All noncardiovascular 483 499 Neoplastic 140–239 316 321 Lip, oral cavity, and pharynx 140–149 5 12 Digestive organs and peritoneum 150–159 73 88 Colorectal 153–154 28 33 Other gastrointestinal 150–152, 155–159 45 55 Respiratory and intrathoracic organs 160–165 141 122 Lung 162 135 117 Other neoplasms ... 97 99 Respiratory 460–519 25 31 Digestive system 520–579 40 33 Accidents, suicides, and homicides 800–999 55 58 Other non-cardiovascular disease ... 47 56 Cause unknown (death certificate not found) ... 1 1 Total ... 991 1050
Relative Difference, % 7.9 20.4† 0.1 29.0 17.2 ... 5.2 11.0 3.3 1.8 ... 17.2 15.2 18.5 15.2 15.0 2.3 19.2 21.0 5.1 15.9 ... 5.7
*(RR–1) 100%, where the RR (relative risk) is estimated from the proportional-hazards regression model. † p .02; p .05 for all other relative differences. Relative difference is not given if there were 10 deaths in either the SI or UC. ‡ In ICD-9, No. 429.2 is cardiovascular disease, unspecified; in ICD-8 this is coded to coronary heart disease No. 412.4 Source: Reprinted with permission from MO Kjelsberg, MRFIT trial, Circulation, Vol 94, p 948, © American Heart Association.
were taken as evidence that all three interventions, targeting cholesterol, blood pressure, and smoking, conferred benefit that might continue to favor SI over still-longer follow-up. Other Multifactor Primary Prevention Trials in Individuals Three other clinical trials of multifactor intervention were reviewed in the 16-year MRFIT report, studies in Oslo (Norway), Göteborg (Sweden), and Helsinki (Finland), whose results appeared from 1980 through 1995. The Oslo Study included 1232 men aged 40–49 years at entry and free of cardiovascular symptoms and diabetes.12 Cholesterol concentrations were quite high, from 290 to 379 mg/dl, or the coronary risk score was in the upper 25% without high blood pressure; 80% of the men were smokers. Aggressive advice was provided on smoking cessation and dietary change to reduce fat and cholesterol and increase polyunsaturated fats in the diet for men randomized to intervention. At 5 years, several CHD endpoints were significantly less frequent in the intervention
group. After an additional 2.5 years of follow-up, without intervention, risk-factor levels converged between groups: prior smokers in the intervention group relapsed, but prior controls reduced their cholesterol levels. Reported mortality differences remained significant. The Multifactor Primary Prevention Trial (Göteborg, Sweden) sampled men aged 47–55 years, with approximately 10,000 participants in intervention and in each of two control groups.13 In the intervention group, all men were screened to identify those at high risk, for whom intervention was directly implemented. Criteria for high-risk status included quite high levels of blood pressure or cholesterol concentration and smoking of at least 15 cigarettes per day. One of the two control groups was sampled (2% at baseline, 11% at 4 years, and 20% at 10 years) for risk-factor assessment without intervention, and the second control group was not screened at all but only monitored to ascertain occurrence of endpoint events. Reductions in risk factors were closely parallel between intervention and control groups, and mortality after 11.8 years follow-up was not appreciably
46894_CH21_619_656.qxd
624
2/25/10
4:16 PM
Page 624
CHAPTER 21 THE CASE FOR PREVENTION
different between groups. Within the intervention group, there were nonparticipants whose mortality was notably higher than that for participants. Given that only the highest-risk individuals were targeted for intervention and many events arise from more moderate levels of risk in the population, intervention may not have reached a sufficiently broad segment of the population to attain the intended effect. The Helsinki Multifactorial Primary Prevention Trial enrolled more than 1200 business executives, approximately 40–55 years of age at randomization in 1974, in a trial of dietary and hygienic measures and drugs to lower cholesterol concentrations or blood pressure.14 The control group participants were untreated by the investigators. After five years of initial follow-up, total risk showed a net reduction of 46% in the intervention group. Strokes were significantly less frequent but coronary events were more frequent in the intervention group. The first posttrial followup was completed 5 years later, in 1985, and the second after another 5 years, or 15 years after randomization. The intervention group experienced significantly higher mortality for cardiac events and accidents and for all causes together. The use of multiple and changing drugs for therapy throughout the trial period complicated attempts to explain the findings, although particular drugs were implicated by the authors. Regardless, the mortality outcome remained clearly negative. These and other studies through 2001 were described in a Cochrane meta-analysis published by Ebrahim and colleagues in 2006, an update of a 1995 review.15 The intent was to aggregate studies of counseling or education, with or without drug therapy, to reduce more than one CVD risk factor in a general population, occupational group, or high-risk group. (Curiously, the Hypertension Detection and Followup Program was included, although it was specifically targeted to reduce high blood pressure; attention to other risk factors was only incidental in keeping with usual standards of care.) Each of 39 studies was described, allowing the reader to assess the body of work available at the time. Of the 39 selected studies, only 10 reported total or CHD mortality outcomes. The studies ranged widely in size and varied in other aspects of design as well. Overall, they showed significant reductions in blood pressure, blood cholesterol, and smoking, but no significant reductions in total or CHD mortality. The authors concluded:15, p 1 The pooled effects suggest multiple risk factor intervention has no effect on mortality. However, a small, but potentially important, benefit of treatment (about a 10% reduction in CHD mortality)
may have been missed. . . . Interventions using personal or family counseling and education with or without pharmacological treatments appear to be more effective at achieving risk factor reduction and consequent reductions in mortality in high risk hypertensive populations. The evidence suggests that such interventions have limited utility in the general population. It is important to emphasize that the interventions reviewed here were primarily physician advice, not necessarily including either intensive lifestyle intervention nor medication. The general conclusions should be interpreted accordingly. A contrasting view of multifactor intervention in individuals was presented in 2003 by Wald and Law.16 They presented meta-analyses of single-factor interventions targeting LDL-cholesterol, blood pressure, serum homocysteine, or platelet function to suggest the potential benefit of a combination pharmacological intervention. They proposed use of a combination “Polypill” as a “strategy to reduce cardiovascular disease by more than 80%. . . . This strategy would be suitable for people with known cardiovascular disease and for everyone over a specified age (say 55), without requiring risk factor measurement.”16, p 1 The pill would contain a statin, three blood pressurelowering drugs at half standard dose, folic acid, and aspirin. One of their premises in arguing for a pharmacologic approach was their view that, although Western diet and lifestyle underlie the causal risk factors for preventable CVD, changes in these conditions could not be achieved in the short term. Further, they argued, CVD occurs to a greater extent among those at moderate rather than extreme risk; risk is strongly predicted by age; risk factors apart from age contribute little to improving prediction of events; and CVD risk is reduced in proportion to reduction in a given risk factor regardless of the starting level. They concluded:16, p 5 It is time to discard the view that risk factors need to be measured and treated individually if found to be “abnormal”. Instead it should be recognised that in Western society the risk factors are high in us all, so everyone is at risk; that the diseases they cause are common and often fatal; and that there is much to gain and little to lose by the widespread use of these drugs. No other preventive method would have so great an impact on public health in the Western world. Extensive comment, pro and con, followed immediately after the Wald and Law publication. An
46894_CH21_619_656.qxd
2/25/10
4:16 PM
Page 625
EXPERIENCE WITH MULTIFACTOR PRIMARY PREVENTION 625
Internet search reveals continuing interest in development and testing of various multidrug formulations as of mid-2008. (Available at www.medscape .com/viewarticle/575014. Accessed 8 May 2008.)17 Intervention in Communities In contrast to the foregoing examples, in which intervention was directed to individuals at high risk of coronary and other cardiovascular events, other multifactor intervention studies were designed to mobilize community-wide action and thereby modify knowledge, attitudes, and behavior related to the major risk factors. In some instances, risk-factor detection was included, with reinforcement of local health services to improve risk-factor treatment and control, but educational and other community strategies remained the central aspect of these programs. Six studies are summarized in Tables 21-2 and 21-3 and are further described as follows.18 Five of these were prospective studies, each with one or more intervention and control communities; the sixth study was a retrospective analysis of changes in CVD mortality concurrent with preventive programs reaching defined target populations. These programs used various intervention strategies, but the majority had several elements in common. Coupled with the differences among population settings, calendar periods, and other aspects of program design and implementation, variation in intervention content would be expected to yield differences in outcomes. Two programs used professional nursing staff and primary care integration; all included community organization, mass media, schools, groceries and restaurants, and group education. The North Karelia Project Among the earliest community intervention trials was the North Karelia Project, undertaken in 1972 in view of the exceptionally high mortality from coronary heart disease in Eastern Finland in the late 1960s.19 Recognition of the magnitude of the problem had led to recruitment of two cohorts, in East and West Finland, into the Seven Countries Study more than a decade earlier. A comprehensive review of the extraordinary 20-year North Karelia Project and its local, national, and international impact was published in 1995.20 One intervention area, North Karelia (population 210,000), was compared with one reference area, Kuopio (population 250,000). Multiple programs were implemented in North Karelia with the aim of reducing blood cholesterol concentration, controlling high blood pressure, and achieving smoking cessation for as many persons as possible in this target population. Risk-factor distributions in both com-
munities were assessed by cross-sectional surveys in independently drawn random samples, including men and women aged 30–59 years in each population. The surveys were conducted every 5 years from 1972 to 1992. In 1982 a third area, southwestern Finland, was added for further comparison of data in the last three surveys. Mortality was monitored from 1969 through 1992. Evaluation of the impact of intervention in North Karelia is complicated by large concurrent changes in risk factors and coronary mortality in the reference area.21 Risk-factor changes from 1972 to 1992 were substantial in North Karelia: Cholesterol concentration decreased by 13% for men and 18% for women, diastolic blood pressure decreased by 9% for men and 13% for women, and smoking decreased from 53% to 37% for men (but increased from 11% to 20% for women) (Table 21-4). But only during the first five years of the program were changes in these risk factors greater in North Karelia than in Kuopio. They were similar thereafter, and coronary mortality decreased 50% in Finland as a whole from 1970 to 1992. These circumstances, in addition to the sample size of only one intervention unit and one reference or control unit (despite the later addition of a second control area), limited the ability to evaluate program effects. One approach to analysis of risk factor and mortality change was to estimate the relation between each risk factor and coronary mortality within the Kuopio population and then to predict the mortality at each survey year on the basis of the risk-factor levels in the population at that time. The predicted mortality, taking into account smoking, blood pressure, and cholesterol concentration, was then compared with the observed mortality from 1972 to 1992 for men and women aged 35–64 years in Finland. These results are shown in Figure 21-2 for men and Figure 21-3 for women. A marked decline in coronary mortality was predicted throughout the period. For men, the observed decline was close to the prediction until 1985, when it began to exceed the prediction. For women, the decline exceeded the prediction throughout the entire period. Further analysis demonstrated that from 1972 through 1987, for both men and women, the observed and predicted trends were consistent, and it was concluded that the decline in mortality could be explained by the risk-factor changes over the same period. The interventions are not the only plausible explanation of these mortality changes, given that there were risk-factor changes in the reference area. Implementation of a number of prevention activities nationally during the course of intervention in North Karelia may have affected this target area as well as
Northern California
1972–1975
1980–1986
1981–1988
Stanford 3-CP (16)
Stanford 5-CP (17)
Minnesota (18) Minnesota, North and South Dakota
1 similar town
3 similar cities
3 similar cities
About 75,000 per city
Small: 25–40,000 Medium: 75–80,000 Metro: 80–115,000
2 cities
3 cities, small, large, metro
Comparison Community Similar region and Finland
12–15,000 per town
Population 180,000 in region
Improved risk factor knowledge, saturated fat intake, cigarette consumption, plasma cholesterol and blood pressure control, and projected cardiovascular risk by 15% to 20%. Mass media more cost-effective. Sustained improvements in blood pressure but not in physical activity. No reductions in cardiovascular morbidity or mortality. Higher education exposure scores and favorable risk factor changes. No reductions in cardiovascular morbidity or mortality.
Mass media only vs. mass media plus individual attention to high-risk individuals.
Mass media only. No individualized interventions.
Face-to-face communications, public events, TV
Prospective controlled
Prospective controlled
Unique Focus Indigenous impetus. Community ownership. Integration with health care. Sustained focus on risk factors among individuals.
Prospective controlled
Study Design Prospective controlled
4:16 PM
3 towns
Intervention Community North Karelia
Associated Outcomes Relative to Comparison Populations Improved risk factors. Reduced cardiovascular and cancer deaths.
2/25/10
Northern California
Location Rural Finland
Years 1972–1997
Features and Outcomes of Six Major Community CVD Prevention Trials
626
Trial/Reference North Karelia (14)
Table 21-2
46894_CH21_619_656.qxd Page 626
CHAPTER 21 THE CASE FOR PREVENTION
Franklin County
Rural Maine
1974–Present
Franklin, Maine (26)
Retrospective ecologic observation
Adjacent, similar counties and state
40,000 in county
Integration of public health, medical care and community resources. Risk factor counseling, tracking and follow-up over time by 1-on-1 nurse encounters.
Community organization, campaigns; screening, counseling and referral.
Reduced: total, cardiovascular and cancer mortality; cardiovascular and “preventable” hospitalizations and hospital charges; and smoking rates. Dosedependent impact of nurse encounters on death rates.
Transient improvements in risk factors and risk ratio for projected cardiovascular disease rates. No reductions in cardiovascular morbidity or mortality.
Source: Reprinted with permission from PA Ades, TE Kottke, N Houston Miller, et al., Journal of the American College of Cardiology, Vol 40, © 2002 American College of Cardiology Foundation, p 618.
Prospective controlled
1 nearby similar city
70,000 in city
4:16 PM
CVD cardiovascular disease.
1 city
Southern New England
1984–1991
2/25/10
Pawtucket (19)
46894_CH21_619_656.qxd Page 627
EXPERIENCE WITH MULTIFACTOR PRIMARY PREVENTION 627
46894_CH21_619_656.qxd
628
2/25/10
4:16 PM
Page 628
CHAPTER 21 THE CASE FOR PREVENTION
Table 21-3
Intervention Strategies of Six Major Community CVD Prevention Trials Strategies North Stanford Stanford Karelia 3-Community 5-Cities Minnesota Community organization Mass media Environmental modifications 0 0 Community groups 0 Schools Worksites 0 Groceries and restaurants Medical settings 0 Professional education 0 Health agencies collaboration 0 Train local personnel 0 Lay volunteer emphasis 0 0 0 Self-management focus 0 Group education Risk factor screening 0 Individual counseling 0 Referral for medical care 0 0 Client risk factor tracking 0 0 Active client follow-up 0 0 Professional nursing staff 0 0 0 Primary medical care integration 0 0 0
Franklin Maine 0
Pawtucket Print only 0 0 0 0
CVD cardiovascular disease “” indicates characteristic present, but does not imply equivalent intensity of intervention components. Source: Reprinted with permission from PA Ades, TE Kottke, N Houston Miller, et al., Journal of the American College of Cardiology, Vol 40, © 2002 American College of Cardiology Foundation, p 619.
other regions, independently of the planned local programs. Nonetheless, this program contributed importantly to development and implementation of community intervention strategies. It is widely regarded as pioneering work in the field providing tested models for programs applied throughout Finland and elsewhere.
Table 21-4
Community Intervention in the United States The six studies summarized in Tables 21-2 and 21-3 include five conducted in the United States: the threeand five-city phases of the Stanford Heart Disease Prevention Program, the Minnesota and Pawtucket (Rhode Island) Heart Health Programs, and the Franklin Cardiovascular Health Program in Franklin
Mean (Standard Error) Level of Coronary Risk Factors in Subjects in Finland, by Year and Sex 1972 1977 1982 1987 1992
Risk Factors Men Cholesterol (mmol/l) Diastolic blood pressure (mm Hg) Smoking (% of study population who were smokers) Women Cholesterol (mmol/l) Diastolic blood pressure (mm Hg) Smoking (% of study population who were smokers)
6·78 (0·02) 92·8 (0·18) 53 (0·8)
6·55 (0·02) 91·0 (0·18) 47 (0·8)
6·28 (0·02) 87·8 (0·26) 42 (1·0)
6·23 (0·03) 88·4 (0·28) 39 (1·2)
5·90 (0·03) 84·2 (0·37) 37 (1·5)
6·72 (0·02) 91·8 (0·19) 11 (0·5)
6·36 (0·02) 87·6 (0·17) 12 (0·5)
6·10 (0·03) 84·6 (0·25) 16 (0·8)
5·94 (0·03) 83·5 (0·26) 16 (0·9)
5·54 (0·03) 79·6 (0·33) 20 (1·1)
Source: Reprinted with permission from E Vartiainen et al., Changes in Risk Factors Explain Changes in Mortality from Ischemic Heart Disease in Finland, British Medical Journal, Vol 309, pp 24–27, © 1994, BMJ Publishing Group.
46894_CH21_619_656.qxd
2/25/10
4:16 PM
Page 629
EXPERIENCE WITH MULTIFACTOR PRIMARY PREVENTION 629
0
Smoking
Decline in Mortality (%)
10 Blood Pressure
20
Cholesterol
30
Predicted (all risk factors)
40
50
Observed (all risk factors) 60 1972 74
76
78 80
82 Year
84
86
88
90
92
Figure 21-2 Observed and Predicted Decline in Mortality from Ischemic Heart Disease in Men Aged 35–64 in Finland. Source: Reprinted with permission from E Vartiainen et al., Changes in Risk Factors Explain Changes in Mortality from Ischemic Heart Disease in Finland, British Medical Journal, Vol 309, p 25, © 1994, BMJ Publishing Group.
220
Smoking
Decline in Mortality (%)
0
Blood Pressure
20
Cholesterol 40
Predicted (all risk factors)
60
Observed (all risk factors) 80 1972 74
76
78
80
82 Year
84
86
88
90
92
Figure 21-3 Observed and Predicted Decline in Mortality from Ischemic Heart Disease in Women Aged 35–64 in Finland. Source: Reprinted with permission from E Vartiainen et al., Changes in Risk Factors Explain Changes in Mortality from Ischemic Heart Disease in Finland, British Medical Journal, Vol 309, p 25, © 1994, BMJ Publishing Group.
County, Maine. Joint analysis of the Five-City Stanford Program and Minnesota and Pawtucket Heart Health Programs demonstrated the aggregate intervention effects on smoking, systolic and diastolic blood pressure, cholesterol, body mass index, and estimated 10-year risk of CHD mortality (Table 21-5).22 Findings of the Franklin study are described separately. Stanford. Concurrent with initiation of the North Karelia Project, and with close collaboration between the two responsible research groups, the Stanford Three-City Study began in northern California.23 The strategy was to implement community-wide health education in two communities, with supplemental individual counseling for a sample of high-risk individuals identified in one of them; the third community would serve as the control. The initial results after 2 years of intervention suggested that community-wide health education had substantial impact, with an overall reduction in a multivariate risk score of 23–28% in treatment compared with control communities. This favorable early experience led to a second study, the Stanford Five-City Project, based on newly selected communities.7,24 Two communities received a 5-year intervention program based on social-learning theory, a communication-behavior change model, community-organization principles, and social-marketing methods. In addition to continuous education programs, several short-term campaigns were also conducted in the two intervention communities. Two of the remaining communities served as observed controls and the third was subject only to monitoring of event rates as an unobserved control. Stroke and CHD were the primary end points, and change in blood cholesterol concentration, blood pressure, smoking prevalence, body weight, resting pulse rate, and knowledge of risk factors were intermediate end points. The program used several approaches to bring about riskfactor change (Table 24-3). Evaluation of risk-factor change was conducted both in cohorts and in independent cross-sectional samples in four of the study communities. Most risk factors improved from baseline, to a greater extent in the treatment communities. The overall multivariate risk score, based on a model for 12-year coronary event rates in the Framingham Study, was predicted to show a net reduction of 20% in treatment communities. For all-cause mortality, significant decreases in risk score of 16% and 14% were observed at the third and fourth cohort examinations, respectively; coronary event risks were correspondingly reduced 17% and 16% in the cohort evaluation. Minnesota. In 1980, a six-community intervention trial began in the upper Midwest of the United States, the Minnesota Heart Health Program.25 Three pairs of communities were matched on population size, type (from small to urban), and distance from
46894_CH21_619_656.qxd
630
2/25/10
4:16 PM
Page 630
CHAPTER 21 THE CASE FOR PREVENTION
Table 21-5
Cardiovascular Risk Factor Trends in Women and Men Aged 25–64 in the Stanford Five-City Study, Minnesota Heart Health Study, and Pawtucket Heart Health Study, Combined, Adjusted for Age and Education, 1978–1990 Results Expressed as Net Average Annual Change in Treatment Communities Minus Control Communities* Linear Trend in Tests of Net Intervention Linear Trend in Tests of Net Intervention Risk Factor Women (per Year) Effect in Women Men (per Year) Effect in Men Current smoking, % 0.30 0.42 p 0.48 0.31 0.51 p 0.54 Systolic blood pressure, 0.31 0.22 p 0.17 0.10 0.24 p 0.68 mm Hg Diastolic blood pressure, 0.23 0.16 p 0.15 0.09 0.23 p 0.68 mm Hg Cholesterol, mg/dl 0.70 0.48 p 0.15 0.23 0.51 p 0.66 Body mass index, kg/m2 0.06 0.05 p 0.19 0.03 0.04 p 0.46 Log10 estimated 10-year 0.001 0.003 p 0.85 0.001 0.002 p 0.64 coronary heart disease mortality risk *Estimate standard error, from mixed-model analysis of variance with pooled error term. Source: Data from MA Winkleby, HA Feldman, DM Murray, Journal of Clinical Epidemiology, Vol 50, © 1997 Elsevier Science, Inc., pp 651–652.
metropolitan Minneapolis–St. Paul. One community of each pair received a multicomponent education program and the other served as a control. The program was described as a high-intensity campaign that included risk-factor screening, education and intervention programs, involvement of most primary care physicians and many other health professionals in training activities, changes in community organization and environment conducive to heart health, and participation by youth in school-based health promotion programs. An intervention program of 5 or 6 years’ duration was projected to reduce population mean values for blood cholesterol concentration by 7 mg/dl, blood pressure by 2 mm Hg, and cigarette smoking by 3% and to increase energy expenditure in physical activity by 50 kcal/day. In the final analysis of event rates, for neither stroke nor CHD was there a significant difference in rates between intervention and control communities.26 Pawtucket. The third major multifactor community intervention trial of cardiovascular disease prevention in the United States was the Pawtucket (Rhode Island) Heart Health Program, in which one intervention and one control community were enrolled.27 The aim was to evaluate the impact on single risk factors and projected CVD rates on the basis of a composite risk score for persons aged 35–64 years in the Framingham Heart Study. Low-cost approaches to community behavior change and environmental changes conducive to the desired behaviors were implemented from 1984 to 1991. Heart health curricula in schools from grades one to high school were included. As in the Stanford and Minnesota programs, both cross-sectional and cohort examinations
were conducted for evaluation, beginning with preintervention assessment in 1981. Overall, the changes observed were generally favorable but modest, and they were similar between intervention and comparison communities. This outcome was attributed to the relatively small proportion of total media messages affecting health-related behavior represented by the program, in contrast with the total environment of advertising, marketplace displays, and countless inputs from other sources. Table 21-5 showed for women and for men in the age range from age 25 to 64 years the results of change in single risk factors (cigarette smoking, blood pressure, total cholesterol concentration, and body mass index) and in the composite coronary heart disease mortality risk score, for the three studies combined. Overall, 9 of 12 comparisons were favorable in direction, showing greater change in the intervention communities (five for women and four for men), but none of these results was statistically significant. Franklin Cardiovascular Health Program. Retrospective evaluation of trends in CVD and total death rates in Franklin County, Maine, in comparison with statewide rates from 1974–1994, suggested lower mortality during periods of intensified community-wide prevention programs (Figure 21-4).28 The intervention, staffed by professional nurses, comprised education, screening, counseling, referral, tracking, and follow-up, all integrated with primary care. The key indicator of program activity was the annual number of program encounters related to high blood pressure (1974–1985) or cholesterol and high blood pressure control together (1986–1994). Encounters due to these programs fluctuated in fre-
46894_CH21_619_656.qxd
2/25/10
4:16 PM
Page 631
EXPERIENCE WITH MULTIFACTOR PRIMARY PREVENTION 631
1.16 Relative Death Rates
1.1
Total Death Relative Rates
1.06 Maine
1 0.96 0.9 0.86 Franklin
0.8
Encounters per Year
0.76 3000
Annual Encounters per 10,000 Franklin Population
2400 2000 1600 1000
Growth
Decline
Growth
HBP
Decline
CHOL
600 1.15 Relative Death Rates
1.1 1.05 1
Heart Death Relative Rates Maine
0.95 0.9 0.85 0.8 0.75
96
94
19
92
19
90
19
88
19
86
19
19
82
19
80
19
78
19
76
19
74
19
19
84
Franklin
0.7
Figure 21-4 Age-Adjusted Total and Heart Disease Death Rates for Franklin County and Maine, 1960–1994, with Study Periods and Franklin Program Phases. Source: Reprinted with permission from NB Record, DE Harris, SS Record, et al., American Journal of Preventive Medicine, Vol 19, © 2000 American Journal of Preventive Medicine.
quency over time with changing funding levels. During periods of increasing frequency of such encounters, death rates from heart disease or CHD decreased relative to statewide rates. A statistically significant inverse correlation was found between encounters and death rates over the total period of analysis. Community Trials in Europe Communities are definable in other than geographic terms, and employment communities or worksites have also been used for multifactorial cardiovascular
risk reduction programs. A major example of this approach is the European Collaborative Trial of Multifactorial Prevention of Coronary Heart Disease, conducted in 80 factories in Belgium, Italy, Poland, and the United Kingdom.29 The intent was to assess the degree to which educational efforts, undertaken at modest cost, could bring about risk-factor change and thereby reduce incidence and mortality from CHD. Within each country, factories or other large occupational units were matched as pairs and randomly allocated to intervention or control. In intervention factories, all men aged 50–69 years underwent risk-factor screening. This was followed for the higher-risk men—those in the upper 10–20% of risk in accordance with a simple risk function—by treatment or advice for lowering cholesterol, smoking cessation, daily physical exercise, weight reduction, and blood pressure-lowering drug therapy. Less intensive education was provided to the remaining men. In the control factories, only a 10% random sample of men were examined for risk-factor status, leaving 90% unaffected or even unaware of the trial. Follow-up for risk-factor change occurred in independent samples of men in the intervention factories, and the 10% sample in control factories underwent a repeat examination. Morbidity and mortality data were obtained through continuous monitoring of the study sites. Recruitment into the trial in the four countries took place in the early 1970s, and 60,881 men were entered. Results were presented in 1986, with extension of the analysis and some corrections appearing in 1987.30,31 After adjustment for baseline differences in risk factors within each pair of intervention and control factories, risk-factor change in intervention factories relative to control factories was found to be significantly associated with reduction in 6-year incidence of fatal coronary heart disease, total coronary heart disease, and total mortality. The actual reduction in coronary disease incidence was only 62% of the amount predicted from the observed risk-factor changes, and pooled results for the whole trial (in contrast to comparisons within countryspecific pairs) were not significant by two-tailed statistical tests; outcomes were least favorable in the United Kingdom factories and most favorable in Belgium, where exceptionally large reductions in risk factors were observed early in the trial during the most intensive intervention.32 German Cardiovascular Prevention Study. This study illustrates a different design, with outcomes in selected intervention regions compared against samples of the total national population as the control.33 In six regions of former West Germany, a 7-year
46894_CH21_619_656.qxd
632
2/25/10
4:16 PM
Page 632
CHAPTER 21 THE CASE FOR PREVENTION
program was implemented to achieve reductions in smoking prevalence and body mass index. From 1985 to 1991, significant net reductions were found between intervention communities and control samples for systolic and diastolic blood pressure and total cholesterol concentration. Concentration of HDLcholesterol increased initially and then declined, and the overall net decrease was not significant. Overall, change in body mass index showed no net difference between groups. Results over the full intervention period were similar for women and men, with somewhat greater net decreases in blood pressure for women. Northern Sweden. The Västerbotten Intervention Program, located in a rural area of Northern Sweden, combined multiple interventions strategies: health surveys for persons at ages 30, 40, 50, or 60 years; media programs and health counseling; worksite and school programs; food labeling and cooking activities; cultural programs; and group and public meetings. Integration with primary care was a prominent feature.34,35 The study used multiple cross-sectional surveys of the intervention area and periodic surveys of the neighboring MONICA Project reference area to monitor program impact. Reductions in risk of CHD death within the intervention and the control regions after 10 years of observation were estimated from the North Karelia risk model (Table 21-6). The crude or unadjusted risk reduction was 34% in the intervention area and 9% in the reference area; differences between areas were roughly similar by age and sex. By education, however, strikingly more favorable changes occurred for the lowest stratum, with 33% improvement in the intervention area and 15% increase in the reference area over the study period. These findings were taken to indicate that this community-wide intervention did not widen but rather narrowed the gap in CHD risk between lower and higher social strata. Concern that intervention will inevitably widen health disparities because of differential access and benefit across socioeconomic strata is belied by this experience. Community Intervention in Developing Countries In addressing evidence for community intervention in low- and middle-income countries, Gaziano and colleagues cited reports from China, Mauritius, Poland, and South Africa.36 Only one of these—the Coronary Risk Factor Study (CORIS) in the Cape Province of South Africa—was designed to compare intervention and control communities, but the others deserve brief mention.37 In the Noncommunicable Disease Intervention Areas in Tianjin, China, high blood pressure was ad-
dressed through efforts to reduce salt intake and body weight and increase antihypertensive therapy for persons with high blood pressure.38 These efforts were undertaken in the context of major policy changes regarding organization of primary care throughout the municipality of Tianjin.39 They were monitored by introduction of cross-sectional surveys of CVD risk factors and, separately, of salt intake between 1989 and 1996. Prevalence of hypertension declined over this period especially among persons aged 45–64 years; overweight and obesity decreased overall but increased in younger adults; and salt intake (12 g/day) was unchanged.38 In Mauritius, a multicomponent primary prevention program was evaluated by pre- and postintervention surveys conducted in 1987 and 1992.40 The program comprised mass media, fiscal measures, legislation, and health education activities in schools, workplaces, and communities to promote improved nutrition and physical activity, smoking cessation, and reduced alcohol intake. Statistically significant decreases in prevalence of hypertension, serum total cholesterol concentration, cigarette smoking, and heavy alcohol consumption were observed. Leisure time physical activity increased. Population distributions of blood pressure, cholesterol, and a composite risk-factor score improved, although prevalence of overweight or obesity increased. Especially noteworthy was a marked reduction in mean serum total cholesterol concentration, by approximately 0.80 mmol (32 mg/dl), a decrease attributed to nationwide substitution of soy oil for palm oil as the principal cooking oil.41 In Poland, consumer subsidies for purchase of animal products were withdrawn, resulting in a marked reduction in use of such foods, concurrent increase in fresh fruit and vegetable intake, and aggressive marketing of (low trans-fat) margarine.42 Subsequently, from 1991–1994, exceptionally sharp declines in mortality from ischemic heart disease ( 25%) were observed and have been described as greater than any previously known decline in CVD in peacetime. Stroke rates also declined in this short period ( 10%). Subsequent analysis of food availability data for Poland, as well as trends in medical care, smoking, alcohol consumption, and stress, has been taken as supporting the argument for abrupt change in nutrition as the most plausible explanation of the decline in CVD mortality.43 CORIS compared three Afrikaner communities in the Cape Province of South Africa—one community receiving low-intensity intervention (mass media only), one receiving high-intensity intervention (adding individual-level intervention for persons at
6.46 2.34 4.40
7.98 2.61 5.30 5.46 2.47 8.60
Age-and-education-adjusted All Women Men 3.49 1.66 5.50
5.32 2.56 4.00
3.49 1.65 5.26 0.67 5.89 3.37
4.28 1.67 6.73
6.43 2.12 4.37
33 2 24 36 33 36
4.28 1.62 6.85 0.81 7.51 4.28
34 33 36 30 38 37
4.39 1.78 7.01
7.33 2.27 4.80
4.30 1.71 6.97 0.76 7.48 4.23
Source: Reprinted with permission from L Weinhall, G Hellsten, K Boman, et al., Scandinavian Journal of Public Health, Vol 29, © Taylor and Francis 2001, p 65.
a
Risk equation for men: IHD 1/(1 e12.73 0.108 age 0.806 smoking 0.021 diastolic blood pressure 0.384 total cholesterol). Risk equation for women: IHD 1/(1 e14.90 0.104 age 1.240 smoking 0.0306 diastolic blood pressure 0.365 total cholesterol).
4.33 1.90 6.63
4.11 1.77 6.31 0.65 7.60 4.25
5.28 2.45 8.16 0.95 9.51 5.38
4.25 1.77 6.66
7.34 2.32 4.94
3.91 1.55 6.39 0.68 6.76 3.83
1 6 1
15 9 13
9 4 7 16 10 11
4:16 PM
Crude All Women Men 30–40 years 50–60 years Age-adjusteda Education Low Medium/high Education-adjusted
Trend for Estimated CHD Mortality Risk (%) in Intervention and Reference Areas. The Risk Estimationa is Based on Data from North Karelia, Finland. Unadjusted, Age-Adjusted, Education-Adjusted, and Joint Age-and-Education-Adjusted Estimates Shown Separately. All Adjustments According to the Distribution in the Reference Area in 1986 Intervention Area Reference Area Estimated Estimated Estimated Risk Estimated Estimated Estimated Risk Risk Risk Risk Difference Risk Risk Risk Difference 1985–87 (I) 1988–91 (II) 1992–94 (III) I–III (%) 1986 (I) 1990 (II) 1994 (III) I–III (%)
2/25/10
Table 21-6
46894_CH21_619_656.qxd Page 633
EXPERIENCE WITH MULTIFACTOR PRIMARY PREVENTION 633
46894_CH21_619_656.qxd
634
2/25/10
4:16 PM
Page 634
CHAPTER 21 THE CASE FOR PREVENTION
high CVD risk), and one isolated community serving as control.37 Extensive intervention included many of the features described for the North Karelia and United States studies, described previously. Evaluation of the program was performed through pre- and postintervention surveys bracketing a 4-year intervention period. Significant net reductions were found in blood pressure, smoking, and a simple risk score based on the observed ranks in levels of risk factors— systolic blood pressure, total cholesterol/HDLcholesterol ratio, and tobacco consumption. Little difference was observed between outcomes of lowand high-intensity intervention, suggesting that the low-intensity approach was nearly as effective as the alternative, at one-fourth the cost. Intervention in Youth In addition to trials of interventions to reduce single risk factors during childhood and adolescence, addressed in Part III as preceded, multifactor interventions have been conducted in school-age populations since the 1980s. An early example of such a study is the evaluation of the Know Your Body Program, a school-based health education program for children at several grade levels.44–46 This program was in part a major descriptive epidemiologic study based on school examinations conducted in several countries. In addition, random assignment among six New York school districts, children in fourth grade (approximately 9 years old) in 11 schools received a curriculum beginning in grade four and continuing through grade nine, addressing nutrition, physical fitness, and cigarette smoking prevention. Those in 11 other schools served as controls. The impact of intervention was evaluated in two settings in or near New York City at 1, 3, and 5 years from initiation. After 1 year, diastolic blood pressure and serum thiocyanate (a marker of smoking) were significantly lower in intervention than control schools, although ponderosity index (an index of overweight or obesity) was greater.44 At 3 years, or at average ages of approximately 12 years, further evaluation was restricted to four of the six districts, as two had withdrawn because the curriculum was judged to require too much class time.45 Diastolic blood pressure and plasma total cholesterol concentration were significantly reduced in intervention relative to control schools, and trends for dietary practices and cigarette smoking were also favorable. At the 5-year evaluation, students in 22 elementary schools in Bronx, New York, who had been in the fourth grade in 1980, were added.46 Cholesterol concentration was reduced further in intervention than control
schools in Westchester but not in the Bronx. Trends were favorable for dietary intake of fat and knowledge of the risk factors in both study areas. It was concluded that educational programs could have favorable, though small, effects on cholesterol concentration in children and that there were no effects on blood pressure, body mass, or physical fitness. On the basis of the Know Your Body experience and others through the 1980s, a major new multicenter school-based intervention program was undertaken in the United States, known initially as the Children’s Activity Trial for Cardiovascular Health, more recently as the Coordinated Approach to Child Health (CATCH, in either case).47 Baseline assessment in 1991 was followed by 3 years of intervention and postintervention assessment in 1994. Participants were 5106 students in third grade at entry who attended one of 56 intervention or 40 control schools in California, Louisiana, Minnesota, and Texas. Intervention was designed to improve school food service and physical education programs and included classroom health curricula; half of the intervention schools also received a family education program. Outcomes were evaluated at two levels: at the school level, change in fat content of school lunches and the amount of moderate-to-vigorous physical activity in physical education programs, and at the individual level, dietary and activity patterns and change in serum cholesterol concentration. The results at the school level were significantly favorable. The relative decrease in fat content of lunch menus is shown in Table 21-7. Moderate-to-vigorous physical activity was also improved significantly in intervention as compared with control schools. Changes in self-reported dietary and activity behaviors were also significantly more favorable among intervention than control children, but changes in cholesterol concentration and other physiologic characteristics did not differ between groups of children. It was judged that the dietary changes were insufficient to produce a detectable change in cholesterol concentration, especially in view of pubertal influences on total cholesterol concentration. However, ability to modify the school environment and curriculum in favorable ways appeared to be well established. The CATCH program has become widely disseminated in the United States and now reaches several other countries as well. Some Implications of Experience with Multifactor Intervention A much wider representation of experience in community intervention can be found in Worldwide
46894_CH21_619_656.qxd
2/25/10
4:16 PM
Page 635
EXPERIENCE WITH MULTIFACTOR PRIMARY PREVENTION 635
Table 21-7
School Lunch Menu Analysis at Three Time Points, 1991 Through 1994, Child and Adolescent Trial for Cardiovascular Health Intervention (I) and Control (C) Groups Value Measured Group Baseline Interim Follow-Up Total energy content, MJ C 2.97 (0.04) 3.04 (0.04) 3.12 (0.04) I 3.01 (0.04) 2.93 (0.04) 2.86 (0.04) p .05 p .001 Energy from total fat, % C 38.9 (0.5) 36.2 (0.5) 36.2 (0.5) I 38.7 (0.4) 32.5 (0.4) 31.9 (0.4) p .001 p .001 Energy from saturated fat, % C 15.1 (0.3) 13.6 (0.3) 13.7 (0.3) I 14.8 (0.2) 12.1 (0.2) 12.0 (0.2) p .02 p .007 Cholesterol content, mg C 80.3 (2.4) 75.2 (2.4) 83.2 (2.4) I 77.7 (2.0) 72.3 (2.0) 74.9 (2.0) p .95 p .17 Sodium content, mg/MJ C 386 (7) 415 (7) 421 (7) I 377 (6) 401 (6) 423 (6) p .64 p .34 Potassium content, mg/MJ C 325 (5) 333 (5) 327 (5) I 331 (4) 350 (4) 357 (4) p .18 p .004 Note: Data for baseline, interim, and follow-up are adjusted means (SE) from repeated-measures analysis of variance, adjusted for site and school random effect. P values compare C with I, adjusting for baseline difference. The school family intervention group did not differ from the school-only group for any endpoint (P .20). 1 MJ 239 kcal. 1 mg/MJ 4.184 mg/1000 kcal. Source: Reprinted with permission from RV Luepker et al., Journal of the American Medical Association, Vol 275, No 10, p 772, © 1996, American Medical Association.
Efforts to Improve Heart Health—the supplement to the report of the Second International Heart Health Conference, The Catalonia Declaration. Investing in Heart Health.48 This resource identified more than 75 heart health programs in Africa, the Americas, Asia, Australia, and Europe, 25 of them being multifactor community programs. References are provided for each program and contact information is given for members of each of six international Heart Health Networks, including the extensive Countrywide Integrated Noncommunicable Diseases Intervention (CINDI) Network. The outcomes of intervention to prevent CHD in various settings (geographic communities, worksites, schools, religious organizations, and medical care organizations), although mixed, have identified components of intervention that appear to have contributed importantly to reductions in risk. The morbidity and mortality outcomes that were positive were taken as reinforcing the possibility of benefit from intervention, but these studies have often presented challenges in design, conduct, and analysis. Statistical power may be seriously limited by study of only a small number of units of observation in the usual community comparison design, and strong, sometimes overriding, favorable secular trends in risk factors and event rates in control communities.
The expectations of favorable outcomes that would demonstrate reduced mortality were high as the formal trials were initiated. For those who were most optimistic about CVD prevention, the positive results stood out regarding both disease outcomes and intermediate behavioral or risk-factor effects. Findings after extended postintervention follow-up in some trials support the view, suggested in many reports, that appropriate evaluation requires longer than usual observation and perhaps more sustained and broadbased intervention than was planned at the outset of these studies. One commentary on “great expectations” contrasted the worlds of community interventions and advertising.49 The implication of that essay was that the likely impact of some public health interventions may be important but too small, at least for the relative investment made and under the conditions of study, to meet conventional criteria for statistical significance. But in the context of advertising, small effects, even if achieved through sometimes very large investments, are considered highly successful. The advice was not to abandon appropriate statistical standards but to ensure that assumptions about effect sizes and other design considerations are more realistic, and therefore more conservative. For example, much larger sample sizes should be provided.
46894_CH21_619_656.qxd
636
2/25/10
4:16 PM
Page 636
CHAPTER 21 THE CASE FOR PREVENTION
These and other issues in design and interpretation of community intervention trials were reviewed in the mid-1990s to update a 1978 symposium on community trials in CHD.50 The theme was that the preceding 15 years or more of experience provided a solid foundation for future community trials. Another review based on the Stanford, Minnesota, and Pawtucket studies proposed broader intervention approaches to include public policy initiatives.51 Complementary small-scale studies would target especially low-income groups and others not yet reached effectively by preventive measures. Outcomes of value for further research would include not only biomedical but also behavioral and attitudinal changes and qualitative assessments of programs. In his essay on the “tribulations of trials,” Susser considered community trials in the context of social movements and policy change.52 His conclusions suggested that the scope of experimental evaluation is unlikely ever to match the larger forces operating on the phenomena of interest. Understanding the social and policy milieu of communities in which interventions are to be evaluated is essential for appropriate choices of time, place, and characteristics of the target population, intervention strategies, and outcomes for evaluation. Overall, experience with multifactor primary prevention has accrued from a large number of studies in the United States and other countries. Widespread interest especially in community approaches suggests an increasing level of readiness over the past decade to take further action. Lessons of experience indicate foremost a need to implement the most promising and comprehensive interventions, in multiple populations, on a large enough scale and with sufficient duration to permit rigorous evaluation. This would offer the greatest opportunity to identify intervention approaches with potential for widespread dissemination and adaptation to local needs and resources. In his personal synthesis of this experience, Farquhar concluded that behavior can be influenced by community intervention; research is needed on options in the mix and sequencing of components in an intervention program; and environmental change (in the sense of the policy setting in which intervention occurs) must be incorporated for effective community change. Further, “The greatest unmet need is now for dissemination research and concurrent international technology transfer of the vast number of lessons learned and the many widely accepted practice principles of community-based interventions” (J Farquhar, personal communication, 1997).
THE BURDEN OF RISK Global Presence of Increased CVD Risk A vast body of evidence identifies the main determinants of CVD, including the major established risk factors, as presented in Part III. Essential to the case for CVD prevention is the knowledge that these risks are widespread on a national and global scale; they themselves constitute a public health challenge of huge proportions; and they underlie large disparities in health within and among populations. Further, individuals whose risk remains low experience better health throughout adult life, and this more favorable course is becoming well documented. Together, this knowledge from epidemiologic experience strongly supports public health strategies for both prevention of risk in the first place and reduction of risk when this becomes necessary. To demonstrate the magnitude and variation in CHD risk among adults in the United States, Ford and others presented estimates of 10-year risk for CHD based on data from the Third National Health and Nutrition Examination Survey, 1988–1994 (NHANES III) (Table 21-8).53 The risk model of the National Cholesterol Education Program’s Adult Treatment Panel III guidelines was applied to riskfactor assessments for more than 11,000 persons aged 20–79 years examined in NHANES III. Persons with CHD or a CHD risk equivalent (diabetes, peripheral vascular disease, or stroke) were excluded from analysis. Risk factors included in the model were age, total cholesterol, HDL cholesterol, systolic blood pressure, hypertension treatment, and current smoking. The outcome was CHD defined as self-reported heart attack or angina pectoris. Risks of 10% or greater of having a first CHD event within 10 years were present in 18.4% of the total population. (The proportion was 27.4% if those with CHD or CHD equivalent were included.) When extrapolated from the survey sample to the total population, 18.4% corresponds to 27 million persons. These risks increased sharply with age but were already greater than 10% of persons aged 40–49 years, increasing to more than 29% at age 50–59 years, more than 45% at age 60–69 years, and more than 65% at age 70–79 years. Risks were greater for men than for women by several times at all but the oldest age (only by two times at age 70–79 years) and differed significantly by race or ethnicity among women, being greater for African Americans and Mexican Americans than for Whites. Global Burden of Disease and Risk Factors drew upon several thousand data sources worldwide to
46894_CH21_619_656.qxd
2/25/10
4:16 PM
Page 637
THE BURDEN OF RISK 637
Table 21-8
Age-Specific and Age-Adjusted Distribution of Risk for Coronary Heart Disease Among United States Adults Age 20 Years Without Self-Reported Coronary Heart Disease or a Coronary Heart Disease Equivalent,* National Health and Nutrition Examination Survey III, 1988 to 1994 10-Year Risk for Coronary Heart Disease 10% to 20%, 10%, % (SE) 20%, % (SE) n Weighted, n % (SE) p† 11,611 138,463,213 81.7(0.5) 15.5 (0.5) 2.9 (0.2) — 0.001 5,481 67,159,469 66.0 (0.6) 28.7 (0.8) 5.3 (0.4) 6,130 71,303,744 94.8 (0.3) 4.3 (0.3) 0.9 (0.1)
Total Gender Men Women Age (yrs) Total 20–29 30–39 40–49 50–59 60–69 70–79 Men 20–29 30–39 40–49 50–59 60–69 70–79 Women 20–29 30–39 40–49 50–59 60–69 70–79 Race or ethnicity Total White African American Mexican American Other Men White African American Mexican American Other Women White African American Mexican American Other
0.001 2,951 2,778 2,022 1,345 1,486 1,029
34,410,851 36,429,134 27,827,102 17,152,023 13,884,733 8,759,370
99.9 (0.1) 95.7 (0.8) 89.7 (0.8) 70.9 (1.7) 55.0 (1.2) 34.5 (1.6)
0.1 (0.1)‡ 3.7 (0.8) 8.5 (0.8) 25.3 (1.7) 40.1 (1.3) 51.5 (1.5)
0.0 (0.0) 0.6 (0.2)‡ 1.8 (0.5) 3.8 (0.6) 5.0 (0.6) 14.0 (1.4)
1,388 1,256 988 614 767 468
17,394,942 18,193,189 13,843,244 8,120,105 6,101,047 3,506,943
99.8 (0.2) 91.4 (1.5) 80.2 (1.7) 40.6 (2.4) 8.4 (1.4) 2.5 (0.7)
0.2 (0.2)‡ 7.4 (1.6) 16.2 (1.6) 52.0 (2.6) 80.8 (1.7) 75.5 (2.4)
0.0 (0.0) 1.2 (0.4) 3.6 (0.9) 7.4 (1.3) 10.8 (1.4) 22.0 (2.5)
1,563 1,522 1,034 731 719 561
17,015,909 18,235,945 13,983,858 9,031,919 7,783,687 5,252,426
100.0 (0.0) 99.9 (0.1) 99.1 (0.4) 98.2 (0.6) 91.5 (1.2) 55.9 (2.5)
0.0 (0.0) 0.1 (0.1)† 0.8 (0.4)† 1.4 (0.5)† 8.2 (1.2) 35.5 (2.1)
0.0 (0.0) 0.0 (0.0) 0.1 (0.0)‡ 0.4 (0.2)‡ 0.3 (0.2)‡ 8.6 (1.5)
4,573 3,290 3,249 499
106,562,092 13,978,761 6,937,452 10,984,907
81.6 (0.5) 80.5 (0.6) 80.6 (0.8) 83.1 (1.4)
15.6(0.5) 16.1 (0.6) 16.7 (0.7) 14.0 (2.0)
2.8 (0.3) 3.5 (0.3) 2.8 (0.3) 2.9 (1.2)‡
2,077 1,527 1,658 219
51,612,956 6,480,578 3,727,827 5,338,108
65.6 (0.8) 66.0 (1.3) 67.0 (1.1) 69.7 (2.4)
29.2 (0.8) 28.2 (1.2) 28.2 (1.2) 24.0 (3.6)
5.2 (0.4) 5.8 (0.6) 4.8 (0.5) 6.3 (2.7)‡
2,496 1,763 1,591 280
54,949,136 7,498,183 3,209,625 5,646,799
95.1 (0.3) 92.7 (0.6) 94.3 (0.7) 94.5 (1.1)
4.1 (0.3) 5.8 (0.6) 5.1 (0.7) 5.2 (1.1)
0.8 (0.1) 1.5 (0.3) 0.6 (0.3)‡ 0.3 (0.2)‡
0.001
0.001
0.289
0.275
0.002
*Self-reported myocardial infarction, angina pectoris, history of stroke, peripheral vascular disease, or diabetes mellitus; †p value is for chi-square test for unadjusted results and for Cochran-Mantel-Haenszel chi-square test for age-adjusted data; ‡estimate may be unstable and should be interpreted cautiously. Source: Reprinted with permission from ES Ford, WH Giles, AH Mokdad, Journal of the American College of Cardiology, Vol 43, © 2004 American College of Cardiology Foundation, p 1793.
46894_CH21_619_656.qxd
638
2/25/10
4:16 PM
Page 638
CHAPTER 21 THE CASE FOR PREVENTION
assess the contribution of common risk factors to death and disability as of 2001.54 Several of these findings are discussed in Part III. Taking together the category of “nutrition-related” risk factors (high blood pressure, high cholesterol, overweight and obesity, and low fruit and vegetable intake) and physical activity, and adding smoking, some 23 million deaths and 320 million disability-adjusted life years lost (DALYs) were attributed to these factors worldwide— 5 million deaths and 62 million DALYs in high-income and 18 million deaths and 263 million DALYs in low- and middle-income countries. For putting the distribution of estimated CVD risk in this global perspective, risk prediction charts from the WHO provide risk estimates for 14 subregions of the world defined by geography and pattern of mortality (A, very low child, very low adult mortality; B, low child, low adult mortality; C, low child, high adult mortality; D, high child, high adult mortality; and E, high child, very high adult mortality).55,56 Factors included in the risk models used were age, sex, smoking, blood pressure, blood cholesterol, and presence of diabetes. The CVD outcome to be predicted comprised acute coronary, cerebrovascular, or peripheral vascular events. Results are summarized in Table 21-9 for one selected age group of men and women across these regions. For reference, illustrative countries in Americas A, Europe A, and Western Pacific A, respectively, are the United States, the United Kingdom, and Japan. The estimates indicate the worldwide presence of risks of these CVD outcomes within 10 years of
Table 21-9
Proportion of the Population at 10% or Greater Total CVD Risk at Ages 50–59 Years, by Sex, by WHO Subregion Subregion Men (%) Women (%) Africa D 13.8 16.7 Africa E 13.6 16.7 Americas A 30.7 13.9 Americas B 22.8 14.4 Americas D 10.1 9.9 Eastern Mediterranean B 16.1 18.3 Eastern Mediterranean D 17.9 15.6 Europe A 17.4 5.5 Europe B 23.5 16.1 Europe C 30.3 20.9 South-East Asia B 14.2 10.6 South-East Asia C 17.0 15.4 Western Pacific A 16.1 8.1 Western Pacific B 16.0 8.6 Source: Data from World Health Organization, Prevention of Cardiovascular Disease: Guidelines for Assessment and Management of Total Cardiovascular Risk, © World Health Organization pp 75–78.
10% or more for 10 to 30% of men and 5 to 20% of women at ages 50–59 years. Proportions at this level of risk are greater for men than for women in most, but not all populations. For men they are exceptionally high in Americas A and in Europe C (including the Russian Federation and other Central European countries and Central Asian Republics), where they are also high in women. Proportions are notably lower for men in Africa and for women in the Western Pacific as well as Americas D (certain countries in Central and South America) and Europe A. Differences in results for the United States in these two analyses at age 50–59 years are striking—10% or greater risks of CHD (heart attack or angina) were 59.4% for men and 1.8% for women in the first report. In the WHO report, corresponding risks of combined CVD outcomes were 30.7% for men and 13.9% for women. The difference in sex ratios of these outcomes (30-fold in one report and 2-fold in the other) is especially puzzling, even recognizing major differences in methods; reasons are not clear from information in the reports. The WHO report, based on one consistent approach across all populations, is especially valuable in permitting geographic comparisons and demonstrating the near ubiquitousness of risk. Discussion of the INTERHEART Study in Chapter 4 indicated the consistent association of several major risk factors with presence of CHD in every population in this multinational case control study (see Table 4-10).57 Even in Africa, where the proportion of men at increased risk appears lower than elsewhere in the WHO analysis, odds ratios for current/ former smoking, diabetes, hypertension, abdominal obesity, and elevated ApoB/ApoA-1 ratio were significant for Black, Colored, European, and other Africans with the sole exception of smoking for Blacks.58 The contribution of these factors to epidemic CVD, as measured by the corresponding population-attributable fractions, is virtually universal as is their importance for CVD prevention. As described by Rodgers and others:59, p 851 High blood pressure, cholesterol, and bodyweight are responsible for a large and increasing proportion of the global burden of disease. Although historically these risks have been regarded as “Western,” their impact is now recognized as global: they are already leading causes of disease in middle-income countries and of emerging importance in low-income countries. It has been noted earlier that what is now “emerging” is appreciation of longstanding developments in low- and middle-income countries, where death rates from CHD and stroke were already substantial, ac-
46894_CH21_619_656.qxd
2/25/10
4:16 PM
Page 639
THE BURDEN OF RISK 639
cording to World Bank estimates for 1985; the impact of the major risk factors throughout the world has in fact been evident for more than two decades (see Chapter 1). The Importance of Low Risk In a number of reports, beginning in 1999, Stamler and colleagues have presented data on long-term health experience of persons whose major risk factors were assessed in middle age or earlier.60 Table 21-10 presents data from two cohorts: first, men aged 35–39 and 40–59 years screened for the Multiple Risk Factor Intervention Trial (MRFIT), and, second, men aged 18–39 years and both men and women aged 40–59 years at examination during the Chicago Heart Association Detection Project in Industry (CHA). “Low risk” was defined as serum cholesterol below 200 mg/dl, blood pressure 120/80 mm Hg, and not currently smoking. Persons with a history of diabetes or myocardial infarction were excluded; in three of the five cohorts, those with electrocardiographic abnormalities were excluded as well. The prevalence of low risk was low, from 5% to 10% across the several cohorts. Death rates (per 100,000) for CHD and for all CVD among those at low risk relative to others in the corresponding cohorts were determined after 16 and 22 years of follow-up in MRFIT and CHA, respectively. Being at low risk at the indicated ages was associated with only 8% to 23% of CHD mortality
and 15% to 28% of CVD mortality experienced by the remainder of the study population. The authors concluded that long-term reduction in mortality and increased life expectancy could be achieved for many more people if prevalence of low risk were increased. Prevalence of risk factors is low in early childhood and increases by early adulthood, continuing generally to do so throughout most of the remaining years of life. Preserving the low-risk that is characteristic of early childhood would contribute importantly to achieving greater prevalence of low risk in the early or middle adult years as found in the experience of MRFIT and CHA follow-up. Several further contributions to this topic have indicated that low risk at middle age is associated with favorable health-related quality of life at older ages and lower healthcare expenditures in later years, including the last year of life.61–62 Low risk is less prevalent in African Americans than Whites in the United States, and this difference in risk distribution accounts for the excess CVD incidence in African Americans.63 The new emphasis on low risk underscores the reality for many people, and the potential for many more, to preserve optimum cardiovascular health throughout their lives. Analysis of long-term outcomes among Framingham Heart Study participants with two or more examinations between ages 40 and 50 showed increased longevity and survival free of major co-morbid conditions up to age 85 on the basis of favorable risk-factor levels at middle age.64 In
Table 21-10
Mortality from Coronary Heart Disease and All Cardiovascular Diseases for Low-Risk Subcohorts and Others* Age-Adjusted RR (95% CI), Low-Risk Cohort† No. Low-Risk Subcohort Others Subcohorts vs Others Coronary Heart Disease Mortality‡ MRFIT men aged 35–39 y 72,144 11 (0.2) 735 (1.5) 0.14 (0.08–0.25) CHA men aged 18–39 y 10,025 1 (0.6) 126 (5.9) 0.08 (0.01–0.61) MRFIT men aged 40–57 y 270,671 126 (4.4) 9578 (19.9) 0.22 (0.18–0.26) CHA men aged 40–59 y 7490 6 (8.8) 516 (38.1) 0.23 (0.10–0.51) CHA women aged 40–59 y 6229 2 (3.5) 181 (14.5) 0.21 (0.05–0.84) MRFIT men aged 35–39 y CHA men aged 18–39 y MRFIT men aged 40–57 y CHA men aged 40–59 y CHA women aged 40–59 y
72,144 10,025 270,671 7490 6229
All Cardiovascular Disease Mortality‡ 16 (0.3) 1022 (2.1) 3 (1.4) 163 (7.7) 190 (6.7) 13247 (27.5) 10 (15.8) 714 (53.1) 4 (5.3) 281 (22.6)
0.15 (0.09–0.24) 0.20 (0.06–0.62) 0.24 (0.21–0.28) 0.28 (0.15–0.52) 0.27 (0.10–0.72)
*MRFIT indicates the Multiple Risk Factor Intervention Trial; CHA, Chicago Heart Association Detection Project in Industry; RR, relative risk; and CI, confidence interval. † Ages are baseline ages; follow-up averaged 16 years in the MRFIT study and 22 years in the CHA study. ‡ Data presented as no. of deaths (age-adjusted mortality rate per 10 000 person-years). Source: Reprinted with permission from Journal of the American Medical Association, Vol 282, J Stamler, R Stamler, JD Neaton, et al., p 2015, © 1999 American Medical Association.
46894_CH21_619_656.qxd
640
2/25/10
4:16 PM
Page 640
CHAPTER 21 THE CASE FOR PREVENTION
a further development of analytic methods, lifetime risk of CVD among Framingham participants up to age 95 years was greatest for persons with no risk factors, and increased with each additional risk factor, at age 50.65 Public health strategies to increase prevalence of low risk through middle age are strongly supported on the basis of such evidence. The burden of risk is global in extent, and no region of the world is free of CVD on an epidemic scale. Distributions of particular risk factors vary among populations as do, therefore, specific populationattributable fractions for CHD and stroke. But the same factors are accountable everywhere, at populationspecific levels, as demonstrated by the INTERHEART Study.57
ECONOMIC CONSIDERATIONS The cost of health care is a prominent concern at present in many, perhaps all, countries. Discussion of economics in connection with health is often limited to the cost of care and its implications. The United States is exceptional for its high rate of expenditure for health care. The costs of CVD in the United States were projected for 2008 as shown in Table 21-11.66 The total direct costs, inclusive of all services for all
Table 21-11
CVD, were estimated to be $296.4 billion, which is nearly $1000 per person per year for the entire population of slightly more than 300 million people. These costs are attributed separately to each of several (nonexclusive) categories of CVD. How to pay for needed care at the individual level is a prominent social and political issue at local, state, and national levels. The implications of cost for healthcare insurers is also a major concern.67 Further, the United States, especially in comparison with other countries, is not achieving a level of health commensurate with this extraordinary spending on health care. This observation adds further to the economic concern.68 This point is emphasized in an analysis of health system impact on “amenable mortality”—deaths attributed to conditions considered preventable or treatable through health care, including half of CHD deaths—among the United States and 18 peer countries.69 From 1997–1998 to 2002, CHD death rates declined in every country, by large margins in several of them; the decline was less for the United States than for all but two of these countries (Greece and Japan); and the rate was higher at the end of this period than in all but two others (Finland and Ireland). This was despite the United States being “the most prolific health spender.”69, p 58
Estimated Direct and Indirect Costs (in Billions of Dollars) of CVD and Stroke: United States: 2008 Hypertensive Heart Disease* CHD Stroke Disease HF Total CVD†
Direct costs Hospital Nursing home Physicians/other professionals Drugs/other Medical durables Home health care Total expenditures† Indirect costs Lost productivity/morbidity Lost productivity/mortality‡ Grand totals†
$99.3 $22.7 $22.8
$51.0 $11.9 $12.9
$18.9 $15.7 $3.6
$7.6 $4.6 $12.8
$18.8 $4.3 $2.3
$140.1 $46.6 $44.4
$21.0 $7.0 $172.8
$9.7 $2.1 $87.6
$1.3 $4.2 $43.7
$24.1 $2.2 $51.3
$3.1 $3.2 $31.7
$49.5 $15.8 $296.4
$23.1 $91.4 $287.3
$10.2 $58.6 $156.4
$6.7 $15.1 $65.5
$8.1 $10.0 $69.4
... $3.1 $34.8
$37.6 $114.5 $448.5
Ellipses (. . .) indicate data not available. *This category includes CHD, HF, part of hypertensive disease, cardiac dysrhythmias, rheumatic heart disease, cardiomyopathy, pulmonary heart disease, and other or ill-defined “heart” diseases. † Totals do not add up because of rounding and overlap. ‡ Lost future earnings of persons who will die in 2008, discounted at 3%. Sources: Direct costs: Extrapolation from 1995 cost estimates for CVD in Hodgson and Cohen1 to the 2008 national health expenditure projections by the Centers for Medicare and Medicaid Services;2 indirect morbidity costs extrapolated to 2008 from indirect cost estimates by disease in 1980 by Rice et al3 after application of a 1980 to 2008 inflation factor computed from mean earnings published by the US Census Bureau;4 indirect mortality costs estimated by multiplying the numbers of deaths by age, sex, and cause in 20045 (NCHS mortality statistics) times estimates of the present value of lifetime earnings for 2003 by age and sex (unpublished estimates) furnished by Rice, Max, Michel, and Sung (University of California, San Francisco, 2007). All estimates prepared by Thomas Thom, NHLBI. Reprinted with permission from American Heart Association. Heart Disease and Stroke Statistics—2008 Update. Dallas, Texas: American Heart Association; 2008. ©2008, American Heart Association.
46894_CH21_619_656.qxd
2/25/10
4:16 PM
Page 641
ECONOMIC CONSIDERATIONS 641
But the remainder of the table points to another aspect of economics of CVD—indirect costs due to lost productivity by those afflicted by these conditions. This is a result of disability or death within the working years. These costs add another $150 billion— or another $500 per person per year for the whole population. Although these numbers are especially large for the United States, the global dimensions of CVD make the economics of CVD and of chronic diseases more generally important everywhere, as suggested by Ebrahim:70, p 225 . . . the burden of disease attributable to chronic diseases is dominant and rising in low- and middle-income countries; these diseases hit younger people at the peak of their productive lives; health systems cannot be made for each disease but need to be integrated to meet all health needs; and, finally, there are cost-effective interventions available that, if implemented, would save lives, reduce suffering and poverty. Aspects of economics addressed by Ebrahim extend far beyond the cost of care and include the impact of CVD and other chronic diseases on economic productivity and development; adequacy of health systems to provide affordable, accessible, high-quality care; and cost-effectiveness of available population-wide interventions that could have broad societal impact. Evidence presented throughout this book indicates that developed or high-income countries must also confront these issues. The broader economic considerations regarding CVD prevention include the following points, presented in the 2007 series of reports by The Lancet on chronic diseases in low- and middle-income countries: In the 23 countries with the most reliable data, chronic diseases (predominantly CVD, diabetes, cancer, and chronic respiratory diseases) accounted for 50% of the total disease burden in 2005. Heart disease, stroke, and diabetes alone were projected to cost these countries $83 billion in lost economic productivity between 2006 and 2015. A newly articulated goal for global health is to accelerate reduction in chronic disease mortality by 2% per year in these countries, through available interventions over these same 10 years. Success in this effort would avert 24 million deaths and reduce economic loss by $8 billion.71 A further point favoring investment in prevention of cardiovascular diseases was made in A Race Against Time:72, p 1 “Investments in health not only reduce the burden of disease, but also stimulate economic growth, which in turn raises a society’s ability to invest in public health.” At another level, cost-effectiveness analysis of particular clinical and public health interventions is
used widely as an aid to decision making in allocating resources. The principles and methods of this approach are presented in detail elsewhere, as is a review of economic evaluations specific to primary prevention of CVD through 2005.73,74 That review found that few cost-effectiveness evaluations had extended beyond clinical studies of lipid-lowering drugs to address broader health promotion strategies; most were conducted in the United States or the United Kingdom; sponsorship was needed from government and not only from the drug industry for a more balanced research agenda; and evaluation methods tended to bias results in favor of treatment over prevention. In the interest of increasing access to methods of cost-effectiveness analysis for public health decision making, WHO has developed and utilized the WHOCHOICE (Choosing Interventions that are CostEffective) project.75 Rather than focusing on “technical efficiency,” regarding a particular intervention alone, WHO-CHOICE addresses “allocative efficiency,” or alternative uses of resources given a wide range of potentially available interventions benefitting different segments of the population. Cost-effectiveness has also been used to assess feasibility of certain interventions for primary prevention of chronic diseases, including CVD. Resulting evidence has been judged to support bringing three population-wide interventions to scale on a national level: price increases for tobacco products, reduction of salt intake, and a low-cost multidrug regimen for people at high risk of CVD ( 25% or 15% in 10 years).36 The first two interventions could avert 13.8 million deaths in these 23 countries over 10 years, at costs from less than US$0.40 to $1.00 per capita per year.76 The corresponding costs for a drug regimen comprising a statin, aspirin, and two blood pressurelowering medications for the benefit of a high-risk subgroup of the population would range from about US$0.40–$2.90 per capita per year across the 23 countries. This intervention was projected to avert 17.9 million deaths over the 10 years, or three-fourths of the goal of a 2% per year reduction in chronic disease mortality “with a moderate increase in health expenditure.”77, p 2054 (The latter point that projected savings will require an increase in initial investment is a key element in the economics of CVD prevention.) At a macroeconomic level, CVD and other chronic diseases demand a level of attention and urgency of action that have been seriously underappreciated until quite recently. The economic and social impact of lost productivity, especially—but not exclusively—in lowand middle-income countries, would seem to compel action, the cost of continued inaction being unacceptable. On the basis of cost-effectiveness analysis of
46894_CH21_619_656.qxd
642
2/25/10
4:16 PM
Page 642
CHAPTER 21 THE CASE FOR PREVENTION
available interventions, substantial progress could be made by implementing largely affordable preventive measures today.
MODELS FOR EXPLANATION AND PREDICTION Models extend information beyond direct observation to complex mathematical representations that can be explored in various ways. This permits analysis of the influence of multiple factors operating simultaneously on an outcome of interest. Also, the parameters of the model derived from one period of observation can be used to project the future course of events under varying assumptions about influential conditions. These uses of modeling are distinct in purpose from those for estimating absolute CVD risk, or evaluating cost-effectiveness of interventions. As discussed in Chapter 4, the unanticipated downturn in age-standardized CHD mortality in the United States that was recognized in the late 1970s stimulated major research projects on trends in occurrence of CVD. Much of this work was simply to compile the available observations for descriptive purposes or to implement new surveillance activities for better documentation of contemporary events. The most far-reaching initiative in this area was the WHO MONICA Project, discussed in Chapter 4 and elsewhere. Following up on the Decline Conference of 1978, the National Heart, Lung and Blood Institute convened a conference in 1986 on the influence of medical care on CHD mortality trends. The proceedings present an extensive body of work with contributions from both epidemiologists and cardiologists, with the overall conclusion that the downturn was real and continuing, and “the relative roles of diverse causal factors [were] being clarified.”78, p 270 Further insight was to await expanded use of modeling in this area. Explaining Secular Trends and Current Burdens Understanding the contributions of prevention versus treatment, or of behavioral versus medical interventions, was and is a matter of central importance for public health decision making and resource allocation. A still broader scope of analysis today includes influences of policy and systems change. No less important is projection into the future of the likely burden of CVD and other chronic diseases, under varying assumptions. For these purposes, numerous approaches to modeling have been developed. Some of these, including the Global Burden of Disease and IMPACT models discussed in Parts I and II, are described in Table 21-12.79 In their review of “CHD
policy models,” Unal and colleagues found wide variation among more than 40 such models in quality, utility, and other characteristics. They selected six of these as the principal current approaches. The IMPACT model developed by Capewell and others, discussed in Chapter 4, is illustrated in Figure 21-5.80 Results are shown from 10 studies in 6 countries, 5 of them based on the IMPACT model and 5 on other methods. The importance of these findings to the case for CVD prevention is in part the consistent result, in 8 of these 10 studies, that 50% or more of past declines in CHD mortality were attributable to population-wide risk-factor changes. Details of the analysis for the United States over the period 1980–2000 included a net contribution of 44% for risk-factor reduction, given a cumulative positive effect of 61% for reductions in total cholesterol, systolic blood pressure, smoking, and physical inactivity that was offset by 17% because of adverse changes in body mass index and diabetes. In addition to this explanatory insight, the magnitude of decreases in CHD mortality over this period demonstrates forcefully the combined influence of changes in environmental conditions in the broadest sense: age-adjusted CHD mortality per 100,000 declined from 1980 to 2000 from 542.9 to 266.8 for men and from 263.3 to 134.4 for women, resulting in 341,745 fewer CHD deaths in 2000 than if the 1980 rates had persisted. Some qualification of this success is needed, however. An analysis of CHD mortality in the United States from 1980 through 2002 was conducted for specific age-sex strata to identify calendar periods in which trends might have changed for each subgroup.81 For both men and women aged 35–54 years, marked slowing of the decline occurred between 1980–1990 and 1991–2002; for women, the rate appeared to have reversed and increased from 2000–2002. No modeling of determinants of these changes was presented, but knowledge of continuing uptake of effective treatments was taken to support the view that offsetting adverse changes in risk factors were the likely explanation of the overall decline. To characterize the current global burden of disease and risk factors, as of the year 2000, Ezzati and colleagues in the Comparative Risk Assessment Collaborative Group estimated the joint contribution of 20 risk factors to occurrence of 10 leading causes of disease, injuries, and combined causes of death separately for low- and high-mortality developing countries, developed countries, and the world.82 (Analogous estimates of the proportionate contributions of each risk factor to both mortality and burden [in DALYs] of ischemic heart disease and stroke by economic region were presented in Chapter 1 [Table 1-8].)
Smoking, cholesterol, systolic blood pressure
England and Wales, Up to 85 years. Men and women
Micro simulation
CHD Policy Analysis (Sanderson and Davies)
Smoking, cholesterol, hypertension, obesity, physical activity, alcohol Smoking, total cholesterol, DBP, glucose intolerance, age
Netherlands; Denmark, England Depending on the purpose aged 65
Smoking, total cholesterol, DBP and weight to estimate CHD risk using Framingham Equations
Canada, Adult men and women, age group not clear
Cell based
PREVENT (GunningScheppers)
USA, Men and women aged 35–84
Did not consider CHD disease categories but treatments can be considered for primary prevention Angina (stable and unstable). AMI, post MI, CABG, PTCA None
None
Angina, AMI, sudden death, post MI, CABG, PTCA Specific treatments considered in different studies eg statins, aspirin, beta-blockers etc
643
Deaths prevented, morbidity prevented, CHD & non-cardiac deaths, unstable angina admissions, investigations, angiograms, PTCA, CABG
One-way
One way, different scenarios
Number of deaths prevented, life years gained
Years of life saved, cost per life year saved, years of life without CHD symptoms
In the initial model none. Subsequently papers reported one way sensitivity analysis
Sensitivity Analysis
Number of deaths prevented, LYG, CHD incidence (number of arrests, angina, AMI), CHD prevalence, CHD mortality, cost per life year
Outcomes
No validation reported
Calibrated
Not checked
Model was calibrated using 1986 mortality data. Validity: Model estimates were compared with 1990 observed— 92–98% fit reported.
Validation
continues
Separate risk factor and treatment components. Future model may include secondary prevention treatments. No sensitivity analyses yet. Model fit appears better for men than women.
This model uses hypothetical cohorts of participants. In most of the papers, time and the specific population are not clear.
Mainly a primary prevention model. Developed and adopted in several different populations.
First policy model rather basic. Steadily refined since then. Many papers in high impact journals
Strengths and Limitations
4:16 PM
Life table analysis— Markov model from 1998 onwards
State transition Markov Model
CHD Policy Model (Weinstein and Goldman)
Risk Factors Included
Disease Groups & Treatments Included
2/25/10
CHD Life Expectancy Model (Grover et al)
Type of Model
Model Setting & Study Population(s)
Summary of the Six Principal CHD Policy Models
Name of the Model (Author)
Table 21-12
46894_CH21_619_656.qxd Page 643
644 Malnutrition, poor water, unsafe sex, alcohol, tobacco, occupation, hypertension, physical activity, illicit drugs, and air pollution
Initially smoking, cholesterol, blood pressure—then also obesity, diabetes and physical activity and deprivation
None
This model is comprehensive and considers all principal CHD categories and over 20 specific CHD treatments
Disability adjusted life years (DALYs)
Deaths prevented or postponed, life years gained.
Sensitivity Analysis
Multi-way sensitivity analysisdiscounting and age weighting
Multi way sensitivity analysis using analysis of extremes method.
Source: Reprinted with permission B Unal, S Capewell, JA Critchley, Biomed Central Public Health, Vol 6, on the basis of Open Access, © 2006 Unal et al.
World divided into eight geographic regions M-F all ages
Scotland, England & Wales, New Zealand. Initially men and women aged 45–84. IMPACT Model for England and Wales includes 25–84
Outcomes
None
Estimated falls in CHD mortality were compared with observed falls over specific time period stratified by age and sex.
Validation
A comprehensive and global model for WHO strategies. Well documented and described. CHD is included, and modelled as caused by tobacco use, hypertension and physical inactivity, and reduced by alcohol. Data quality. Extremely variable depending on the region
Considers all major effective treatments available for CHD and all major risk factors. Data quality adequate, used trial and meta-analyses: National population statistics and results from representative studies
Strengths and Limitations
4:16 PM
Population attributable risk method
Spread-sheet
IMPACT (Capewell, Critchley and Unal)
Risk Factors Included
Disease Groups & Treatments Included
2/25/10
Global Burden of Disease (Murray and Lopez)
Type of Model
Model Setting & Study Population(s)
Summary of the Six Principal CHD Policy Models—continued
Name of the Model (Author)
Table 21-12
46894_CH21_619_656.qxd Page 644
46894_CH21_619_656.qxd
2/25/10
4:16 PM
Page 645
MODELS FOR EXPLANATION AND PREDICTION 645
Risk Factors
Treatments
Unexplained
United States, 1968–7614
40
54 6
New Zealand, 1974–8115*
40
60
46
Holland, 1978–8517
44 10
43
United States, 1980–9013
50 7
IMPACT Scotland, 1975–9418
35
55 10
IMPACT New Zealand, 1982–9319
35
60 5
IMPACT England and Wales, 1981–200020
37
IMPACT United States, 1980–2000 (this study)
52 11 47
44 9
24
Finland, 1972–9216
23
IMPACT Finland, 1982–9722
0
76 53
50 Decrease in Deaths (%)
24 100
Figure 21-5 Percentage of the Decrease in Deaths from Coronary Heart Disease Attributed to Treatments and Risk-Factor Changes in Our Study Population and in Other Populations. Source: Reprinted with permission from ES Ford, UA Ajani, JB Croft, et al., New England Journal of Medicine, Vol 356, © 2007 Massachusetts Medical Society, p 41.
The report addressed here extended this type of analysis to examine contributions of these risk factors to healthy life expectancy (HALE). The results showed that, beyond preventing CVD or chronic diseases alone, elimination of the 20 risk factors would increase healthy life expectancy globally and reduce differentials across regions and subregions; for example, removing risks associated with alcohol, tobacco, high blood pressure, and high cholesterol would contribute importantly to an 8- to 11-year increase in HALE in much of eastern and central Europe and the former Soviet Union. Of particular interest was development of models to account for the joint effects of coexisting risk factors, by considering separately their direct and mediated effects—for example, the effects of body mass index mediated through blood pressure or other factors as distinct from its direct effects. Effect modification, resulting from influence of one risk factor on another, was also estimated and taken into account. The results regarding the factors associated with ischemic heart disease and stroke for the world as a whole are shown in Table 21-13. Contributions to the global burden of disease attributed to these factors range from 1.3% for physical inactivity to 4.4% for high blood pressure, to which tobacco and alcohol were nearly equivalent.
To put these values in perspective, among all 20 leading risk factors only two accounted for a greater percentage of disease burden than did blood pressure— underweight in children (9.5%) and unsafe sex (6.3%). Even unsafe water, sanitation, and hygiene contributed just 3.7%, less than any one of the three leading CVD risk factors. Further, the definitions of the CVD risk factors should be understood clearly. The “theoretical minimum” for each factor is the reference value against which any higher level is considered to increase risk. For systolic blood pressure, for example, the reference value to identify increased risk is not a clinical cut-point, such as 160, 140, or even 120 mm Hg, but 115 mm Hg. The approach to this definition is described in Global Burden of Disease and Risk Factors as based on “the lowest levels at which meta-analyses of cohort studies have characterized dose-response relationships.”83, p 246 From this perspective, joint population-attributable fractions of the risk factors to worldwide burden and mortality from ischemic heart disease were estimated. The resulting values were 83% and 78%, respectively, after adjustment for mediated effects of the risk factors. Corresponding estimates for stroke were 70% and 60%. That is to say, if these several risks were eliminated, 83% of the CHD burden and 70% of the stroke burden would be averted. (These results
46894_CH21_619_656.qxd
646
2/25/10
4:16 PM
Page 646
CHAPTER 21 THE CASE FOR PREVENTION
Table 21-13
Risk Factors for Ischemic Heart Disease and Stroke: Definitions and Contributions to the Global Burden of Disease, 2000 Contribution to Global Burden of Risk Factor Exposure Variable (See Source for Details) Theoretical Minimum Disease (%) High blood pressure Usual systolic blood pressure 115 mm Hg (SD 6) 4.4 High cholesterol Usual total blood cholesterol 3.8 mmol/L (SD 0.6) 2.8 High BMI BMI (weight over height squared) 21 kg/m2 (SD 1) 2.3 Low fruit and vegetable Fruit and vegetable intake per day 600 g (SD 50)/day for adults 1.8 intake Physical inactivity Three categories of inactive, All having at least 2.5 h/week insufficiently active, and sufficiently of moderate-intensity activity active, taking into account discretionary or equivalent (4000 KJ/week) time, work, transport 1.3 Tobacco Current values of smoking impact ratio; No tobacco use 4.1 prevalence of oral tobacco use Alcohol Current alcohol consumption volumes No alcohol use 4.0 and patterns Source: Data from M Ezzati, S Vender Hoorn, A Rodgers, et al., The Lancet, Vol 362, © 2003, pp 273–274.
were similar but not identical to those shown in Table 1-8.) The general concordance of these and the INTERHEART Study findings is noteworthy. Predicting Future Burdens Before the 1990s, use of modeling for the purpose of predicting the course of the CVD epidemic or of chronic disease burdens was rare, even at a national level. The Global Burden of Disease: A Comprehensive Assessment of Mortality and Disability from Diseases, Injuries, and Risk Factors in 1990 and Projected to 2020, published in 1996, is a landmark.84 Cited widely in current discussions of health policy on global, regional, and national levels, it was not without its limitations as noted in the Foreword by Dr. Ralph H. Henderson, Assistant Director-General of the World Health Organization:85, pp xiii–xiv The findings published in the Global Burden of Disease and Injury Series provide a unique and comprehensive assessment of the health of populations as the world enters the third millennium. We also expect that the methods described in the various volumes in the series will stimulate Member States to improve the functioning and usefulness of their own health information systems. Nevertheless, it must be borne in mind that the results from an undertaking as ambitious as the Global Burden of Disease Study can only be approximate. The reliability of the data for certain diseases, and for some regions, is extremely poor, with only scattered information available in some cases. . . . The concept of the
DALY as used in this Study is still under development, and further work is needed to assess the relevance of the social values that have been incorporated in the calculation of DALYs, as well as their applicability in different sociocultural settings. In this regard, WHO and its partners are continuing their efforts to investigate burden-of-disease measurements and their use in health policy decision-making. Accordingly, an updated assembly of data and application of further-developed methods led to publication of Global Burden of Disease and Risk Factors in 2006, and a further search for data to update the more recent report is already in progress. Global projections from 1990 to 2020 showing ischemic heart disease and stroke to be the first and second leading causes of death called increased attention to the previously under-appreciated burden of CVD throughout the world.86 Whether the baseline or alternative scenarios—optimistic or pessimistic—were considered, CVD remained dominant; only the gross numbers of deaths were influenced by alternative assumptions. Further work in projecting future burdens of CVD and other chronic diseases is illustrated by A Race Against Time: The Challenge of Cardiovascular Disease in Developing Economies (2004) and by Dying Too Young: Addressing Premature Mortality and Ill Health Due to Non-Communicable Diseases and Injuries in the Russian Federation (2005).72,87 The first of these reports, described in Chapter 1, was published in response to the Millennium Development
46894_CH21_619_656.qxd
2/25/10
4:16 PM
Page 647
VISIONS OF SUCCESS IN CVD AND CHRONIC DISEASE PREVENTION 647
Goals, which omitted reference to chronic diseases. Race Against Time projected CVD burdens from 2000 to 2030, from a macroeconomic perspective. It focused on the narrow window of opportunity, within the next two decades or so, “to prevent the precursors and reverse the negative effects of CVD in developing countries” and thereby rescue the labor force—people of working age—from falling victim to CVD in great numbers, thereby crippling economic and social development in such countries as China, Brazil, India, Russia, and South Africa.72, p 84 The second report, in projecting demographic, economic, and health conditions in the Russian Federation, forecast continuation of a national decline already referred to as “devastation” without precedent among industrial nations.87 Population was shrinking, life expectancy was receding, and CVD mortality was the principal cause of death for working men and the population as a whole, with a rate more than three times that of the United States. However, the response to these projections was optimistic:87, p 100 Reducing NCDs and injury-related mortality rates among Russian working-age adults will have a major macroeconomic and poverty reduction impact, regardless of how this is measured. . . . The expected economic benefits are of a magnitude that easily outweighs the costs of health promotion and disease prevention programs. Given the significant positive effect on economic growth from investing in health [citation given], governmental intervention is urgently needed in Russia to develop healthenhancing policies and programs to address the alarmingly high rates among the working-age population. These efforts should be seen as key investments to help improve the general welfare of the population and secure sustainable economic growth in the future. Modeling contributes importantly to explanation, description, and prediction of past, present, and future occurrence of CVD and other chronic diseases. Extending beyond the sometimes quite limited direct observations available, modeling offers insights that can influence decision making about health policy in positive ways. At the same time, it can stimulate continuing efforts to strengthen data sources for future improvement in the models.
VISIONS OF SUCCESS IN CVD AND CHRONIC DISEASE PREVENTION The two reports discussed previously cited opportunities for prevention, and not only threats, in the
projected burdens of CVD and other chronic diseases. Costs of these conditions due to reduced economic productivity in low- and middle-income countries were seen as being so immense that their prevention would more than repay the foreseeable expenditures. Optimism was also evident in the more specific economic analyses regarding cost-effectiveness of feasible interventions, such as were reported in The Lancet’s 2007 series cited previously. In the United States, Europe, and on the global level, many examples can be found of visionary statements of what is achievable through CVD prevention, not only improving health itself but also enhancing societal well-being through greater health equity and social justice. Several examples can serve to illustrate this point: • At a national level in the United States, both federal and voluntary organizations set explicit goals for CVD prevention, decade by decade. The federal Department of Health and Human Services (DHHS) published Healthy People 2010 early in the year 2000 with its heart disease and stroke prevention goal, cited earlier: prevention of risk factors, detection and treatment of risk factors, early identification and treatment of heart attacks and strokes, and prevention of recurrent cardiovascular events.88 At the Centers for Disease Control and Prevention (CDC), the Division for Heart Disease and Stroke Prevention is the administrative locus for leadership in these efforts. The Division’s mission is “To provide public health leadership to improve cardiovascular health for all, reduce the burden, and eliminate disparities associated with heart disease and stroke” and thereby to achieve the vision of a world that is “Heart-Healthy and Stroke-Free.”89 The American Heart Association’s (AHA’s) mission is “Building healthier lives, free of cardiovascular diseases and stroke.”90 Both DHHS and AHA have set targets for reducing the burden of CVD by 2010. DHHS, through Healthy People 2010, provides multiple objectives by which to measure success nationally, and AHA’s goal for 2010 is “to reduce coronary heart disease, stroke and risk by 25%.” • A landmark joint policy statement by the American Cancer Society, American Diabetes Association, and American Heart Association in 2008 declared that optimum delivery of clinical preventive services could reduce myocardial infarction by 63% and strokes by 31%.91 Cost
46894_CH21_619_656.qxd
648
2/25/10
4:16 PM
Page 648
CHAPTER 21 THE CASE FOR PREVENTION
reduction and increased efficiency would be necessary to realize this full impact, however. • Responding to the “challenges to health and equity” confronting European countries, the WHO Regional Office for Europe (WHOEURO) published in 2006 Gaining Health: The European Strategy for the Prevention and Control of Noncommunicable Diseases.92 Underlying the Strategy are the following key statements:92, p 17 Vision—A health-promoting Europe free of preventable noncommunicable disease, premature death and avoidable disability. Goal—To avoid premature death and significantly reduce the disease burden from NCD by taking integrated action, improving the quality of life and making healthy life expectancy more equitable within and between Member States. Objectives—To take integrated action on risk factors and their underlying determinants across sectors. To strengthen health systems for improved prevention and control of NCD. • Preventing Chronic Disease: A Vital Investment: WHO Global Report, supported by the governments of Canada, Norway, and the United Kingdom, was released in 2005.2 The Report “makes the case for urgent action to halt and turn back the growing threat of chronic diseases; presents a state-of-the-art guide to effective and feasible interventions; provides practical suggestions for how countries can implement these interventions to respond successfully to the growing epidemics.”2, p xiv The chronic conditions addressed are heart disease and stroke (CVD), cancer, asthma and chronic obstructive pulmonary disease, and diabetes. The overview of the report succinctly summarizes the argument:2, p 1 The Problem: 80% of chronic disease deaths occur in low and middle income countries and these deaths occur in equal numbers among men and women. The threat is growing––the number of people, families, and communities afflicted is increasing. This growing threat is an under-appreciated cause of poverty and hinders the economic development of many countries.
The Solution: The chronic disease threat can be overcome using existing knowledge. The solutions are effective—and highly cost-effective. Comprehensive and integral action at country level, led by governments, is the means to achieve success. The Goal: An additional 2% reduction in chronic disease death rates worldwide, per year, over the next 10 years. This will prevent 36 million premature deaths by 2015. The scientific knowledge to achieve this goal already exists. • In 1992, the first of seven three-yearly International Heart Health Conferences took place in Victoria, British Columbia, Canada, and concluded with issuance of The Victoria Declaration on Heart Health.93 The document was foremost a call to action and policy framework that addressed action areas, population groups, risk-factor reduction, strategies, developingworld perspectives, a research agenda, and partnerships to include international agencies, governments, and many other players. The Declaration recognized that “both scientific knowledge and widely tested methods exist to prevent most cardiovascular disease” and called upon virtually all segments of society “to join forces in eliminating this modern epidemic by adopting new policies, making regulatory changes and implementing health promotion and disease prevention programs directed at entire populations.”93, p iv Further Declarations from subsequent conferences in the series addressed particular aspects of heart health promotion. The recommendations from the first six reports, issued from 1992 through 2001, were synthesized in the document, International Action on Cardiovascular Disease: A Platform for Success Based on International Cardiovascular Disease (CVD) Declarations, released in 2005. The report concluded:94, p 22 Cardiovascular disease (CVD) is a challenge of global proportions. It is largely preventable. Unfortunately, overall investment in CVD prevention has been insufficient to achieve optimal results. Applying the existing knowledge with the wisdom that has accumulated over the past twenty years could stem the epidemic of CVD around the world. A strong international effort could encourage, promote, and facilitate country-led initiatives.
46894_CH21_619_656.qxd
2/25/10
4:16 PM
Page 649
CURRENT ISSUES 649
International organizations have a critical role to play in harnessing globalization in the service of health. The visions expressed in these several examples represent judgments by prominent organizations and individuals that go beyond the systematic review of evidence on any specific intervention. They reflect not only a sense of what such evidence says but also what it means in terms of societal interests and values. That such belief in the potential for CVD prevention is expressed strongly by many authoritative sources contributes significantly to the case for prevention.
COUNTER-ARGUMENTS Contrary views mainly follow one of three lines of argument: (1) Knowledge of causation of CVD is inadequate, whether because conventional risk factors fail to account for a sufficient proportion of risk; they exclude more fundamental causes, such as social determinants; or they will be superseded by emerging or yet-to-be-discovered factors, including genetics, that will better identify true causation.95–97 (2) Populationlevel interventions to prevent CVD and other chronic diseases are ineffective; they are of insufficient priority to compete for resources against an unfinished agenda of communicable disease and other health needs; or they would widen disparities in health by inequitable impact, favoring the affluent and leaving the poor at even greater relative disadvantage than at present.15,98,99 (3) Further research is needed in order to know what interventions will work, how to tailor them to particular populations, and what they will save in terms of return on investment, prior to taking any public health action.100 A number of published commentaries respond specifically to these views, some of them being cited elsewhere in this or other chapters. Several, for example, address the “myth” that only 50% or less of heart disease and stroke is accounted for by the major established risk factors.100–103 Others emphasize the value of lessons learned from the experience of population-level interventions, including their successes and their limitations, and the need to focus especially on the most disadvantaged populations, eliminating disparities in CVD and other chronic diseases.7,72,104–106 Finally, the judgment that knowledge is sufficient to support local, national, and global public health efforts in this area is reflected in each of the visions highlighted previously and many others, expressed recently as the global challenge of the “know-do gap.”107–108
That counter-arguments regarding the case for CVD prevention continue to be raised should not be surprising, given competing interests, priorities, or interpretations of the evidence. Weighed against the elements of the argument in favor of CVD prevention, however, they are not persuasive to many in positions of accountability for the public’s health.
CURRENT ISSUES Current understanding of the public health importance and feasibility of preventing CVD and other chronic diseases is based on several fundamental considerations: knowledge of their causation; widespread experience in implementation and evaluation of multifactor CVD prevention programs; accumulated evidence of the increasing mortality and burden due to these diseases globally and their especially heavy toll on particular regions, countries, or populations; results of a variety of modeling approaches to explanation and prediction of secular trends in their occurrence and estimation of the fraction of disease attributable to each of several globally operating risk factors; the macroeconomic impact of these diseases on development, affecting the majority of the world’s population; the potential cost-effectiveness of selected policy and health system interventions; and visionary statements over many years by prominent organizations and individuals. Effective communication is called for to express: (1) the urgency of large-scale implementation of policies and programs for prevention; (2) the great potential impact of prevention; and (3) the consequences of failure to take sufficient and timely action in the interest of the public’s health. In the concluding report in The Lancet 2007 series on prevention of chronic diseases, Beaglehole and coauthors summarize findings presented in the series:109, p 2157 This Series of papers in The Lancet provides evidence that achievement of the global goal for the prevention and control of chronic diseases is both possible and realistic through available interventions. These papers lend support to the rapid scaling-up of efforts to prevent and control chronic disease in low-income and middle-income countries. We appreciate the complex nature of the health problems of national authorities in lowincome and middle-income countries because of competing priorities. We are also aware that the evidence for the actual effects of interventions on reducing the burden of disease within
46894_CH21_619_656.qxd
650
2/25/10
4:16 PM
Page 650
CHAPTER 21 THE CASE FOR PREVENTION
countries is more limited than is the broader evidence base for action. However, the totality of the evidence suggests that large economic and health gains can be achieved in low-income and middle-income countries through increased efforts to prevent and control chronic diseases. A current challenge is to take action based on this case for prevention of CVD and other chronic diseases––in the United States, in other high-income countries, and in the developing world.
REFERENCES 1. Committee for the Study of the Future of Public Health, Division of Health Care Services, Institute of Medicine. The Future of Public Health. Washington, DC: National Academy Press; 1988. 2. World Health Organization. Preventing Chronic Diseases: A Vital Investment: WHO Global Report. Geneva: World Health Organization; 2005.
9. The Multiple Risk Factor Intervention Trial Research Group. The Multiple Risk Factor Intervention Trial (MRFIT). Risk factor changes and mortality results. JAMA. 1982;248:1465–1477. 10. The Multiple Risk Factor Intervention Trial Research Group. Mortality rates after 10.5 years for participants in the Multiple Risk Factor Intervention Trial: findings related to a priori hypothesis of the trial. JAMA. 1990; 263:1795–1801. 11. The Multiple Risk Factor Intervention Trial Research Group. Mortality after 16 years for participants randomized to the Multiple Risk Factor Intervention Trial. Circulation. 1996; 94:946–951. 12. Holme I, Hjermann I, Helgelland A, Leren P. The Oslo Study: diet and antismoking advice. Prev Med. 1985;14:279–292.
3. Horton R. Chronic diseases: the case for urgent global action. Lancet. 2007;370:1881–1882.
13. Wilhelmsen L, Berglund D, Elmfeldt D, et al. The multifactor primary prevention trial in Göteborg, Sweden. Eur Heart J. 1986;7: 279–288.
4. Jamison DT, Breman JG, Measham AR, et al., eds. Disease Control Priorities in Developing Countries. 2nd ed. Washington, DC: International Bank for Reconstruction and Development/ The World Bank; 2006.
14. Strandberg TE, Salomaa VV, Naukkarinen VA, et al. Long-term mortality after 5-year multifactorial primary prevention of cardiovascular diseases in middle-aged men. JAMA. 1991; 266:1225–1229.
5. Lopez AD, Mathers CD, Ezzati M, Jamison DT, Murray CJL, eds. Global Burden of Disease and Risk Factors. Washington, DC: The International Bank for Reconstruction and Development/ The World Bank; 2006.
15. Ebrahim S, Beswick A, Burke M, Davey Smith G. Multiple risk factor interventions for primary prevention of coronary heart disease. Cochrane Database Syst Rev. 2006, Issue 4. Art. No.:CD001561. doi: 10.1002/14651858 .CD001561.pub2.
6. Shigan EN. Integrated programme for noncommunicable diseases prevention and control (NCD). World Health Stat Q. 1988;41:267–273. 7. Farquhar JW, Fortmann SP. Community-based health promotion. In: Aherns W, Pigeot I, eds. Handbook of Epidemiology. Berlin: Springer 2005;1305–1323. 8. The Multiple Risk Factor Intervention Trial (MRFIT). A national study of primary prevention of coronary heart disease. JAMA. 1976; 235:825–827
16. Wald NJ, Law MR. A strategy to reduce cardiovascular disease by more than 80%. Brit Med J. 2003;326:1419–1423. 17. Stiles S. Polypill progress: clinical testing imminent for three-agent secondary-prevention pill. Heartwire 2008. Available at www.medscape .com/viewarticle/575014. Accessed 8 May 2008. 18. Ades PA, Kottke TE, Miller NH, McGrath JC, Record NB, Record SS. Task Force #3—getting
46894_CH21_619_656.qxd
2/25/10
4:16 PM
Page 651
REFERENCES 651
results: who, where, and how? J Am Coll Cardiol. 2002;40:615–630. 19. Vartiainen E, Puska P, Jousilahti P, et al. Twenty-year trends in coronary risk factors in North Karelia and in other areas of Finland. Int J Epidemiol. 1994;23:495–504. 20. Puska P, Tuomilehto J, Nissinen A, Vartiainen E, eds. The North Karelia Project. 20 Year Results and Experiences. Helsinki, Finland: National Public Health Institute, KTL; 1995. 21. Vartiainen E, Puska P, Pekkanen J, Tuomilehto J, et al. Changes in risk factors explain changes in mortality from ischaemic heart disease in Finland. Br Med J. 1994;309:23–27. 22. Winkelby MA, Feldman HA, Murray DM. Joint analysis of three US community intervention trials for reduction of cardiovascular disease risk. J Clin Epidemiol. 1997;50:645–658. 23. Farquhar JW, Maccoby N, Wood PD, et al. Community education for cardiovascular health. Lancet. 1977;i:1192–1195. 24. Farquhar JW, Fortmann SP, Flora JA, et al. Effects of community wide education on cardiovascular disease risk factors: the Stanford Five-City Project. J Am Med Assoc. 1990;264: 359–365. 25. Luepker RV, Murray DM, Jacobs DR Jr, et al. Community education for cardiovascular disease prevention: risk factor changes in the Minnesota Heart Health Program. Am J Public Health. 1994;84:1383–1393. 26. Luepker R, Rastam L, Hannan PJ, et al. Community education for cardiovascular disease prevention: morbidity and mortality results from the Minnesota Heart Health Program. Am J Epidemiol. 1996;144:351–362. 27. Carleton RA, Lasater TM, Assaf AR, et al., Pawtucket Heart Health Program Writing Group. The Pawtucket Heart Health Program: community changes in cardiovascular risk factors and projected disease risk. Am J Public Health. 1995;85:777–785. 28. Record NB, Harris DE, Record SS, GilbertArcari J, DeSisto M, Bunnell S. Mortality
impact of an integrated community cardiovascular health program. Am J Prev Med. 2000;19(1):30–38. 29. World Health Organization European Collaborative Group. An international controlled trial in the multifactorial prevention of coronary heart disease. Int J Epidemiol. 1974; 3:219–224. 30. World Health Organization European Collaborative Group. European Collaborative Trial of Multifactorial Prevention of Coronary Heart Disease: final report on the 6-year results. Lancet. 1986;i:869–872. 31. World Health Organization Collaborative Group. European collaborative trial of multifactorial prevention of coronary heart disease. Lancet. 1987;i:685. 32. De Backer G, Kornitzer M, Dramaix M, et al. The Belgian Heart Disease Prevention Project: 10-year mortality follow-up. Eur Heart J. 1988;9:238–242. 33. Hoffmeister H, Mensink GBM, Stolzenberg H, et al. Reduction of coronary heart disease risk factors in the German Cardiovascular Prevention Study. Prev Med. 1996;25:135–145. 34. Weinehall L, Hellsten G, Boman K, Hallmans G. Prevention of cardiovascular disease in Sweden: the Norsjö community intervention programme—motives, methods and intervention components. Scand J Public Health. 2001;29(suppl 56):13–20. 35. Weinehall L, Hellsten G, Boman K, Hallmans G, Asplund K, Wall S. Can a sustainable community intervention reduce the health gap?— 10-year evaluation of a Swedish community intervention program for the prevention of cardiovascular disease. Scand J Public Health. 2001;29(suppl 56):59–68. 36. Gaziano TA, Galea G, Reddy KS. Chronic diseases 2. Scaling up interventions for chronic disease prevention: the evidence. Lancet. 2007; 370:1939–1945. 37. Roussouw JE, Jooste PL, Chalton DO, et al. Community-based intervention: the Coronary
46894_CH21_619_656.qxd
652
2/25/10
4:16 PM
Page 652
CHAPTER 21 THE CASE FOR PREVENTION
Risk Factor Study (CORIS). Int J Epid. 1993;22:428–438. 38. Yu Z, Song G, Guo Z, et al. Changes in blood pressure, body mass index, and salt consumption in a Chinese population. Prev Med. 1999; 29:165–172. 39. Dong Y. Community health services in China: the experience of Tianjin City. Australian Health Rev. 2001;24:176–182. 40. Dowse GK, Gareeboo H, Alberti KGMM, et al. Changes in population cholesterol concentration and other cardiovascular risk factor levels after five years of the non-communicable disease intervention programme in Mauritius. Brit Med J. 1995;311:1255–1259. 41. Uusitalo U, Feskens EJM, Tuomilehto J, et al. Fall in total cholesterol concentration over five years in association with changes in fatty acid composition of cooking oil in Mauritius: cross sectional survey. Brit Med J. 1996;313: 1044–1046. 42. Zatonski WA, McMichael AJ, Powles JW. Ecological study of reasons for sharp decline in mortality from ischemic heart disease in Poland since 1991. Brit Med J. 1998;316: 1047–1051. 43. Zatonski WA, Willett W. Changes in dietary fat and declining coronary heart disease in Poland: population based study. Brit Med J. 2005;331: 187–188. 44. Walter HJ, Hofman A, Connelly PA, et al. Coronary heart disease prevention in childhood: one-year results of a randomized intervention study. Am J Prev Med. 1986;2: 239–451. 45. Walter HJ, Hofman A, Barrett LT, et al. Primary prevention of cardiovascular disease among children: three year results of a randomized intervention trial. In: Hetzel B, Berenson GS, eds. Cardiovascular Risk Factors in Childhood: Epidemiology and Prevention. New York: Elsevier Science Publishers; 1987:161–181. 46. Walter HJ, Hofman A, Vaughan RD, Wynder EL. Modification of risk factors for coronary heart disease: five-year results of a school-
based intervention trial. N Engl J Med. 1988;318:1093–1100. 47. Luepker RV, Perry CL, McKinlay SM, et al. Outcomes of a field trial to improve children’s dietary patterns and physical activity: the Child and Adolescent Trial for Cardiovascular Health (CATCH). JAMA. 1996;275:768–776. 48. Advisory Board of the Second International Heart Health Conference. The Catalonia Declaration: Investing in Heart Health. Barcelona, Spain: Department of Health and Social Security, Autonomous Government of Catalonia; 1996 49. Fishbein M. Great expectations, or do we ask too much from community-level interventions? Am J Public Health. 1996;86:1075–1076. 50. Koepsell TD, Diehr PH, Cheadle A, Kristal A. Invited commentary: Symposium on Community Intervention Trials. Am J Epidemiol. 1995;142:594–599. 51. Winkleby M. The future of community-based cardiovascular disease intervention studies. Am J Public Health. 1994;84:1369–1372. 52. Susser M. The tribulations of trials—intervention in communities. Am J Public Health. 1995;85:156–158. 53. Ford ES, Giles WH, Mokdad AH. The distribution of 10-year risk for coronary heart disease among U.S. adults. J Am Coll Cardiol. 2004;43:1791–1796. 54. Lopez AD, Mathers CD, Ezzati M, Jamison DT, Murray CJL. Measuring the global burden of disease and risk factors, 1990–2001. In: Jamison DT, Breman JG, Measham AR, et al, eds. Disease Control Priorities in Developing Countries. 2nd ed. Washington, DC: International Bank for Reconstruction and Development/The World Bank; 2006:1–13. 55. World Health Organization. Prevention of Cardiovascular Disease. Guidelines for Assessment and Management of Cardiovascular Risk. Geneva: World Health Organization; 2007.
46894_CH21_619_656.qxd
2/25/10
4:16 PM
Page 653
REFERENCES 653
56. World Health Organization. The World Health Report 2002. Reducing Risks, Promoting Healthy Life. Geneva: World Health Organization; 2002.
65. Lloyd-Jones DM, Leip EP, Larson MG, et al. Prediction of lifetime risk for cardiovascular disease by risk factor burden at 50 years of age. Circulation. 2006;113:791–798.
57. Yusuf S, Hawken S, Ôunpuu S, et al. Effect of potentially modifiable risk factors associated with myocardial infarction in 52 countries (the INTERHEART study): case-control study. Lancet. 2004;364;937–952.
66. Rosamond W, Flegal K, Furie K, et al. Heart disease and stroke statistics—2008 update. A report from the American Heart Association Statistics Committee and Stroke Statistics Subcommittee. Circulation. 2008;117: e1–e122.
58. Steyn K, Sliwa K, Hawken S, et al. Risk factors associated with myocardial infarction in Africa. The INTERHEART Africa Study. Circulation. 2005;112:3554–3561. 59. Rodgers A, Lawes CMM, Gaziano T, Vos T. The growing burden of risk from high blood pressure, cholesterol, and bodyweight. In: Disease Control Priorities in Developing Countries. 2nd ed. Washington, DC: International Bank for Reconstruction and Development/The World Bank; 2006:851–868. 60. Stamler J, Stamler R, Neaton JD, et al. Low risk-factor profile and long-term cardiovascular and noncardiovascular mortality and life expectancy. Findings for 5 large cohorts of young adult and middle-aged men and women. JAMA. 1999;282:2012–2018. 61. Daviglus ML, Liu K, Pirzada A, et al. Favorable cardiovascular risk profile in middle age and health-related quality of life in older age. Arch Int Med. 2003;163:2460–2468. 62. Daviglus ML, Liu K, Pirzada A, et al. Cardiovascular risk profile earlier in life and Medicare costs in the last year of life. Arch Int Med. 2005;165:1028–1034. 63. Hozawa A, Folsom AR, Sharrett AR, Chambless LE. Absolute and attributable risks of cardiovascular disease incidence in relation to optimal and borderline risk factors. Comparison of African American with white subjects—Atherosclerosis Risk in Communities Study. Arch Int Med. 2007;167:573–579. 64. Terry DF, Pencina MJ, Vasan RS, et al. Cardiovascular risk factors predictive for survival and morbidity-free survival in the oldest-old Framingham Heart Study participants. J Am Geriatr Soc. 2005;53:1944–1950.
67. Trogdon JG, Finkelstein EA, Nwaise IA, Tangka FK, Orenstein D. The economic burden of chronic cardiovascular disease for major insurers. Health Prom Pract. 2007;8:234–242. 68. World Health Organization. The World Health Report 2000. Health Systems: Improving Performance. Geneva: World Health Organization; 2000. 69. Nolte E, McKee CM. Measuring the health of nations: updating an earlier analysis. Health Aff. 2008;27:58–71. 70. Ebrahm S. Chronic diseases and calls to action. Int J Epid. 2008;37:225–230. 71. Abegunde DO, Mathers CD, Adam T, Ortegon M, Strong K. Chronic diseases 1. The burden and costs of chronic diseases in low-income and middle-income countries. Lancet 2007; 370:1929–1938. 72. Leeder S, Raymond S, Greenberg H. A Race Against Time: The Challenge of Cardiovascular Disease in Developing Countries. New York: The Trustees of Columbia University in the City of New York; 2004. 73. Gold MR, Siegel JE, Russell LB, Weinstein MC. Cost-Effectiveness in Health and Medicine. New York: Oxford University Press; 1996. 74. Schwappach DLB, Boluarte TA, Suhrcke M. The economics of primary prevention of cardiovascular disease—a systematic review of economic evaluations. Cost Eff Resour Alloc. 2007;5:5. doi:10.1186/1478-7547-5-5. Available at http://www.resource-allocation .com/content/5/1/5.
46894_CH21_619_656.qxd
654
2/25/10
4:16 PM
Page 654
CHAPTER 21 THE CASE FOR PREVENTION
75. Hutubessy R, Chisholm D, Edejer T T-T, WHOCHOICE. Generalized cost-effectiveness analysis for national-level priority-setting in the health sector. Cost Effect Resource Alloc. 2003;1:8. Available at http://www.resource -allocation.com/content/1/1/8. 76. Asaria P, Chisholm D, Mathers C, Ezzati M, Beaglehole R. Chronic diseases 3. Chronic disease prevention: health effects and financial costs of strategies to reduce salt intake and control tobacco use. Lancet. 2007;370: 2044–2053. 77. Lim SS, Gaziano TA, Gakidou E, et al. Chronic diseases 4. Prevention of cardiovascular disease in high-risk individuals in low-income and middle-income countries: health effects and costs. Lancet. 2007;370;2054–2062. 78. Carleton RA. Trends and determinants of coronary heart disease mortality: future prospects and projections. In: Higgins MW, Luepker RV, eds. Trends in Coronary Heart Disease Mortality. The Influence of Medical Care. New York: Oxford University Press; 1988. 79. Unal B, Capewell S, Critchley JA. Coronary heart disease policy models: a systematic review. BMC Public Health. 2006;6:213. doi:10.1186/1471-2458-6-213. 80. Ford ES, Ajani UA, Croft JB, et al. Explaining the decrease in U.S. deaths from coronary disease, 1980–2000. N Engl J Med. 2007;356: 33–43. 81. Ford ES, Capewell S. Coronary heart disease mortality among young adults in the U.S. from 1980 through 2002. Concealed leveling of mortality rates. J Am Coll Cardiol. 2007; 50:2128–2132. 82. Ezzati M, Vander Hoorn S, et al. Estimates of global and regional potential health gains from reducing multiple major risk factors. Lancet. 2003;362:271–280. 83. Mazzati E, Vander Hoorn S, Lopez AD, et al. Comparative quantification of mortality and burden of disease attributable to selected risk factors. In: AD Lopez, CD Mathers, M Ezzati, DT Jamison, CJL Murray, eds. Global Burden
of Disease and Risk Factors. Washington, DC: The International Bank for Reconstruction and Development/The World Bank; 2006:241–396. 84. Murray CJL, Lopez AD, eds. The Global Burden of Disease: A Comprehensive Assessment of Mortality and Disability from Diseases, Injuries, and Risk Factors in 1990 and Projected to 2020. Boston: The Harvard School of Public Health; 1996. 85. Henderson H. Foreword to the Global Burden of Disease and Injury Series. In: Murray CJL, Lopez AD, eds. The Global Burden of Disease: A Comprehensive Assessment of Mortality and Disability from Diseases, Injuries, and Risk Factors in 1990 and Projected to 2020. Boston: The Harvard School of Public Health; 1996: xiii–xiv. 86. Murray CJL, Lopez AD. Alternative visions of the future: projecting mortality and disability, 1990–2020. In: Murray CJL, Lopez AD, eds. The Global Burden of Disease: A Comprehensive Assessment of Mortality and Disability from Diseases, Injuries, and Risk Factors in 1990 and Projected to 2020. Boston: The Harvard School of Public Health; 1996. 87. The World Bank. Dying Too Young. Washington: Europe and Central Asia, Human Development Department, The World Bank; 2005. 88. US Department of Health and Human Services. Healthy People 2010. 2nd ed. With Understanding and Improving Health and Objectives for Improving Health. 2 vols. Washington, DC: US Government Printing Office; 2000. 89. Centers for Disease Control and Prevention. Division for Heart Disease and Stroke Prevention. Available at http://www.cdc.gov/ dhdsp/. Accessed July 6, 2008. 90. American Heart Association. American Heart Association Strategic Goals. Available at http://www.americanheart.org/presenter.jhtml? identifier=4429. Accessed February 11, 2008. 91. Kahn R, Robertson RM, Smith R, Eddy D. The impact of prevention on reducing the burden of cardiovascular disease. Circulation. 2008;118: 576–585.
46894_CH21_619_656.qxd
2/25/10
4:16 PM
Page 655
REFERENCES 655
92. World Health Organization. Gaining Health: The European Strategy for the Prevention and Control of Noncommunicable Diseases. Copenhagen: WHO Regional Office for Europe; 2006. 93. Advisory Board of the International Heart Health Conference. The Victoria Declaration on Heart Health. Victoria, British Columbia: Health and Welfare Canada, BC Ministry of Health and Ministry Responsible for Elders, Heart and Stroke Foundation of Canada; 1992.
102. Stamler J. Low risk—and the “No More Than 50%” myth/dogma. Arch Int Med. 2007;167:537–538. 103. Canto JG, Iskandrian AE. Major risk factors for cardiovascular disease. Debunking the “Only 50%” myth. JAMA. 2003;290: 947–949. 104. Fries JF, Green LW, Levine S. Health promotion and the compression of morbidity. Lancet. 1989;1:481–483.
94. International Heart Health Society. International Action on Cardiovascular Disease: A Platform for Success. Vancouver: International Heart Health Society, 2005. Available at: www.internationalhearthealth .org. Accessed January 21, 2008.
105. Puska P. Community change and the role of public health. In: Marmot M, Elliott P, eds. Coronary Heart Disease Epidemiology: From Aetiology to Public Health. 2nd ed. Oxford (England): Oxford University Press; 2005: 893–907.
95. Syme SL. Rethinking disease: where do we go from here? Ann Epid. 1996;6:463–468.
106. Murray CJL, Kulkarni S, Ezzati M. Eight Americas: new perspectives on U.S. health disparities. Am J Prev Med. 2005;29:4–10.
96. Link BG, Phelan J. Social conditions as fundamental causes of disease. J Health Soc Behav. 1995;36:80–94. 97. Davey Smith G, Ebrahim S, Lewis S, Hansell AL, Palmer LJ, Burton PR. Genetic epidemiology 7. Genetic epidemiology and public health: hope, hype, and future prospects. Lancet. 2005;366:1484–1498. 98. McCormick J, Skrabanek P. Coronary heart disease is not preventable by population interventions. Lancet. 1988;2:839–841. 99. Ebrahim S, Davey Smith G. Exporting failure? Coronary heart disease and stroke in developing countries. Int J Epid. 2001;30: 201–205. 100. Pearson TA. Cardiovascular disease in developing countries: myths, realities, and opportunities. Cardiovasc Drugs Ther. 1999;13: 95–104. 101. Magnus P, Beaglehole R. The real contribution of the major risk factors to the coronary epidemic: time to end the 50% myth. Arch Int Med. 2001;161:2657–2660.
107. The John E. Fogarty International Center. Pathways to Global Health Research. Strategic Plan 2008–2012. NIH Publication No. 08-6261. Bethesda: US Department of Health and Human Services, National Institutes of Health, The John E. Fogarty International Center, May 2008. 108. US Department of Health and Human Services. A Public Health Action Plan to Prevent Heart Disease and Stroke. Atlanta, GA: US Department of Health and Human Services, Centers for Disease Control and Prevention; 2003. 109. Beaglehole R, Ebrahim S, Reddy S, Voûte J, Leeder S on behalf of the Chronic Disease Action Group. Chronic Diseases 5. Prevention of chronic diseases: a call to action. Lancet. 2007;370:2152–2157.
46894_CH21_619_656.qxd
2/25/10
4:16 PM
Page 656
46894_CH22_657_678.qxd
2/25/10
4:16 PM
Page 657
C H A P T E R
22 Taking Action their work for heart disease and stroke prevention, with demonstrated outcomes. Still, despite widely held agreement on needed action, built on a substantial evidence base, prevention policies are not being implemented to full effect at either population or individual levels. Several obstacles at both levels pose challenges for preventive action. Common to both levels and underlying a number of more specific obstacles are issues of economics; health system structure and function; and priorities for CVD and other chronic disease prevention, among other policies within and beyond the health sector. Three strategic priorities for achieving the goals of heart disease and stroke prevention are to strike a new balance in our investment in health; transform our public health agencies into effective instruments for leadership in health system change; and prevent the causes of CVD and other chronic diseases rather than wait to treat their consequences. A continuum of care is envisioned that can link public health and clinical practice within a coherent health system addressing the needs of both populations and patients. Opportunities for effective prevention of the cardiovascular diseases are great, even while further research continues that may still better inform future prevention policies and plans.
SUMMARY Calls to action for prevention of cardiovascular and other chronic diseases are increasingly common at national, regional, and global levels. They variously include broad statements of goals, proposed strategies for achieving such goals, or concrete action plans to be implemented by governments and other players. Examples are cited from the Americas (the United States, Canada, and the Pan American Health Organization (PAHO), Europe (the Russian Federation, Country-wide Integrated Noncommunicable Diseases Intervention [CINDI] Program, and the European Region of the World Health Organization [WHO]), South Asia (Pakistan and the South Asian Association for Regional Cooperation [SAARC]), and other regions as well as at the global level (Heart Health Networks, WHO, and the Disease Control Priorities in Developing Countries Project [DCP2]). As a case study of action plan initiation, implementation, and institutionalization, A Public Health Action Plan to Prevent Heart Disease and Stroke, developed by the US Centers for Disease Control and Prevention (CDC) in partnership with the American Heart Association (AHA), Association of State and Territorial Health Officials (ASTHO), and other partners, is examined in some detail. This example describes a process for moving from explicit national goals to responsive recommendations and proposed action steps to comprise a comprehensive public health strategy for CVD prevention. Further steps to prioritize actions, establish means of their implementation, and accomplish the indicated tasks are also outlined. An action plan can be developed and can guide interested organizations and agencies in
INTRODUCTION: CALLS TO ACTION Parallel developments along multiple lines—knowledge of causation, concepts of prevention, assessments of evidence, and guidelines for clinical and public health practice—culminate in a strong case for prevention. If the force of argument supporting public
657
46894_CH22_657_678.qxd
658
2/25/10
4:16 PM
Page 658
CHAPTER 22 TAKING ACTION
health intervention to prevent cardiovascular and other chronic diseases is sufficient, effective action should be expected to follow. Accordingly, calls to action have been presented by many authoritative organizations and agencies. Some of these are found in the visionary statements illustrated in the preceding chapter. Building on precedents beginning decades ago, recent calls to action in CVD prevention have become more frequent, more comprehensive, and more prominent. For example, in the United States, the sense of urgency about chronic disease prevention led to development of an innovative partnership among three leading national voluntary organizations, the American Cancer Society, American Diabetes Association, and American Heart Association. This new collaboration and a common agenda were announced simultaneously in the principal journals of the three organizations in 2004:1, p 3244 Current approaches to health promotion and prevention of cardiovascular disease, cancer, and diabetes do not approach the potential of the existing state of knowledge. A concerted effort to increase application of public health and clinical interventions of known efficacy to reduce prevalence of tobacco use, poor diet, and insufficient physical activity––the major risk factors for these diseases––and to increase utilization of screening tests for their early detection could substantially reduce the human and economic cost of these diseases. Numerous national, regional, and global initiatives with similar missions have emerged over the past decade or more from governmental agencies and nongovernmental organizations alike. Calls for action have become commonplace, perhaps to excess, as suggested by Ebrahim in a recent editorial in which he notes, “The real challenge for any call to action is to develop and implement a plan for achieving its goals.”2, p 227 A call to action may in itself be a meaningful act of communication, as it conveys the urgency of current conditions, potential impact of intervention, and cost of inaction, yet it may fall short of actual plans for any direct health intervention. Effective response to a call to action requires taking action, which is presumed to have greatest likelihood of impact when based on explicit goals, strategies, and plans. The emergence of well-articulated responses to calls for prevention of cardiovascular and other chronic diseases is illustrated through an overview of examples from North American, European, South Asian, and other regional and global efforts. One example serves as a case study: A Public Health Action
Plan to Prevent Heart Disease and Stroke, first published in the United States in 2003.3
OVERVIEW: GOALS, STRATEGIES, AND ACTION PLANS Americas In the United States, nongovernmental and governmental bodies have gone varying distances along the path to a fully explicit action plan for CVD prevention. The American Heart Association (AHA), for example, has as its mission “Building healthier lives, free of cardiovascular disease and stroke.”4 (www.american heart.org; accessed February 11, 2008) AHA adopted strategic goals for the current decade in order, by 2010, “to reduce coronary heart disease, stroke, and risk by 25 percent.” The goal statement includes indicators of progress related to reduced coronary heart disease and stroke mortality, prevalence of smoking, high blood cholesterol, physical inactivity and uncontrolled high blood pressure, and limited growth of overweight and diabetes. Strategies are outlined by which to meet the targets for these indicators in four areas— knowledge, revenue generation, evaluation, and capacity building. Consistent with its mission and goals, AHA also published its Guide for Improving Cardiovascular Health at the Community Level and the sequel, Taking the Initiative, reviewed in Chapter 20.5,6 Although these documents stop short of proposing specific tasks, timelines, or accountability for achievement of goals and objectives, together they clearly anticipate that responsive action will occur. Coalitions of other nongovernmental organizations have also been established in the United States to address chronic disease prevention generally, such as the Partnership to Fight Chronic Disease, Partnership for Prevention, and Trust for America’s Health7–9 (www.fightchronicdisease.org, www.prevent.org, www.healthyamericans.org; accessed September 21, 2008). In some instances, such Web sites provide opportunities for action by interested visitors, although systematic action plans are not presented. The leading example of explicit goals, strategies, and action plans in the United States is A Public Health Action Plan to Prevent Heart Disease and Stroke.3 This is a comprehensive strategic plan that derived from national goals presented in Healthy People 2010 and that included recommendations, action steps, and a mechanism for long-term implementation, the National Forum for Heart Disease and Stroke Prevention. This plan will be examined in some detail as a case study as follows.
46894_CH22_657_678.qxd
2/25/10
4:16 PM
Page 659
OVERVIEW: GOALS, STRATEGIES, AND ACTION PLANS 659
In Canada, leadership in CVD prevention was represented early by the Canadian Heart Health Initiative, a federal-provincial partnership designed to implement and evaluate community interventions to reduce the burden and risk of CVD across Canada10,11 (www.pac-aspc.gc.ca/ccdpc-cpcmc/cindi/ pdf/chhi-eval_e.pdf; accessed September 21, 2008). Begun in 1986 with a five-phase 18-year plan, the program provided experience in integrated CVD prevention and “healthy living promotion” on which future community-based interventions for chronic disease prevention could be based. After this demonstration program and the series of heart health conferences cited previously, it was recognized that there was “no over-arching pan-Canadian strategy for heart health, nor . . . clear collaborative leadership or a coordinating mechanism to guide such a strategy”12, p 16-1 (www .chhs-scsc.ca; accessed September 21, 2008) With completion planned in late 2008, the direction of the Canadian Heart Health Strategy-Action Plan (CHHS-AP) can be anticipated from its statement of purpose and objectives:12, p 20–21 1. to develop a collaborative and inclusive Canadian Heart Health Strategy and Action Plan that will engage stakeholders with a commitment to improving cardiovascular health for Canadians Output: a comprehensive Canadian strategy document for a collaborative approach to cardiovascular health Output: a business plan to guide and support implementation of the Strategy 2. to agree on six to eight Theme Working Group topics (areas of emphasis) and related strategic priorities on the basis of the best evidence available Output: identification of five or six strategic priorities and related recommendations for action for each identified Theme Working Group Output: development of innovative knowledge synthesis, exchange and translation practices (appropriate for a range of populations, including those who are underserved) to promote the implementation of recommended policies and programs developed by each Theme Working Group 3. to develop a plan for a system (and related components) for a comprehensive and coordinated approach to cardiovascular health surveillance for Canada Output: a five-year staged plan for cardiovascular health surveillance, including key indicators
4. to propose a five-year monitoring and evaluation protocol to document the impact of the Strategy and suggest improvements Output: a document outlining realistic formative and summative approaches to monitoring and evaluation using a range of quantitative and qualitative strategies This Canadian initiative will thus present a true national strategy and action plan for cardiovascular health, combining key elements of collaboration, comprehensiveness, planned resources for implementation, recommended actions in identified priority areas, plans for monitoring and evaluation, and a supporting surveillance system. More broadly for the region, a Health Agenda for the Americas, 2008–2017 was presented by the Ministers of Health of the Americas in 2007.13 (Available at www.paho.org, accessed June 30, 2008.) The document includes a declaration regarding commitment to a broad health agenda and principles, followed by a situation analysis and review of regional health trends. Chronic diseases are recognized as dominant causes of morbidity and mortality throughout the region. The health agenda addresses strengthening the national health authorities, tackling health determinants, increasing social protection and access to quality health services, diminishing health inequalities among countries and inequities within them, reducing the risk and burden of disease, strengthening the management and development of health workers, harnessing knowledge, science, and technology, and strengthening health security (preparing, for example, for natural disasters or pandemic influenza). Regarding risk and burden of disease:13, p 18 Specific actions should be initiated to control diabetes, cardiovascular and cerebrovascular diseases, as well as types of cancer with the greatest incidence, as well as hypertension, dyslipidemia, obesity, and physical inactivity. . . . The health authority should be highly active in promoting healthy lifestyles and environments. Changes in behavior will only be sustained if they are accompanied by environmental, institutional, and policy changes that truly allow people to choose lifestyles that involve healthy eating habits, physical activity, and not smoking. Collaboration with industry, the media, and other strategic partners is needed to produce and market healthier foods, and with the education sector so that schools set an example of good dietary practices and promote healthy habits.
46894_CH22_657_678.qxd
660
2/25/10
4:16 PM
Page 660
CHAPTER 22 TAKING ACTION
The agenda, though far-reaching in scope, gives less direct guidance to action in the cardiovascular or chronic disease areas than, for example, the US and Canadian strategies. However, the fundamental concepts have much in common: expectation of shared impact across several chronic conditions, need for multisector approaches, and emphasis on diet, physical activity, and tobacco. Europe In Europe, at the national level, a very recent example is the document Strategy for the Prevention and Control of Noncommunicable Diseases and Injuries in the Russian Federation.14 This report was released by the Ministry of Health and Social Development of the Russian Federation and the State Research Center for Preventive Medicine, a World Health Organization (WHO) Collaborating Center on Development and Implementation of Noncommunicable Disease Prevention Policy and Programs:14, p 3 The purpose of the Strategy is to create an intersectoral system aimed at preventing the development and progression of NCDs and injuries in Russia by means of integrated measures that promote a healthier lifestyle, adjust the risk factors on which these diseases are based, enhance the effectiveness of treatment so as to improve the quality of and prolong people’s lives and increase the country’s work force and economic capacity. The main objectives of the Strategy are: • to upgrade the priority of preventing NCDs and Is in the programs of national action aimed at strengthening and preserving the health of the population; • to develop a system of interagency and interregional cooperation and partnership on matters of improving health and preventing NCDs and Is; • to determine and create an effective infrastructure for improving health and preventing NCDs and Is; • to propose resources for increasing, and ways of distributing, resources, such a professionals, equipment and funds allocated for health promotion and prevention and monitoring of NCDs and Is; • to define the role of the Strategy for the Prevention of NCDs and Is in the implementation of top-priority national health projects.
This document is aimed above all at decisionmakers: members of the RF Government and the government administration, members of the parliament, to the heads and staff members of the Russian Health Ministry and other ministries and agencies, to the heads of federal districts and regions, the heads of the public-health system in the regions, and to all who may participate in developing and implementing programs to prevent NCDs and Is in Russia. This report illustrates the approach of linking cardiovascular diseases, which account for more than 50% of all deaths in the Russian Federation, with other chronic diseases (together, NCDs) and injuries (Is), another major cause of death and disability. Emphasis is placed on implementation of programs that are nationwide, appropriate in content, and of sufficient size, duration, and intensity to be effective. Healthy lifestyles are to be promoted beginning in childhood and extending throughout life, both preserving physical and mental health and providing health care when needed. The strategy also addresses issues of according priority for prevention of these conditions within the national agenda, the need for cooperation across levels of government, and adequacy of infrastructure and resources. Although it focuses on decision makers throughout government, it also recognizes “all of those who may participate” as a potentially broader constituency for preventing NCDs and Is in Russia. Seven action areas are identified: developing policies and funding sources; improving legislative and regulatory frameworks; strengthening the public health system; training of professionals; educating the public; creating a system for monitoring NCDs, Is, and risk factors; and developing international cooperation. For implementation of the Strategy, specific decisions are required of governmental and parliamentary levels and the level of the Russian Health Ministry. Multinational and regional strategies and action plans have been proposed in Europe at least from the mid-1990s. A decade earlier, in 1986, WHO had established the Country-wide Integrated Noncommunicable Diseases Intervention (CINDI) Program as part of the Health for All by the Year 2000 program of the European Region of WHO.15 As a network of 24 countries primarily in Europe, the CINDI Program developed in 1995 a CINDI-EUROHEALTH Action Plan for NCD prevention and control.16 Priorities for action within the CINDI agenda were: policy development; multiple-risk reduction intervention through primary care and community intervention; enhancing prevention in primary care; children, youth, and families; worksites; and training. Experience based on the multinational character of the CINDI network
46894_CH22_657_678.qxd
2/25/10
4:16 PM
Page 661
OVERVIEW: GOALS, STRATEGIES, AND ACTION PLANS 661
led to recognition that intervention of any type at a local level would require adaptation to local infrastructure, priorities, and capacities. Accordingly, an action plan for the region should be considered as a framework within which countries could implement specific relevant activities. The CINDI-EUROHEALTH Action Plan was the forerunner of a document published in 2006, Gaining Health: The European Strategy for the Prevention and Control of Noncommunicable Diseases.17 The document begins by presenting the challenges to health and equity and to societies and health systems resulting from the burden of NCDs across the 52 member states of the European Region—accounting for 77% of DALYs and 86% of all deaths. Discussion follows of potential gains in health through effective strategies of prevention and a statement of vision and principles to guide strategic planning. The vision is that of “A health-promoting Europe free of preventable noncommunicable disease, premature death and avoidable disability”; the two objectives of the strategy are “To take integrated action on risk factors and their underlying determinants across sectors” and “To strengthen health systems for improved prevention and control of NCD.”17, p 17 Several key messages underlie the proposed actions for European states and may be taken to reflect the prevailing view of personal and societal roles in health:17, p 17 1. Prevention throughout life is effective and must be regarded as an investment in health and development. 2. Society should create health-supporting environments, thereby also making healthy choices easier choices. 3. Health and medical services should be fit for purpose, responding to the present disease burden and increasing opportunities for health promotion. 4. People should be empowered to promote their own health, interact effectively with health services and be active partners in managing disease. 5. Universal access to health promotion, disease prevention and health services is central to achieving equity in health. 6. Governments at all levels have the responsibility to build healthy public policies and ensure action across all the sectors concerned. The European Strategy is rooted in a composite framework that joins the Bangkok Charter for Health Promotion in a Globalized World and WHO Health
Systems Framework to inform six action areas: advocacy, knowledge, regulation and financing, capacities, community support, and health service delivery (Figure 22-1). It also recognizes an array of existing WHO strategies and action plans in areas relevant to NCD prevention. Nine such strategies are indicated as contributing to the specifics of the six lines of action, making the NCD strategy both comprehensive and consistent with previously defined, but independent, approaches (Figure 22-2). The Strategy presents altogether some 80 examples of specific actions under the six headings shown in the figures. The actions are “primarily directed at ministries of health and public health policy-makers, although other stakeholders such as nongovernmental organizations, private sector and community groups will recognize examples of relevance for their own work.”17, p 21 The concluding section, addressing the way forward, outlines areas of support to Member States that could be provided by WHO. These areas are, broadly, strengthening international, bilateral, and multilateral cooperation; facilitation of information exchange, technical cooperation, and capacity-building; and research, monitoring, and surveillance. The Strategy was endorsed by the WHO Regional Committee for Europe in September 2006. South Asia Pakistan’s public-private health partnership, Heartfile, took leadership in developing the National Action Plan for Prevention and Control of NonCommunicable Diseases and Health Promotion in Pakistan.18 This very substantial initiative was published in 2004 with joint sponsorship of the Ministry of Health, Government of Pakistan; WHO, Pakistan Office; and Heartfile. Following discussion of underlying concepts and principles, the document comprises several self-contained sections, first addressing action areas common to multiple NCDs, then specifically cardiovascular diseases, diabetes, tobacco use, chronic respiratory diseases, cancer, injuries, and mental illnesses. Each section concludes with its own action agenda, such as that for cardiovascular diseases reproduced here (Table 22-1). This 18-point agenda emphasizes eight priority items that concern surveillance, communications on physical activity and diet, population-level interventions, nutrition policy, environmental requirements for physical activity, religious support for physical activity among women, and health system capacity for CVD prevention and control. Key to the Plan is an “Integrated Framework for Action” that addresses systematically the elements of implementation and evaluation in three areas integrated across all disease outcomes:
46894_CH22_657_678.qxd
662
2/25/10
4:16 PM
Page 662
CHAPTER 22 TAKING ACTION
BANGKOK CHARTER FOR HEALTH PROMOTION
HEALTH SYSTEMS FRAMEWORK
Advocate for health based on human rights and solidarity
Stewardship and governance: influencing, formulating and implementing policy
Invest in sustainable policies, actions and infrastructure to address the determinants of health Build capacity for policy development, leadership, health promotion practice, knowledge transfer and research, and health literacy
Health financing: raising revenues, pooling funds and purchasing services
Regulate and legislate to ensure a high level of protection from harm and enable opportunity for health and well-being for all people
Resource generation: creating human resources, infrastructure and consumables
Partner and build alliances with public, private, nongovernmental and international organizations and civil society to create sustainable actions
Service delivery: efficiently producing high quality and accessible personal and non-personal services
ADVOCACY KNOWLEDGE REGULATION AND FINANCING CAPACITIES COMMUNITY SUPPORT HEALTH SERVICE DELIVERY
Figure 22-1 A Composite Framework for Action on NCD. Source: Reprinted with permission from WHO Regional Office for Europe, Gaining Health: The European Strategy for the Prevention and Control of Noncommunicable Diseases, WHO Regional Office for Europe, © World Health Organization 2006, p 18.
surveillance, an integrated behavioral change communication strategy, and an integrated reorientation of health services. The breadth of proposed actions has much in common with guidelines and policies for prevention reviewed in previous chapters, and it reflects a common perspective with advocates for CVD prevention elsewhere. A second initiative in the region was published in 2002 jointly by Heartfile and the South Asian Association for Regional Cooperation (SAARC), which links Bangladesh, Bhutan, India, Maldives, Nepal, Pakistan, and Sri Lanka to address issues of common concern to countries throughout South Asia (www.saarc-sec.org; accessed January 21, 2008).19 Preventing Coronary Heart Disease in South Asia: SAARC Cardiac Society Guidelines and Recommendations is more than a clinical guideline.20 It provides a broad contextual framework for addressing CVD prevention in the region and policy and program recommendations for public health as well as clinical practice. The document addresses the partic-
ular situation of south Asian populations with respect to demographics, economics, healthcare resources, and epidemiologic considerations. Analogous to the European Strategy, this report anticipates action tailored to the circumstances of each member country. Other Regional and Global Initiatives Heart Health Networks The need for action to promote heart health at community and national levels has been expressed continuously in the successive Declarations issued by the International Heart Health Society at its conferences from 1992 through 2004 and summarized in the document International Action on Cardiovascular Diseases: A Platform for Success.21 (Available at www.internationalhearthealth.org.) Action recommendations in this synthesis of previous Declarations were directed to:21, pp 23–24
46894_CH22_657_678.qxd
2/25/10
4:16 PM
Page 663
OVERVIEW: GOALS, STRATEGIES, AND ACTION PLANS 663
GLOBAL STRATEGY FOR PREVENTION AND CONTROL OF NONCOMMUNICABLE DISEASES 2000 HELSINKI DECLARATION & MENTAL HEALTH ACTION PLAN 2005 EUROPEAN STRATEGY FOR TOBACCO CONTROL 2002 EUROPEAN ACTION PLAN FOR FOOD AND NUTRITION POLICY 2000 EUROPEAN STRATEGY FOR CHILD AND ADOLESCENT HEALTH AND DEVELOPMENT 2005 EUROPEAN FRAMEWORK FOR ALCOHOL POLICY 2005 EUROPEAN ALCOHOL ACTION PLAN 2000 – 2005 STOCKHOLM DECLARATION ON YOUNG PEOPLE AND ALCOHOL 2001
ADVOCACY KNOWLEDGE REGULATION AND FINANCING CAPACITIES COMMUNITY SUPPORT HEALTH SERVICE DELIVERY
A C T I O N S
CHILDREN’S ENVIRONMENT AND HEALTH ACTION PLAN FOR EUROPE 2004 GLOBAL STRATEGY ON INFANT AND YOUNG CHILD FEEDING 2002 GLOBAL STRATEGY ON DIET, PHYSICAL ACTIVITY AND HEALTH 2004
Figure 22-2 A Comprehensive Action-Oriented Approach. Source: Reprinted with permission from WHO Regional Office for Europe, Gaining Health: The European Strategy for the Prevention and Control of Noncommunicable Diseases, WHO Regional Office for Europe, © World Health Organization 2006, p 19.
WHO, World Heart Federation and the International Heart Health Society and the US Centers for Disease Control and Prevention: Undertake a leadership role in facilitating international collaboration and communication among all countries and international organizations. International organizations and countries’ internal infrastructures: Adopt and promote of the 5 core values of the European Region of WHO as the foundation of CVD prevention and control programs; implement public policies across ministries to address social and behavioral determinants of CVD; develop strategies with the private sector to take economic and health effects of globalization in to account; recognize and create policy based on economic benefits of primary prevention of CVD; collaborate to address inequities between developed and developing countries, the rich and poor within countries, and genders at all ages.
Countries’ internal infrastructures: Facilitate effective collaboration between government, non-governmental organizations, the private sector, community groups, and partners outside the health sector to assure systematic and strategic planning, development, implementation and evaluation of CVD prevention programs. Governments: Invest in international CVD coalitions; create a balanced, coordinated and comprehensive prevention and treatment health system that is founded on the five health goals of the Declarations; includes CVD in an allNCD program; adds CVD prevention and control elements to existing programs; provides a comprehensive set of strategies including programs and services, public policies, community action, research and surveillance, and training of health service providers and other professionals; disseminates and implements effective prevention strategies; and invests adequate
46894_CH22_657_678.qxd
664
2/25/10
4:16 PM
Page 664
CHAPTER 22 TAKING ACTION
Table 22-1
Cardiovascular Diseases—Action Agenda
• Integrate surveillance of cardiovascular risk factors with a population-based NCD surveillance system; develop and validate tools of assessment for the Pakistani population. Integrate public health programme monitoring and evaluation with NCD surveillance.† • Promote physical activity and a healthy diet as a cultural norm as part of the NCD behavioural change communication strategy. Create awareness about the risks of CVD and its mitigates, prevention of RF and RHD and screening approaches.† • Promote strategies for mitigation of cardiovascular risks through population-level approaches.† • Revisit health policy on diet and nutrition to expand its current focus on under-nutrition.† • Develop a nutrition and physical activity policy seeking guidance from the WHO Global Strategy on Diet and Physical Activity.† • Develop policies and strategies to limit the production of, and access to, ghee as a medium for cooking. • Develop agricultural and fiscal policies that increase the demand for, and make healthy food more accessible. • Create an enabling physical and social environment for physical activity.† • Generate support from religious leaders to endorse the need for participation of women in physical activity.† • Enforce effective legislation to stipulate standards for urban planning. • Utilize available open spaces for physical activity where feasible and appropriate. • Integrate concerted primary and secondary prevention programmes into health services as part of a comprehensive and sustainable, scientifically valid, culturally appropriate and resource-sensitive CME programme for all categories of healthcare providers. • Promote screening for raised blood pressure at the population level. Promote high-risk screening for dyslipidaemia and diabetes in high-risk groups only. • Focus attention on improving the quality of prevention programmes within primary and basic health sites. • Ensure availability of aspirin, beta blockers, thiazides, ACE inhibitors, statins and penicillin at all levels of healthcare. • Conduct clinical end-point trials in the native Pakistani setting to define cost-effective therapeutic strategies for primary and secondary prevention of CVDs. • Build capacity of health systems in support of CVD prevention and control.† • Build a coalition or network of organizations at the national, provincial and local levels facilitated by federal and provincial health services to add momentum to CVD prevention and control as part of a comprehensive NCD prevention effort. †
Priority Action Areas. Priorities within other Action Areas will be determined subsequently.
Source: Reprinted with permission from Ministry of Health, Government of Pakistan; the World Health Organization, Pakistan Office; Heartfile, National Action Plan for Prevention and Control of Non-Communicable Diseases and Health Promotion in Pakistan, © Tripartite Collaboration on NCDs in Pakistan 2004, p 42.
resources in public health and primary health care while balancing attention to specialty care. Government and other research bodies: Fund interdisciplinary CVD research with a focus on CVD prevention policies. Here, then, is a consolidated call to action with a charge to international, national, and internal domestic organizations and agencies to undertake the essential tasks of a global movement for CVD prevention. Following the second International Heart Health Conference, in Barcelona in 1995, a supplemental report was published in 1997, Worldwide Efforts to Improve Heart Health.15 This report compiled information regarding a great many local and national programs and included six international heart health networks—in addition to the Canadian Heart Health Initiative and CINDI Program noted previously, these were networks in francophone countries, Chinese language countries and areas, the South Asian Heart Health Network, and the globally constituted
InterHealth organization. This report provides general descriptions of program activities in each instance, but it does not offer details of action plans guiding these efforts. This information may be available through identified contacts for leaders and members of each program. The World Heart Federation has called attention to work of the newly established African Heart Network, founded in 2001, whose members represent Cameroon, Democratic Republic of Congo, Ghana, Ivory Coast, Kenya, Mozambique, Nigeria, Rwanda, South Africa, Sudan, and Tunisia (www.world-heartfederation.org; accessed January 21, 2008).22 Its principal focus is on tobacco control. The World Health Organization (WHO) in 2005 published Cardiovascular Disease Prevention and Control: Translating Evidence into Action.23 This report described activities undertaken in keeping with the global strategy for NCD prevention and control endorsed by the World Health Assembly in 2000. Three areas of emphasis for country-based activities
46894_CH22_657_678.qxd
2/25/10
4:16 PM
Page 665
CASE STUDY: A PUBLIC HEALTH ACTION PLAN TO PREVENT HEART DISEASE AND STROKE 665
were CVD risk reduction and secondary prevention of heart attacks and strokes, as well as secondary prevention of rheumatic heart disease. The report centered on two focal areas: scaling up secondary prevention through implementation of the WHOPREMISE (Prevention of Recurrences of Myocardial Infarction and StrokE) project and the CVD-Risk Management Project, an integrated approach to primary prevention of CVD in high-risk populations. More recently, the World Health Assembly in April 2008 received a report, Prevention and Control of Noncommunicable Diseases: Implementation of the Global Strategy, that presented a draft action plan based on broad input from Member States and other interested parties.24 The purpose of the plan is:24, pp 3–4 • mapping the emerging epidemics of noncommunicable diseases and analysing their social, economic, behavioural and political determinants as the basis for providing guidance on the policy, programmatic, legislative and financial measures that are needed to support and monitor the prevention and control of noncommunicable diseases; • reducing the level of exposure of individuals and populations to the common modifiable risk factors for noncommunicable diseases–– namely, tobacco use, unhealthy diet and physical inactivity, and the harmful use of tobacco––and their determinants, while at the same time strengthening the capacity of individuals and populations to make healthier choices and follow lifestyle patterns that foster good health; and • strengthening health care for people with noncommunicable diseases by developing evidence-based norms, standards and guidelines for cost-effective interventions and by reorienting health systems to respond to the need for effective management of diseases of a chronic nature. Objectives of the plan are to raise priority of NCD prevention and integrate efforts across government departments; strengthen national policies and plans; reduce the risk factors common to multiple chronic diseases; promote research on NCD prevention and control; promote partnerships; and monitor NCDs and their determinants in order to evaluate progress at national, regional, and global levels. Further calls to action on a global scale, with special reference to low- and middle-income countries, are found, for example, in work from the Disease Control Priorities Project (DCP2), such as its
publication Priorities in Health, discussed in previous chapters.25 Here, cost-effectiveness of strategies to address the cluster of CVD, diabetes, high blood pressure, cholesterol, and body weight is addressed under the broad categories of lifestyle and medical interventions. Its “blueprint for action” acknowledges the limited specificity of proposals for action presented on a global level. “Even though the selection and design of interventions is not something that can be characterized in a single universal plan, some common features do emerge from DCP2.”25, p 180 These features, seemingly calling for action at global, regional, and national levels, address delivery and availability for everyone of cost-effective interventions; adequacy of public financing; a higher level of international financial and technical assistance; multisector collaboration; strengthened health systems; and building the knowledge base in basic and applied sciences and management. Similarly aimed at the global level, A Race Against Time concludes by stating that action is needed on a wide range of fronts to combat CVD effectively. These actions include:26, pp 85–87 Putting CVD in the developing world on the international health and development agenda . . . Deeper documentation of the prevalence and costs of CVD . . . Developing partnerships at the macroeconomic level with national governments in key developing countries . . . Train the trainer initiatives in health education . . . Undertaking trial treatment and prevention interventions . . . Longer term research and interventions . . . This agenda is seen as a basis for confidence that CVD control can be advanced in developing countries and as a framework for collaboration around the world.
CASE STUDY: A PUBLIC HEALTH ACTION PLAN TO PREVENT HEART DISEASE AND STROKE The preceding examples attest to the breadth of concern about epidemic CVD at national, regional, and global levels and the growing number of calls for action to intervene. It remains to consider one example more closely with respect to its initiation, implementation, and institutionalization, as it may guide future
46894_CH22_657_678.qxd
666
2/25/10
4:16 PM
Page 666
CHAPTER 22 TAKING ACTION
action for CVD prevention at local, state, and national levels in the United States and elsewhere: A Public Health Action Plan to Prevent Heart Disease and Stroke.3 Background The Action Plan was created during 2002 and released by the Secretary of the US Department of Health and Human Services (DHHS) in April 2003. The background of its development was reviewed on the occasion of its first 5-yearly update presented in March 2008.27,28 In 1994, in the absence of federal funding to states in support of CVD prevention, the Centers for Disease Control and Prevention (CDC) and National Heart, Lung and Blood Institute (NHLBI) supported a steering committee to develop a document, Preventing Death and Disability from Cardiovascular Diseases: A State-Based Plan for Action.29 This report called for federal support for much-needed efforts to address the documented CVD burden in the states and presented outlines of a plan that could bring about the needed improvements. Subsequently, in 1998, the US Congress appropriated funds enabling CDC to initiate “a comprehensive cardiovascular program, with particular emphasis on risk factors and the promotion of healthy behaviors.”30 A key further development setting the stage for the Action Plan was release in January 2000 of Healthy People 2010, presenting the nation’s goals for improving health throughout the new decade.31 In view of its responsibility for the just-established state CVD prevention program, CDC was designated with the National Institutes of Health (NIH) as “co-lead agency” within DHHS to work for progress toward achieving the national goal for heart disease and stroke prevention. This new role and accountability, adding to CDC’s vision for future growth of the state program, provided the impetus to develop a longrange strategic plan for public health efforts in this area. Its purpose would be “To chart a course for the Centers for Disease Control and Prevention (CDC) and collaborating public health agencies, with all interested partners and the public at large, to help in promoting achievement of national goals for preventing heart disease and stroke over the next two decades––through 2020 and beyond.”3, p 1 Initiation Preparation for the planning process began in January 2000 and continued to mid-2001. Leadership came from CDC and co-lead partners, the American Heart Association (AHA) and the Association of State and Territorial Health Officials (ASTHO). The conceptual foundation of the Action Plan would be the goal for
heart disease and stroke prevention presented in Healthy People 2010.31 This goal could be considered in four complementary parts: prevention of risk factors, detection and treatment of risk factors, early identification and treatment of heart attacks and strokes, and prevention of recurrent cardiovascular events. A comprehensive public health strategy would be expected to address each of these four goals and the needs and opportunities for effective prevention under each of them. A second underlying element of the plan would be a limited set of focus areas within which recommendations could be developed. A panel of experts, each with some 12–15 members representing major partner organizations, state and local health agencies, and other interested parties, would be appointed for this purpose. Their work would be guided by a working group similarly composed. The five designated panels and the respective topical areas or “essential components” of the plan were described in the Action Plan:3, pp 7–8 Taking action––Translating current knowledge into effective public health action (Expert Panel A). Strengthening capacity––Transforming public health agencies with new competencies and resources and expanding partnerships to mount and sustain such action (Expert Panel B). Evaluating impact––Systematically monitoring and evaluating the health impact of interventions to identify and rapidly disseminate those most effective (Expert Panel C). Advancing policy––Defining the most critical policy issues and pursuing the needed prevention research to resolve them and expedite policy development (Expert Panel D). Engaging in regional and global partnerships––Multiplying resources and capitalizing on shared experience with others throughout the global community who are addressing similar challenges (Expert Panel E). Each group met on two occasions at two-month intervals between January and May 2002 to define their areas more fully and develop recommendations, proposed action steps, and expected outcomes of these activities. These draft recommendations were compiled by the Working Group, which identified two cross-cutting topics, or “fundamental requirements,” to be recognized in addition to the five essential components: effective communication and public health leadership, partnership, and organization. In September 2002, in order to engage a broader constituency in support of the plan, poten-
46894_CH22_657_678.qxd
2/25/10
4:16 PM
Page 667
CASE STUDY: A PUBLIC HEALTH ACTION PLAN TO PREVENT HEART DISEASE AND STROKE 667
tially interested organizations and agencies were invited to participate in a National Forum for Heart Disease and Stroke Prevention. Nearly 70 organizations were represented by 120 attendees, who reviewed the draft recommendations and provided extensive input, leading to completion of the draft report in December 2002. The Action Plan is presented in four sections, following an executive summary: (1) Heart Disease and Stroke Prevention: Time for Action; (2) A Comprehensive Public Health Strategy and the Five Essential Components of the Plan: Platform for Action; (3) Recommendations: A Call to Action; and (4) Implementation: Mobilizing for Action. The conceptual foundation of the plan is represented in a detailed graphic, Strategic Framework for a Comprehensive Public Health Strategy to Prevent Heart Disease and Stroke (see Figure 18-9). The core of the plan is its 22 final recommendations, each with three to four proposed action steps and their expected outcomes. The recommendations can be expressed in summary form under each of the two fundamental requirements and five essential components, as follows:3, pp 8–11 Effective communication: Communicate to the public at large and to policy makers the urgent need and unprecedented opportunity to prevent heart disease and stroke in order to establish widespread awareness and concern about these conditions, as well as confidence in the ability to prevent and control them. Public health leadership, partnership, and organization: Transform the nation’s public health infrastructure to provide leadership and to develop and maintain effective partnerships and collaborations for the action needed. Taking action: Develop policies for preventing heart disease and stroke at national, state, and local levels to assure effective public health action, including new knowledge on the efficacy and safety of therapies to reduce risk factors. Implement intervention programs in a timely manner and on a sufficient scale to permit rigorous evaluation and the rapid replication and dissemination of those most effective. Promote cardiovascular health and prevent heart disease and stroke through interventions in multiple settings, for all ages groups, and for the whole population, especially high-risk groups. Strengthening capacity: Strengthen public health agencies to assure that they develop and maintain sufficient capacities and competencies, including their laboratories. Create opportunities for training, offer model standards for preventing chronic diseases, and
make consultation and technical support continuously available to public health agencies, including their laboratories. Monitoring and evaluation: Define criteria and standards for population-wide health data sources. Expand these sources as needed to assure adequate long-term monitoring of population measures related to heart disease and stroke. Upgrade and expand health data sources to allow systematic monitoring and evaluation of policy and program interventions. Advancing policy: Emphasize the critical roles of atherosclerosis and high blood pressure, which are the dominant conditions underlying heart disease and stroke, within a broad prevention research agenda. Develop innovative ways to monitor and evaluate policies and programs, especially for policy and environmental change and population-wide health promotion. Engaging in regional and global collaboration: Reap the full benefit of shared knowledge and experience from regional and global partners through information exchange in the area of heart disease and stroke prevention. Work with regional and global partners to develop prevention policies, formulate strategies for use of global media for health communications, and assess the impact of globalization on cardiovascular health. The Action Plan as released in April 2003 represented a substantial advance from the statement of goals in Healthy People 2010 by proposing specific recommendations and a wide range of supporting action steps to promote cardiovascular health and prevent heart disease and stroke. Broad participation in the process represented national, state, and local government, national and international organizations, academia, health professional organizations, community interests, and others. Thorough review and approval across the many agencies within DHHS and introductory messages from the Secretary and from directors of CDC and NIH gave high-level authority to the plan. (It should be noted that the Action Plan did not include specific attention to policies to improve healthcare quality, such as pay for performance and health information technology. These and more recent developments require consideration at the interface between public health goals and delivery of clinical services.) To put the plan into action, two major further steps were required: first, to implement the plan and, second, to institutionalize it.
46894_CH22_657_678.qxd
668
2/25/10
4:16 PM
Page 668
CHAPTER 22 TAKING ACTION
Implementation A strength of the plan, given the purpose of setting an agenda for two decades or more, is the number and scope of its 22 recommendations and some 70 proposed action steps. It became apparent soon after its release that this strength was also a liability, in that direction and focus were needed to begin implementation. Setting priorities was undertaken with participation from the Working Group and many of the participants in the National Forum meeting of 2002.32 For accomplishing this, a systematic review and prioritization of all proposed action steps was undertaken. In a second meeting of the National Forum, in April 2004, eight “concrete tasks” were adopted, one in each of six of the recommendation areas, and two in the seventh. For each of the priorities, a task group was based on the original Expert Panels
and was charged to complete the planning process and carry out each of the priorities. The product at this stage was a charge to each Task Group presenting the originally proposed action step, specific task, and expected outcome, with the supporting rationale and statement, “What success will look like.” Figure 22-3 represents by example the charge to the Monitoring and Evaluation: Surveillance Task Group. With this charge, related guidance, and some limited staff support, the group undertook the task and published its review and recommendations for strengthening CVD surveillance at national, state, and local levels as an AHA Scientific Statement in January 2007.33 Work is in progress to implement the recommendations, beginning with establishment of a national CVD surveillance unit within the Epidemiology and Surveillance Branch,
Monitoring and Evaluation: Surveillance Task Group Action:
Task:
Outcome:
Bring key partners and stakeholders together to address gaps in heart disease and strokerelated data systems Identify data requirements and gaps and propose remedies to insure optimum data collection, management, and reporting Presentation of proposals for improved heart disease and stroke-related health data systems
Rationale This task is intended to advance implementation of the Action Plan by addressing the need for improved cardiovascular health data systems. Such improvements are needed both to monitor more adequately the burden and disparities attributable to heart disease and stroke in the population as a whole and to create the potential for evaluating the impact of preventive programs and policies at the level of communities or larger units of observation.
What success will look like Proposals for improved data systems will include an inventory of the relevant existing data sources (e.g., those relating to cardiovascular events and conditions, risk factors, behaviors, underlying determinants, and current practices, programs, and policies) and their principal collective strengths and limitations; a listing of the most critical data elements that are lacking (e.g., incidence of heart disease and stroke, incidence of risk factors, estimates based on adequate sample sizes for population subgroups, etc.); and a proposed approach to filling these gaps (e.g., by strengthening existing systems or creating new ones (e.g., building in longitudinal components of NHANES, BRFSS, or YRBS, * increasing sample sizes in existing surveys, or establishing comprehensive surveillance of quality of life, events, risk factors, treatments, and other elements, in multiple sentinel communities). Additional considerations include estimated resource requirements (e.g., budget, personnel, training) for effective implementation and utilization of the enhanced data systems that are proposed. Finally, strategies to achieve implementation of the needed improvements should be addressed.
*NHANES, National Health and Nutrition Examination Survey; BRFSS. Behavioral Risk Factor Surveillance System; YRBS, Youth Risk Behavior Survey
Figure 22-3 Monitoring and Evaluation: Surveillance Task Group.
46894_CH22_657_678.qxd
2/25/10
4:16 PM
Page 669
OBSTACLES TO TAKING ACTION 669
Division for Heart Disease and Stroke Prevention, CDC. This illustration demonstrates how implementation can proceed through successive steps from a plan into a completed action. Institutionalization The second step to move the Action Plan forward was to institutionalize a means of long-term implementation. The National Forum appeared to be the most appropriate vehicle, with major support for its operation continuing to be provided by CDC. Rather than formalize its organization a priori, the National Forum evolved as functioning of the task groups developed and elements of a formal structure emerged from that experience. The Forum meets annually near the anniversary date of release of the Action Plan and in 2006 adopted an organizational plan illustrated in Figure 22-4. The foundation of the National Forum rests on the seven Implementation Groups (formerly task groups), with their respective Chairs serving as members of its governing body, the Coordinating Board. Additional members represent lead organizations and others with delegates elected by the full membership. Resource and Membership committees provide supporting functions, and a Leadership Council is planned for appointment in due course. The Action Plan recommendations were reviewed and reaffirmed by the National Forum in 2008, at the 5-year anniversary of their original publication.28
The Forum remains to date a voluntary membership organization operating under agreed principles and by-laws rather than as a legal entity. Its Vision is “Working together for a heart-healthy and stroke-free world”; its Mission is “To provide leadership and encourage collaboration among organizations committed to heart disease and stroke prevention.” A schematic view of the function of the Forum and its seven Implementation Groups in relation to the four goals for heart disease and stroke prevention is presented in Figure 22-5. One can consider, for example, the distinct communication functions and tasks required to serve each of the four goals or, alternatively, how each of the Implementation Groups can serve differently to achieve progress under goal 1 or any of the other three goals. With full future development of its membership, the National Forum for Heart Disease and Stroke Prevention may include the broad array of organizations and agencies whose collective interests align with each of the Implementation Groups and each of the four goals of prevention. By then, the agenda anticipated by the Action Plan will have reached the stage of full implementation.
OBSTACLES TO TAKING ACTION Action plans, policies, guidelines, and recommendations for CVD prevention are being implemented to varying degrees. The major declines in CHD death
National Forum for Heart Disease and Stroke Prevention Leadership Council Resource Committee
Coordinating Board Executive Committee
Membership Committee
Action Priorities Group Monitoring and Evaluation Group
Organizational Capacity Group
Regional and Global Collaboration Group
Communications Group
Policy Research Group Public Health Leadership and Partnership Group
Figure 22-4 National Forum for Heart Disease and Stroke Prevention.
46894_CH22_657_678.qxd
670
2/25/10
4:16 PM
Page 670
CHAPTER 22 TAKING ACTION
The Goals of Prevention Prevention of risk factors
X X X
X
X
PH leadership
X
X
Communications
X
2
X
al Go
X
X
X
1 al
X
X
X
X
X
X
3
Action priorities
X
X
X
X
X
X
al
Organizational capacity
Go
X
Go
Monitoring & evaluation
X
X
4
Policy research
X
Prevention of recurrent cardiovascular events
al
Regional & global collaboration
Early identification and treatment of heart attacks and strokes
Detection and treatment of risk factors
Go
The scope of CVD prevention: 7 action areas ⫻4 goals
Figure 22-5 The Scope of CVD Prevention: 7 Action Areas 4 Goals.
rates in the United States and several other countries have been attributed about equally to populationwide shifts in risk-factor distributions and to application of treatment of persons with CVD. Both of these effects indicate real impacts of prevention. Yet strategies of prevention are underutilized. Prevention lags, from the United States and other industrial nations that are far along in experience of epidemic CVD to the low- and middle-income countries where the epidemic, though long-established, is only recently recognized. What impedes full implementation of existing knowledge about the urgency of the problem, the great potential of prevention, and the cost of inaction? It appears that obstacles remain that must be surmounted if taking action is to become fully effective. The goal of prevention of the atherosclerotic and hypertensive cardiovascular diseases globally is to reduce the morbidity, mortality, disparities, and costs— both social and economic—of these conditions. In the United States, the national goal for heart disease and stroke prevention spans the course of CVD from prevention of risk factors to prevention of recurrent cardiovascular events.31 The central issues in prevention of atherosclerotic and hypertensive diseases concern not the scientific basis for sound recommendations nor the availability of policy statements and guidelines at both community- or population-wide and high-risk levels, but rather the actual implementation of these by the organizations, agencies, or health professionals on whom implementation de-
pends. The apparent obstacles include divergent opinions, competing interests, issues of population diversity and heterogeneity, limitations of the policy framework, and failure to commit needed resources. Population-Wide and Community Intervention Divergent Opinions. Opponents of community- or population-wide intervention strategies can cite conflicts of opinion found in the scientific literature. However, these are perhaps among the least important obstacles to effective prevention. For example, Oliver’s feature article in Circulation in the mid-1980s protested: “Cardiologists and physicians throughout the world are being persuaded by health educationalists, and some epidemiologists who have also assumed this role, that the only really effective way to prevent coronary heart disease (CHD) is to endorse and promulgate changes in lifestyle of the public at large. . . . Much of the faith in the value of changing lifestyles is little more than wishful thinking.”34, p 1 Oliver mischaracterized the epidemiologic or “health educationalist” approach as exclusively populationwide in its strategy and advocated instead a focus on prevention of precipitating conditions immediately preceding acute coronary events, an approach that is of theoretical interest but has limited applicability. He further argued against continued investment in large-scale trials of population-wide interventions, although he identified such trials as the only source of convincing evidence of the effectiveness of preventive measures.
46894_CH22_657_678.qxd
2/25/10
4:16 PM
Page 671
OBSTACLES TO TAKING ACTION 671
Rose emphasized the view that only those interventions that tended to restore “biological normality,” such as improved dietary balance, increased physical activity, and cessation of smoking, could be strongly advocated as elements of population-wide interventions in the absence of extensive experimental evidence of safety as well as effectiveness.35 Rose also addressed what he characterized as the “prevention paradox”:35, p 1850 We arrive at what we might call the prevention paradox—“a measure that brings large benefits to the community offers little to each participating individual.” It implies that we should not expect too much from individual health education. People will not be motivated to any great extent to take our advice, because there is little in it for each of them, particularly in the short and medium term. Change in behaviour has to be for some the larger and more immediate reward. From these comments, it seems the “educationalists” and epidemiologists cannot fairly be charged with lack of circumspection about the nature of appropriate population interventions or with altogether unrealistic views of what may be achieved. Waiting to rely on intervention among persons already experiencing acute coronary events or their immediate precipitating conditions ignores both the high proportion of persons dying early in the course of the initial event and the inability to reverse entirely the high risk of recurrence and death among those who survive. For these reasons and because of the increasing evidence of the effectiveness of population interventions, these strategies remain widely advocated despite the opinions illustrated here. Competing Interests. In some areas central to policy development and implementation, competing interests exert strong influence. For example, James and Ralph noted in discussion of strategies for dietary change at the national level, with particular reference to the United Kingdom, “We are now in a position where the public health priorities are clear, but where the food policies of the government and of the farming and food industries are geared to completely different goals.”36, pp 524–528 The theme of the review was that “informed choice” as the sole determinant of individual action concerning diet is illusory in view of the many pressures on consumers that result in “the consumer’s dilemma”—the inability to make an informed “free” choice. The authors rejected as inappropriate the argument that governmental nutrition policy concerning health should be limited to a role of providing infor-
mation. Governmental policy development is needed that includes food protection, agricultural strategies, nutrition education, and food labeling, among other components. According to James and Ralph, “Analyses of different policies suggest that health issues are readily squeezed out of discussion by economic and vested interests unless able promoters of the health issues are involved in the discussions.”36, p 537 A specific example is the conflict between salt manufacturers and governmental policies in the United Kingdom, illustrated in Chapter 8 by reference to a revealing set of papers and commentaries that appeared in the British Medical Journal in May 1996.37 Population Diversity and Heterogeneity. Populations are diverse in many respects that may be potentially relevant to population-wide intervention. However, the nature and extent of such diversity and the need to tailor prevention policies for different groups are less clear. In the United States, development of culturally sensitive intervention programs has long been advocated. The Stanford Five City Project provided information on responsiveness to community-wide intervention, as measured by change in a composite, multivariate risk score for all-cause mortality among several subgroups of the population.38 The authors concluded that specific interventions should be developed for different age, socioeconomic, and cultural subgroups. On the international level, Janus and colleagues addressed the impact of modernization in Asia and its implications for the occurrence of CHD in that region.39 Population diversity with respect to coronary heart disease and its risk factors was illustrated by examples from Taiwan, the Philippines, Malaysia, Indonesia, and India. The premise of the review was that thorough knowledge of a region is required for development of acceptable and effective prevention programs, and the conclusion was that “a single prevention strategy for all of Asia may be inappropriate.”39, p 2671 In particular, it was suggested that recommendations developed in Europe, North America, or Australia might require substantial modification for different Asian countries. Importantly, however, “broader statements on exercise, obesity, smoking, and diet and on the prevention and control of diabetes and hypertension might be appropriate.”39, p 2673 Findings of the INTERHEART Study offer increased confidence that the causal factors are much the same across a very wide range of populations and support this latter view.40 Moreover, it would be contrary to principles of public health and public policy to focus on differences among specific groups or individuals to the neglect of common characteristics that call for and justify broad and inclusive intervention strategies. Within any population, how-
46894_CH22_657_678.qxd
672
2/25/10
4:16 PM
Page 672
CHAPTER 22 TAKING ACTION
ever, the details of intervention must nevertheless take local contextual factors into account. An extreme proposition is that lipid-lowering therapy should be individualized on the basis of genetic heterogeneity. A requirement that treatment be based on individual pretreatment genetic typing could undermine public health recommendations altogether.41 If individual-level genetic assessment were needed to evaluate potential risks and benefits of interventions, little might remain of population-wide recommendations. The effect would be a conceptual atomization of populations as mere aggregates of individuals for whom no common action could be proposed. Some vigilance may be required to protect the public’s health and public health policy from this reductionist view. Limited Scope of the Policy Framework for Intervention. Prevention policy for the atherosclerotic and hypertensive diseases is well established on the basis of scientific knowledge that was available 40 years ago and that has been greatly strengthened since then. At the broadest levels, however, as suggested by James’s observations concerning food and nutrition policy, the framework is incomplete because many large societal questions have not yet been resolved. Therefore consistency of policies across the spectrum from land use and food production to recommended dietary behaviors of individuals has not yet been attained. The unusual case of Poland’s removal of tax subsidies for animal products in the national food supply and ensuing sharp decline in CHD mortality was noted in Chapter 21. Similarly, Roemer’s review of tobacco legislation in developing countries suggests important gaps and implies that these will remain in any particular country until national data are available to provide the needed justification.42 Thus despite much support for existing policies, further policy development is needed to enhance the implementation of currently proposed interventions. Failure to Commit Needed Resources. Resource limitations are another obstacle to effective action. It might be assumed that this would be a less serious factor in the United States than elsewhere. But in 1997 the National Center for Chronic Disease Prevention and Health Promotion, CDC, reported the following:43, p 12 [T]he nation’s public health system framework is severely underdeveloped to address the tremendous burden of chronic disease. . . . Coordinated and comprehensive national chronic disease prevention efforts have not been nearly adequately or systematically applied.
Seven years ago, in 1989, only $245 million— less than 3%—of the $9.5 billion spent by state health agencies was directed toward the prevention and control of chronic disease. Of the 48,000 full-time employees in state health agencies that year, only 2% were employed in chronic disease programs. In 1996, the very small percentage of resources reflected in these numbers remains essentially unchanged. Without strong, well-coordinated state-based programs aimed at chronic disease and supported by essential national elements, state and local health departments, and indeed, this nation, cannot hope to address the current burden of chronic disease, and efforts will increasingly fall short as the population ages. To the extent that investment in cardiovascular disease prevention elsewhere is proportionately less than that in the United States, significant challenges remain to approach the minimum resource requirements for effective public health action. Adding to the discussion of resources is the matter of costs of intervention. Two contributions illustrate current views on this issue. As reviewed in Chapter 21 cost-effectiveness analysis identifies two populationwide interventions (increasing taxes on tobacco products and reducing sodium content of manufactured foods) and one high-risk level intervention (combination low-dose medication to reduce risk factors in high-risk individuals) as practical for implementation in low- and middle-income countries.44 Trust for America’s Health, a nongovernmental organization supporting disease prevention policies, estimates that (available at www.healthyamericans .org; accessed September 1, 2008):9, p 3 an investment of $10 per person per year in proven community-based disease prevention programs could yield net savings of more than $2.8 billion annually in health care costs in one to 2 years, more than $16 billion annually within 5 years, and nearly $18 billion annually in 10 to 20 years (in 2004 dollars). With this level of investment, the country could recoup nearly $1 over and above the cost of the program for every $1 invested in the first one to 2 years of these programs, a return on investment (ROI) of 0.96. Within 5 years, the ROI could rise to 5.6 for every $1 invested and rise to 6.2 within 10 to 20 years. This return on investment represents medical cost savings only and does not include the significant gains that could be achieved in worker productivity, re-
46894_CH22_657_678.qxd
2/25/10
4:16 PM
Page 673
CURRENT ISSUES 673
duced absenteeism at work and school, and enhanced quality of life. Individual and High-Risk Intervention Many obstacles operate at the individual or high-risk level, as at the population level, to impede implementation of prevention policy. They have much in common with those discussed previously. Divergent Opinions. The idea that health promotion in clinical practice is unethical was advanced in The Lancet by McCormick in the mid-1990s. His commentary was in opposition to an announced policy for general practice in the United Kingdom that proposed attention to blood pressure, smoking, body mass index, alcohol consumption, family history, diet, and physical activity.45 His premise was that this activity constituted unsolicited intrusion into the lives of patients and lacked necessary support of “conclusive evidence that screening can alter the natural history of disease in a significant proportion of those screened,”45, p 390 a criterion attributed to Cochrane and Holland. He concluded, “Health promotion as encouraged by [the recommendations being addressed] falls far short of meeting the ethical imperatives for screening procedures, and moreover diminishes health and wastes recourse [sic]. General practitioners would do better to encourage people to live lives of modified hedonism, so that they may enjoy, to the full, the only life that they are likely to have.”45, p 391 This laissez-faire recommendation, of course, also lacks conclusive evidence. Limited Influence of Guidelines. The concern that published guidelines for preventive services often have less than intended impact was a special focus of the 1995 Bethesda Conference report and has been addressed extensively elsewhere.46,47 Four types of barriers were identified in that report, related to the patient, the physician, healthcare settings, and the community or society at large. These included, among others, lack of knowledge and motivation, lack of access to care, cultural and social factors, lack of policies and standards, and lack of reimbursement. Questions of Cost. Increasingly, in the present US healthcare situation and elsewhere, evaluation of procedures in terms of cost has intensified. The example of cholesterol lowering has been addressed in numerous reports, one of which illustrates the point:48, pp 329–330 Although preventive care is intuitively appealing and is often advocated as a means to reduce health care costs, formal economic analyses demonstrate that, similar to most preventive
care, cholesterol lowering for primary coronary prevention does not “pay for itself.” Nonetheless, most analyses suggest that drug treatment for young and middle-aged men with moderate-to-severe elevations of serum cholesterol ( 240 mg/dl) and multiple other risk factors for CHD has a cost-effectiveness ratio below $40,000 per year of life saved—similar to federally funded programs such as outpatient hemodialysis and many other widely practiced medical interventions. . . . The appropriateness of cholesterol reduction for other populations, including young men with isolated mild hypercholesterolemia, women, and the elderly is less certain, however. A more recent analysis of clinical interventions came to a similar conclusion:49, p 576 Aggressive application of nationally recommended prevention activities could prevent a high proportion of the CAD [coronary artery disease] events and strokes that are otherwise expected to occur in adults in the United States today. However, as they are currently delivered, most of the prevention activities will substantially increase costs. If preventive strategies are to achieve their full potential, ways must be found to reduce the costs and deliver preventive activities more efficiently. The dependence of these estimates on the costs of lipid-lowering agents has been emphasized.50 The tendency to identify intervention to lower cholesterol with use of drugs often clouds the question of whether intervention is justified in terms of cost or potential risks of side effects. Until more extensive experience is accumulated to indicate that substantial reductions in cholesterol concentration can be achieved even in high-risk persons by nonpharmacologic means, this issue will probably remain an obstacle to policies for intervention in high-risk persons except, perhaps, those with already-manifest coronary heart disease.
CURRENT ISSUES Economics, health systems, and policy priorities are central issues in determining what action will be taken to prevent CVD and other chronic diseases. Current discussions about global health commonly focus on one or more of these issues. Similarly, they were recognized in the Action Plan in 2003 as three “strategic imperatives,” to be addressed in order for the plan to be fully implemented:3
46894_CH22_657_678.qxd
674
2/25/10
4:16 PM
Page 674
CHAPTER 22 TAKING ACTION
1. strike a new balance in our investment in health, by putting prevention first; 2. transform our public health agencies into effective instruments for policy and environmental change, supporting the entire range of public health approaches for heart disease and stroke prevention; and 3. prevent the causes of heart disease and stroke––no longer waiting to treat the causes or their consequences, while their prevention was possible. More recent reflection on these issues suggests some refinements in thinking about the strategic imperatives. Both Healthy People 2010 and the Action Plan point to four levels of prevention. They reach from the “upstream” prevention of risk factors to detection and treatment of risk factors, early identifica-
tion and treatment of heart attacks and strokes, and the furthest “downstream” prevention of recurrent CVD events among those who have survived a first one. They represent a “continuum of care” that extends from populations to patients, from the broadest population-wide prevention-oriented policies and programs to the most individualized practices and rescue victims of CVD. This view of a continuum is intended to demonstrate the nexus linking public policy, public health, and community interventions on the one hand with health care, management guidelines, and clinical practice on the other. Figure 22-6 completes the representation of the “action framework” by incorporating this fundamental aspect. From this perspective the issues of economics, health systems, and policy priorities can be revisited. “Putting prevention first” in the context of investment in health is a strategy for sharpening the fo-
Action Framework for a Comprehensive Public Health Strategy To Prevent Heart Disease and Stroke A Vision of the Future Social and Environmental Conditions Favorable to Health
Behavioral Patterns Promote Health
Policy and Environmental Change
Behavior Change
‘upstream’
Few Events/ Only Rare Deaths
Low Population Risk
Continuum of care Risk Factor Detection and Control
Emergency Care/Acute Case Management
Intervention Approaches Unfavorable Social and Environmental Conditions
Adverse Behavioral Patterns
The Present Reality First Event/ Major Risk Sudden Factors Death
Full Functional Capacity/ Low Risk of Recurrence
Good Quality of Life Until Death
‘downstream’ Rehabilitation/ Long-term Case Management
End-ofLife Care
Fatal CVD Complications/Decompensation
Disability/ Risk of Recurrence
The Healthy People 2010 Goals Increase Quality and Years of Healthy Life Eliminate Disparities Goal 1
Prevention of risk factors
Goal 2
Goal 3
Detection and treatment of risk factors
Early identification and treatment of heart attacks and strokes
Goal 4
Prevention of recurrent cardiovascular events
Figure 22-6 Action Framework for a Strategy to Prevent Heart Disease and Stroke. Source: Adapted from US Department of Health and Human Services. A Public Health Action Plan to Prevent Heart Disease and Stroke. p 6. Atlanta, GA: US Department of Health and Human Services, Centers for Disease Control and Prevention; 2003.
46894_CH22_657_678.qxd
2/25/10
4:16 PM
Page 675
REFERENCES 675
cus on prevention at all levels, but especially on prevention of the acute and irreversible coronary and cerebrovascular events with their toll of death, disability, disparities, and cost. Both detection and treatment of risk factors and prevention of risk factors have this potential, although having established risk factors and their accompanying pre- or subclinical consequences already entail disability, disparities, and cost. Prevention of risk factors in the first place is necessary to arrest the progression from optimum cardiovascular health into the state of increased risk and especially to protect future generations from following the course of past and present ones. This is the area where our under-investment in health is most critical. “Supporting the entire range of public health approaches for heart disease and stroke prevention” depends importantly on transforming our public health agencies. New capacities are needed for policy development and implementation and for effective leadership in creating and sustaining broad-based, multisector partnerships and collaborations for prevention of CVD and other chronic diseases. But pressures are intensifying, especially in the United States, for healthcare transformation in order to resolve problems of cost, insurance coverage, access, quality, and equity in the healthcare arena. From the perspective of the continuum of care, a focus on health care to the neglect of public health would be seriously incomplete. “Health system transformation” is therefore advocated by many as the needed focus. Progress would, in this view, extend beyond payment mechanisms that support the existing healthcare apparatus (not in the usual sense a “system” at present). Rather, it would incorporate the broader issues of prevention that put upstream strategies up front. Public health agencies and related organizations must engage effectively for health system transformation at this level to succeed. “No longer waiting to treat the causes or their consequences, while their prevention was possible” calls again for attention to prevention of the risk factors in the first place. A high priority is to preserve optimum cardiovascular health through effective upstream approaches—policy and environmental change and population-wide behavior change that promote health and protect from adverse influences. Prevention of risk factors must begin in childhood and adolescence (or even in utero) and continue throughout life to have fullest effect. An integrated strategy of investing in prevention, establishing a coherent health system, and focusing on promoting health as a first priority offers the best prospect for achieving the goals of preventing CVD and other chronic diseases.
A final consideration concerns monitoring and evaluation of cardiovascular health, at the population level, in order to know the extent to which the challenge of CVD prevention is being met: Is the plan being fully realized, with the expected impact on the health of the population? REFERENCES 1. Eyre H, Kahn R, Robertson RM, and the ACS/ ADA/AHA Collaborative Writing Committee. Preventing cancer, cardiovascular disease, and diabetes. A common agenda for the American Cancer Society, the American Diabetes Association, and the American Heart Association. Circulation. 2004;109:3244–3255. 2. Ebrahim S. Chronic diseases and calls to action. Int J Epid. 2008;37:225–230. 3. US Department of Health and Human Services. A Public Health Action Plan to Prevent Heart Disease and Stroke. Atlanta, GA: US Department of Health and Human Services, Centers for Disease Control and Prevention; 2003. 4. American Heart Association. American Heart Association strategic goals. Available at www.americanheart.org/presenter.jhtml? identifier=4429. Accessed February 11, 2008. 5. Pearson TA, Bazzarre TL, Daniels SR, et al. American Heart Association guide for improving cardiovascular health at the community level: a statement for public health practitioners, healthcare providers, and health policy makers from the American Heart Association Expert Panel on Population and Prevention Science. Circulation. 2003;107:645–651. 6. Veazie MA, Galloway JM, Matson-Koffman D, et al. Taking the initiative. Implementing the American Heart Association guide for improving cardiovascular health at the community level. Circulation. 205;112:2538–2554. 7. Partnership to Fight Chronic Disease. About the PFCD. Available at www.fightchronic disease.org/about/index.cfm. Accessed September 21, 2008. 8. Partnership for Prevention. Partnership news and updates. Available at www.prevent.org. Accessed September 21, 2008.
46894_CH22_657_678.qxd
676
2/25/10
4:16 PM
Page 676
CHAPTER 22 TAKING ACTION
9. Trust for America’s Health. Prevention for a Healthier America: Investments in Disease Prevention Yield Significant Savings, Stronger Communities. July 2008. Available at www.healthyamericans.org. Accessed September 21, 2008.
16. World Health Organization. Bridging the Health Gap in Europe. A Focus on Noncommunicable Disease Prevention and Control. The CINDI-EUROHEALTH Action Plan. Copenhagen (Denmark): WHO Regional Office for Europe; 1995.
10. Robinson K, Farmer T, Elliott SJ, Eyles J. From heart health promotion to chronic disease prevention: contributions of the Canadian Heart Health Initiative. Prev Chron Dis. 2007;4: A29. Available at www.cdc.gov/pcd/issues/ 2007/apr/06_0076.htm. Accessed September 21, 2008.
17. World Health Organization. Gaining Health: The European Strategy for the Prevention and Control of Noncommunicable Diseases. Copenhagen (Denmark): WHO Regional Office for Europe; 2006.
11. Blair L, MacLean D, National Evaluation Steering Group and Writing Group. Canadian Heart Health Initiative: Process Evaluation of the Demonstration Phase. Conference of Principal Investigators of Heart Health. Available at www.med.ca/chhdbc/. Accessed September 21, 2008. 12. Canadian Heart Health Strategy and Action Plan. About the CHS-AP. Available at www .chhs-scsc.ca/web/?page_id=16. Accessed September 21, 2008. 13. Ministers of Health of the Americas. Health Agenda for the Americas 2008–2017. Presented by the Ministers of Health of the Americas in Panama City, June 2007. Available at www.paho.org. Accessed September 1, 2008. 14. Ministry of Health and Social Development of the Russian Federation and State Research Center for Preventive Medicine. Strategy for the Prevention and Control of Noncommunicable Diseases and Injuries in the Russian Federation. Moscow, Russian Federation: Ministry of Health and Social Development of the Russian Federation and State Research Center for Preventive Medicine; 2008. 15. Grabowsky TA, Farquhar JW, Sunnarborg KR, Bales VS. Worldwide Efforts to Improve Heart Health: A Follow-up to the Catalonia Declaration––Selected Program Descriptions. Atlanta, GA: US Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion; 1997.
18. National Action Plan for Prevention and Control of Non-Communicable Diseases and Health Promotion in Pakistan. Islamabad, Pakistan: tripartite collaboration of the Ministry of Health, Government of Pakistan; WHO, Pakistan office; and Heartfile; 2004. 19. South Asian Association for Regional Cooperation. A brief on SAARC. Available at www.saarc-sec.org/main.php?id=76. Accessed January 21, 2008. 20. Nishtar S. Preventing Coronary Heart Disease in South Asia. SAARC Cardiac Society Guidelines and Recommendations. Islamabad, Pakistan: SAARC Cardiac Society and Heartfile, 2002. 21. International Heart Health Society. International Action on Cardiovascular Disease: A Platform for Success. Vancouver: International Heart Health Society; 2005. 22. World Heart Federation. African Heart Network gathers and grows. Heart Beat; June 2007. Available at www.world-heartfederation.org/publications/heart-beate-newsletter. Accessed January 21, 2008. 23. World Health Organization. Cardiovascular Disease Prevention and Control. Translating Evidence into Action. Geneva: World Health Organization; 2005. 24. World Health Organization. Prevention and Control of Noncommunicable Diseases: Implementation of the Global Strategy. Report by the Secretariat. A61/8. Geneva: World Health Organization; 2008.
46894_CH22_657_678.qxd
2/25/10
4:16 PM
Page 677
REFERENCES 677
25. Jamison DT, Breman JG, Measham AR, et al., eds. Priorities in Health. Washington DC: The World Bank; 2006. 26. Leeder S, Raymond S, Greenberg H. A Race Against Time: The Challenge of Cardiovascular Disease in Developing Countries. New York: The Trustees of Columbia University in the City of New York; 2004. 27. Labarthe DR. Heart-healthy and stroke-free, 2008. Prev Chron Dis. 2008;5(2). Available at: www.cdc.gov/pcd/issues/2008/apr/07_0259 .htm. Accessed May 1, 2008. 28. US Department of Health and Human Services. Update to a Public Health Action Plan to Prevent Heart Disease and Stroke. Atlanta, GA: Centers for Disease Control and Prevention; 2008. 29. CVD Plan Steering Committee. Preventing Death and Disability From Cardiovascular Diseases: A State-Based Plan for Action. Washington, DC: Association of State and Territorial Health Officials; 1994. 30. US House Committee on Appropriations. Conference report to accompany H.R. 2264. Making appropriations for the Departments of Labor, Health and Human Services, and Education, and related agencies for the fiscal year ending September 30, 1998, and for other purposes. (H. Rpt. 105-390.) Washington, DC: Government Printing Office; 1997. 31. US Department of Health and Human Services. Healthy People 2010. 2nd ed. With Understanding and Improving Health and Objectives for Improving Health. 2 vols. Washington, DC: US Government Printing Office; 2000. 32. Labarthe DR, Biggers A, Goff Jr DC, Houston M. Translating a plan into action. A Public Health Action Plan to Prevent Heart Disease and Stroke. Am J Prev Med. 2005;29(5S1): 146–151. 33. Goff Jr DC, Brass L, Braun LT, et al. Essential features of a surveillance system to support the prevention and management of heart disease and stroke. A Scientific Statement from the American Heart Association Councils on Epidemiology and Prevention, Stroke, and
Cardiovascular Nursing and the Interdisciplinary Working Groups on Quality of Care and Outcomes Research and Atherosclerotic Peripheral Vascular Disease. Circulation. 2007;115:127–155. 34. Oliver MF. Prevention of coronary heart disease—propaganda, promises, problems, and prospects. Circulation. 1986;73:1–8. 35. Rose G. Strategy of prevention: lessons from cardiovascular disease. BMJ. 1981;282: 1847–1851. 36. James WPT, Ralph A. National strategies for dietary change. In: Marmot M, Elliott P, eds. Coronary Heart Disease Epidemiology: From Aetiology to Public Health. Oxford (England): Oxford Medical Publications; 1992:525–540. 37. Godlee F. The food industry fights for salt. BMJ. 1996;312:1239–1240. 38. Winkleby MA, Flora JA, Kraemer HC. A community-based heart disease intervention: predictors of change. Am J Public Health. 1994;84:767–772. 39. Janus ED, Postiglione A, Singh RB, Lewis B, Council on Arteriosclerosis of the International Society and Federation of Cardiology. The modernization of Asia: implications for coronary heart disease. Circulation. 1996;94: 2671–2673. 40. Yusuf S, Hawken S, Ôunpuu S, et al. Effect of potentially modifiable risk factors associated with myocardial infarction in 52 countries (the INTERHEART study): case-control study. Lancet. 2004;364;937–952. 41. Omenn GS. Comment: genetics and public health. Am J Public Health. 1996;86: 1701–1704. 42. Roemer R. Legislative Action to Combat the World Smoking Epidemic. Geneva, Switzerland: World Health Organization; 1982. 43. Report of the National Center for Chronic Disease Prevention and Health Promotion, Centers for Disease Control and Prevention. Unrealized Prevention Opportunities: Reducing the Health and Economic Burden of
46894_CH22_657_678.qxd
678
2/25/10
4:16 PM
Page 678
CHAPTER 22 TAKING ACTION
Chronic Disease. Bethesda, MD: Public Health Service, US Department of Health and Human Services; 1997. 44. Gaziano TA, Galea G, Reddy KS. Chronic diseases 2. Scaling up interventions for chronic disease prevention: the evidence. Lancet. 2007;370:1939–1945. 45. McCormick J. Health promotion: the ethical dimension. Lancet. 1994;344:390–391. 46. Pearson TA, Fuster V. Executive summary. J Am Coll Cardiol. 1996;27:957–1047. 47. Pearson TA, McBride PE, Miller NH, Smith SC Jr. Task Force 8: organization of preventive cardiology service. J Am Coll Cardiol. 1996; 27:1039–1047.
48. Cohen DJ, Goldman L, Weinstein MC. The cost-effectiveness of programs to lower serum cholesterol. In: Rifkind BM, ed. Lowering Cholesterol in High Risk Individuals and Populations. New York: Marcel Dekker, Inc.; 1995:311–336. 49. Kahn R, Robertson RM, Smith R, Eddy D. The impact of prevention on reducing the burden of cardiovascular disease. Circulation. 2008;118: 576–585. 50. Lloyd-Jones DM. Is an ounce of prevention worth a pound of cure? Or does it take 16 ounces of prevention (or more)? AHA Communities Learning Library, July 7, 2008. Available at http://pt.wkhealth.com/pt/re/aha/ addcontent. Accessed July 10, 2008.
46894_CH23_679_696.qxd
2/25/10
4:17 PM
Page 679
C H A P T E R
23 Epidemiology and a CVD Prevention Research Agenda dependent on collaboration with other disciplines today than previously. How should epidemiology be understood today in view of its expanding horizon, growing complexity, and potential for greater impact on population health? It has been suggested that “black box” epidemiology (focusing on specific risk factors in isolation) has given way to “Chinese box” epidemiology (putting specific observations into their hierarchical social context). But further reflection suggests still another developmental stage. First, from a global perspective on CVD and other chronic disease prevention, the classic epidemiologic dimensions of person, place, and time require consideration on a much larger than usual scale. An example of this view is the theory of epidemiologic transition, with its global reach over decades. Second, the recognized value of collaboration across scientific disciplines, communication between science and other societal interests, and partnerships extending beyond health to reach all sectors, call for consideration of “connectivity.” Effective connections of these kinds may require a new outlook and unfamiliar skills on the part of epidemiologists but can greatly enhance the relevance and impact of our work. The goals of epidemiologic research can be summarized as understanding causes, identifying means of prevention, and monitoring populations to assess the burden of disease and impact of interventions. Strategies of epidemiologic investigation afford a range of approaches, among them cross-sectional population surveys, retrospective and prospective analytic studies, and intervention trials. Other relevant strategies include surveillance, program evaluation, policy analysis, modeling, and economic analysis. Newer areas of
SUMMARY The scope and complexity of CVD epidemiology are increasing, for several reasons. First, although the population is its ultimate concern, its interests extend across the full spectrum of biomedical and community health research. Epidemiology necessarily connects with laboratory and clinical research disciplines in its wide-ranging pursuits, as evidenced in Part III on determinants of atherosclerotic and hypertensive diseases. Second, social determinants of health are receiving new emphasis as being fundamental to causation and prevention of adverse health conditions, especially as they afflict disadvantaged groups within society. This emphasis calls for epidemiologic research extending further into social and behavioral sciences, economics, public policy, and other related fields. Third, epidemiologic research, with surveillance and program evaluation, are required to fulfill the three core functions of public health: to assess the health status of communities and populations, develop policies that will foster conditions in which people can be healthy, and assure that these policies are being implemented with the intended benefits. These roles require collaboration between epidemiology and other disciplines concerned with population health and public health practice. Fourth, in parallel with public health action for CVD prevention at local, national, regional, and global levels, epidemiologic research and evaluation of programs and policies are needed at every level, from community health to global health. For these reasons, research in epidemiology and prevention of cardiovascular diseases—including surveillance and evaluation—is relevant on a wider scale and still more
679
46894_CH23_679_696.qxd
680
2/25/10
4:17 PM
Page 680
CHAPTER 23 EPIDEMIOLOGY AND A CVD PREVENTION RESEARCH AGENDA
implementation science and public health services and systems research are also applicable. Proposals for an epidemiologic research agenda in CVD prevention come from multiple sources, including the National Institutes of Health, the World Health Organization, the International Heart Health Society, and others. A Public Health Action Plan to Prevent Heart Disease and Stroke emphasizes the importance of research that provides evidence for policy and other public health decision making. Capacity to implement the research agenda and sustain the proposed research depends on sponsorship and funding, training, and institutionalization. These key elements of infrastructure depend, in turn, on effective efforts by those who recognize and value the potential contributions of cardiovascular epidemiology. In view of the large gap between the current level of research and the need, it has been suggested that a social movement is required to develop sufficient support for research to meet the global challenge of CVD and other chronic disease prevention. Epidemiology is an essential component of a larger scientific enterprise. Recognition of this could provide a frame of reference for epidemiology regarding its place among related disciplines. It might also help to shape the research agenda through which epidemiology can make its best contributions to advancing population health. In the present period, described elsewhere as “the -omics era,” this wider enterprise might be designated “populomics”—a term to represent what may be thought of as “integrative human ecology,” with population health as its focus and epidemiology as its core scientific discipline.
INTRODUCTION A challenge was posed to the field of epidemiology by Remington, in a symposium honoring the late Geoffrey Rose on his retirement in 1991:1, p 517 [A]s epidemiology passes from aetiological investigations through longitudinal studies to community intervention programmes and evaluative trials of CHD control programmes, it recapitulates the several stages en-compassed in this publication. The concern of this chapter is with the far end of that transition—the role of organized public health in applying the fruits of epidemiological investigation to cardiovascular disease prevention in the community. This transition from research to policy formation to application has received too little attention
among epidemiologists and, for that matter, among professional public health workers. Yet the transition itself has much to do with the uses of epidemiology in improving community health. Put another way, in the language of the Institute of Medicine report, the substance of public health is considered to be “organized community efforts aimed at the prevention of disease and promotion of health. (Public health) links many disciplines and rests upon the scientific core of epidemiology.” Practical application is the principal focus here. Substantial accountability rests with epidemiology to apply what decades of research have previously established, that the major cardiovascular diseases are largely preventable. As Remington emphasizes, the transition from research through policy formation to application has largely been neglected by epidemiologists. Within contemporary epidemiology as a whole, the scope of research interest has been considered previously as very broad, reaching from molecules to societies (see Figure 17-4). Much of current epidemiologic research concerns mechanisms of disease at the molecular level, and such research seems likely to continue. Remington’s observation implies different questions: What could epidemiologic research contribute to policy development? Is this occurring? What impact is policy development having on population health? These practical questions should be strongly influential in shaping a research agenda for CVD prevention. Such questions were anticipated when recommendations for research priorities were developed in the Public Health Action Plan to Prevent Heart Disease and Stroke.2 The charge to the Expert Panel on Advancing Policy was to consider issues in policy development, adoption, or implementation where epidemiologic research could potentially make a decisive contribution. The underlying premise was that in certain areas of CVD prevention progress might be impeded either by lack of a sufficient science base or by inadequate synthesis and communication of existing evidence. In the latter case, a renewed evaluation of pertinent evidence might suffice. But in the former case, one or more “critical investigations” should be devised with the intent to resolve the policy issue and promote action in accordance with the research findings. For example, such a critical investigation could address the lack of widespread adoption of communitywide interventions to prevent atherosclerotic and hypertensive diseases beginning in childhood, as called for in several published recommendations and guide-
46894_CH23_679_696.qxd
2/25/10
4:17 PM
Page 681
CONCEPTS OF EPIDEMIOLOGY 681
lines. The research approach would be to assess current policy statements and their scientific foundation; consider what impediments to action could be resolved with new research; plan and conduct the research; and communicate the findings to those positioned to mobilize the indicated action. The generic questions posed in the preceding extend beyond a narrow understanding of epidemiologic research to include both surveillance and program evaluation as well as other approaches. These latter activities are sometimes distinguished from research. A commonly cited definition of public health surveillance is “the ongoing systematic collection, analysis, and interpretation of outcome-specific data for use in the planning, implementation, and evaluation of public health practice.”3, p 1 Public health practitioners generally are said not to consider surveillance as research, on grounds that its purpose is to support taking public health actions rather than to acquire generalizable knowledge. This distinction exempts surveillance activities of official public health agencies from encumbering regulations that apply to research, such as informed consent of participants and institutional review board approval, so long as individually identifiable information is not collected.4 (see also 5) Program evaluation has been described as “a systematic way to improve and account for public health actions by involving procedures that are useful, feasible, ethical, and accurate” and as “an essential organizational practice in public health.”6, p 1 Further, “Evaluation is the only way to separate programs that promote health and prevent injury, disease, or disability from those that do not; it is a driving force for planning effective public health strategies, improving existing programs, and demonstrating the results of resource investments. Evaluation also focuses attention on the common purpose of public health programs and asks whether the magnitude of investment matches the tasks to be accomplished.”6, p 34 “Program” in this context refers to “any organized public health action.”6, p 3 Evaluation is sometimes distinguished from research by its purposes of assessing programs and supporting decision making rather than testing prior hypotheses. While recognizing these distinctions in the present discussion, epidemiologic research, surveillance, and program evaluation are considered together within a larger set of approaches to research in CVD prevention. All of these contribute to a potential research agenda with its principal purpose of providing evidence for policy development, while supporting the two other core public health functions, assessment and assurance.7
CONCEPTS OF EPIDEMIOLOGY The Array of Biomedical and Community Health Sciences A scientific discipline is characterized by the array of phenomena with which it is concerned. In his 1980 essay, “To Advance Epidemiology,” Stallones depicted the biomedical sciences as arrayed across a scale of biological organization, ranging from the phenomena of “submolecular particles” to those of whole societies and from the discipline of molecular biology to epidemiology (see Figure 17-4).8 While noting discontinuities along this spectrum due to specialization in science, Stallones observed that the true underlying continuity “should be an integrating force, not a divisive one, and ultimately the knowledge accrued by different subspecialties must all fit together without internal contradictions.”8, p 70 The sphere of interest of epidemiology was shown in his depiction to reach across four levels of organization, from individuals to families, communities, and societies. The individual, at the midpoint of the spectrum from particles to societies, was seen as the focal point for integration of all of the biomedical sciences, “and the totality of biomedical research should be coherent, with different disciplines sustaining and supporting each other.”7, p 70 Developments to the mid-1990s suggested some revisions to this scheme. For example, Rose argued persuasively for much greater prominence, indeed primacy, of the population level in thinking about the impact of preventive measures.9 Determinants of differences in rates of disease between populations are, under his argument, the foremost concern of public health. This is because the impact of interventions at the population level can be far broader than that of individual-level, patient-oriented interventions. Epidemiology that is limited to the individual level may be irrelevant to the larger forces of disease causation at the population level and therefore incapable of contributing to achievement of public health goals. It is true that population-wide and high-risk intervention strategies have been widely acknowledged as complementary. However, from the public health perspective, the population as a whole is the primary point of reference and the principal orientation of its core functions of assessment, policy development, and assurance. Consequently, adequacy of the term “biomedical” to represent the full array of health research has been questioned. The concern is that the essential contributions of research at the societal or community level may be undervalued in an era of exuberant expectations of laboratory research, for
46894_CH23_679_696.qxd
682
2/25/10
4:17 PM
Page 682
CHAPTER 23 EPIDEMIOLOGY AND A CVD PREVENTION RESEARCH AGENDA
example, in genetics and molecular biology. Thus, in the report of the first 10 years’ experience in CDC’s Prevention Research Centers Program, the expression “biomedical and community health research” was proposed to connote the full research spectrum.10 From “Black Box” to “Chinese Boxes” A further development of the 1980s and 1990s was growing recognition of fragmentation among the health sciences, especially epidemiology. This development was cited by Susser and Susser in a two-part essay on the future of epidemiology.11,12 They argued that: (1) Epidemiology in previous eras was rooted in public health; (2) the latter half of the 20th century focused chiefly on individual risks, out of context of the population as a whole; and, therefore, (3) epidemiology had evolved to be out of touch with public health. They proposed a “different paradigm” that has close parallels with the thought of Stallones and Rose and with evolution of cardiovascular epidemiology in the most recent decades:12, pp 674, 677 Encompassing many levels of organization— molecular and societal as well as individual—this paradigm, termed Chinese boxes, aims to integrate more than a single level in design, analysis and interpretation. Such a paradigm could sustain and refine a public health-oriented epidemiology. But preventing a decline in this new era will require more than a cogent scientific paradigm. Attention will have to be paid to the social processes that foster a cohesive and humane discipline. . . . Without intense socialization and learning, we may well find—because of the natural momentum and narrow focus that specialization generates—that the links between the values of public health and its specialized disciplines dissolve as we watch. In this respect, epidemiology is most certainly at risk. The Chinese boxes symbolized for the Sussers an advance, or perhaps a recovery, from the “black box” paradigm of previous decades. In this period, the authors had seen a profusion of epidemiologic research on “decontextualized risk factors” that had concluded with finding associations between exposures and outcomes but failed to integrate these findings into a larger picture of social phenomena. The image of a multilevel hierarchy of different levels of organization, represented by nested Chinese boxes, conveys the idea of their proposed paradigm. In their concept, epidemiology could gain breadth by exploiting modern information systems to communicate within and across levels of organization. Greater depth could result from incorporating new biomedical techniques.
This view of epidemiology, consistent with the illustration in Figure 17-4, incorporates the multilevel concept common to the views of Stallones, Rose, and the Sussers. It also indicates a preeminent place for society as a whole and shows the territory of epidemiology extending across the whole array as a basic discipline of biomedical and community health research. Whether this is a sufficient representation of contemporary epidemiology will be considered further as follows.
GOALS Why Do Epidemiology? The goals of cardiovascular epidemiology and prevention can be characterized as understanding causes, identifying means of prevention, and monitoring populations to assess both the burden of cardiovascular diseases and the impact of interventions to control them. Common reference to “epidemiology and prevention” points to a dual role and emphasizes that “epidemiology” alone, without “prevention,” is incomplete. Much has been said in preceding chapters about engagement of epidemiology and epidemiologists in development of guidelines, recommendations, and policies for CVD prevention. These activities, and their extension into communication about, and advocacy for, needed interventions in the interest of public health, are essential aspects of epidemiology and prevention. Contributing critical evidence and translating evidence into effective public health strategies are defining roles for epidemiology and prevention. Ultimately, the goal of epidemiology and prevention is achievement of better population health through disease prevention and health promotion, based on sound, relevant science. Outcomes of Epidemiologic Research What the cardiovascular epidemiologist does through research can be seen in part as continually adding new increments to Stallones’s “n-dimensional complex” of accumulated evidence described in Chapter 17. For reasons discussed there, this pursuit has scientific and intellectual value, even elegance, in itself. It represents the purely scientific or theoretical aspect of epidemiology. But to draw also on the Stallones metaphor of “directed pathways” within the complex, results of epidemiologic research can have major practical utility. Through this applied aspect, epidemiology contributes more directly to improving population health. Meeting the global challenge of preventing cardiovascular diseases depends importantly on progress in applied epidemiology, embracing the broad concept
46894_CH23_679_696.qxd
2/25/10
4:17 PM
Page 683
STRATEGIES OF INVESTIGATION 683
of research that includes surveillance and evaluation. The outcomes of this research extend from immediate study findings to syntheses of evidence; from resulting recommendations, guidelines, or policies to implementation strategies; and from impact assessments to updated syntheses of the evidence, taking practice experience into account. Dissemination of policies and programs may follow, with further research regarding effectiveness of interventions when adapted for particular population settings. The outcomes of epidemiologic research are illustrated throughout the preceding chapters, whether describing the global distribution of the major atherosclerotic and hypertensive diseases, epidemiologic evidence regarding their determinants, or translation of this evidence into recommendations, guidelines, policies, and public health strategies and action plans.
STRATEGIES OF INVESTIGATION Overview Epidemiologists are well aware of the range of conventional approaches to investigation that includes cross-sectional population surveys, retrospective and prospective analytic studies, and intervention trials. Each is useful and necessary for addressing one or another type of epidemiologic question, as demonstrated abundantly throughout Parts I–III. Other approaches are also needed that are less often considered within the purview of epidemiology. Surveillance, program evaluation, and policy analysis are described more fully here, and brief mention of implementation science and public health services and systems research are noted as well. Modeling and economic analysis were discussed in Chapter 21. Surveillance As addressed here, surveillance is viewed as integral to the topic of research for CVD prevention and control. Within A Public Health Action Plan to Prevent Heart Disease and Stroke, a priority task has been to assess the strengths and limitations of existing surveillance systems in the United States and recommend changes as appropriate. The National Forum for Heart Disease and Stroke Prevention, through its Monitoring and Evaluation Implementation Group, commissioned such an assessment that resulted in a comprehensive report with nine key recommendations for improvement.13 National, state, and local surveillance systems are critically reviewed, including the National Health and Nutrition Examination Survey, Behavioral Risk Factor Surveillance System, Youth Risk Behavioral Surveillance System, National
Health Interview Survey, and several others. Overarching recommendations concerned establishment of a National Heart Disease and Stroke Surveillance Unit, making major CVD events reportable conditions; strengthening current data collection practices and linkages between data systems; validating data from self-reported sources; and adopting variable sampling fractions among specific geographic or other defined populations of high priority to permit more adequate representation. Complemented by recommendations specific to the four Healthy People 2010 goals for heart disease and stroke prevention, these provisions for improved surveillance at local, state, and national levels would contribute greatly to research in CVD epidemiology throughout the United States. Concurrently, improvements in cardiovascular surveillance in Europe have been proposed by the Section on Prevention and Health Policy, European Association for Cardiovascular Prevention and Rehabilitation (EACPR).14 The European Cardiovascular Indicators Surveillance Set (EUROCISS) comprises operations manuals for three areas— population-based registers of both acute myocardial infarction and stroke, and cardiovascular surveys. The context of this proposal was as follows:14, p S1 Europe is now facing the challenge to implement preventive actions, identify persons in need of treatment, apply the European Guidelines for CVD Prevention in Clinical Practice and verify improving effectiveness. The development, testing and implementation of effective surveillance systems for CVD will produce reliable and comparable indicators, thus enabling policy makers to trace differences within and between countries and to make better decisions on planning and evaluation of prevention programs, healthcare delivery, resource allocation, and research. WHO has been active as well in developing and implementing a strategy to strengthen and standardize population-level data collection supporting noncommunicable disease prevention efforts globally.15 The WHO STEPwise approach to Surveillance (STEPS) program proposes an incremental strategy for developing national surveillance incorporating questionnaires at step 1, adding physical measurements at step 2, and further adding biochemical measurements at step 3. Each step begins with core data elements, with expanded scope and further optional increments as resources permit. Emphasis here is on the critical need for good information as a basis for decision making. At the same time, availability of such data permits population-level research that is otherwise impossible.
46894_CH23_679_696.qxd
684
2/25/10
4:17 PM
Page 684
CHAPTER 23 EPIDEMIOLOGY AND A CVD PREVENTION RESEARCH AGENDA
Program Evaluation Program evaluation, as defined previously, is recognizable at least implicitly in some or all of the research agendas described here. It was the subject of a 1995 monograph, Evaluating Community Efforts to Prevent Cardiovascular Diseases.16 That report noted the complexity of evaluating community programs conducted by multiple partner organizations and individuals, owing to the multifaceted activities often undertaken. The following core questions for evaluation were presented:16, p 14 • Was the community mobilized to reduce risks for CVD? • What changes in the community resulted from the initiative? • Is there a change in behavior related to risks for CVD? • Does the initiative have a community-level outcome related to risks for CVD? • Is community-level outcome related to changes facilitated by the initiative? The nature of these questions combines qualitative assessments of program processes with measures of community changes and outcomes that require attention to program design, data collection, and appropriate analysis and interpretation. These latter activities merge into research, making the distinction of program evaluation seem exaggerated. One approach to program evaluation that is widely used is the RE-AIM framework discussed in Chapter 19. A Public Health Action Plan to Prevent Heart Disease and Stroke includes a recommendation to elevate the prominence of program evaluation by convening a “watershed conference” calling attention to its methods, contributions, and the need for broader visibility and support of this critical public health function.2 Success would be seen in public health agencies attaining greater capacity for program evaluation, state and local public health agencies receiving needed assistance in planning for program evaluation, and a plan for meeting these requirements being in place. The proposed conference took place in late summer 2007, and a report of the proceedings is forthcoming. Policy Analysis Policy analysis in the present context concerns the relation between epidemiologic evidence and policy development—whether current policy is consistent with epidemiologic evidence, whether existing evidence is sufficient to support policy change, and whether identifiable gaps in evidence suggest that
new research is needed to advance the development and implementation of policy for CVD prevention. To take up the approach of policy analysis in this sense suggests considering types of issues that might constitute deterrents to policy development, adoption, and implementation—for example, issues surrounding primordial prevention, with respect to need, assets, costs, priorities, and impact: Is the policy needed? Evidence indicates that investment in primordial prevention is low; prevention can be effective; and risk factors and their consequences begin to develop early in life and progress with age, at differential rates such that some groups (such as African Americans) are at early and continuing disadvantage. Although these observations would support the need for the policy, some decision makers might require further evidence on one or another of these points or raise other questions about need. Potential research question: What proportion of the population is at low risk of CVD, who are they, and what interventions are effective in preserving low risk? What assets are required and available to implement the policy? Increased investment would require adequate infrastructure, access, and utilization to be effective. Research may be needed to determine whether these conditions are present or could be established and with what reach with respect to populations of special concern due to disparities. Potential research question: What conditions, behaviors, or practices are expected to change in consequence of increased investment in prevention, what is their nature and prevalence in the population and groups of special concern, and what specific interventions are known to be feasible for the target population? What are the costs of prevention? Alternative prevention scenarios, each with its evidence base, costs, and expected effectiveness would be desirable inputs to decision making on this policy. Potential research question: How much gain would be expected in what measures of health from a given increase in investment in primordial prevention, and for what segments of the population? How would other priorities be affected by this policy? Within the health arena, increased investment in primordial prevention might be instituted by containing cost increases or reducing some treatment expenditures. Or resources might be shifted from another sector. Potential research question: Would gains from increased support of prevention “upstream” be offset by reduced investment causing higher rates of disability or recurrence of CVD events “downstream”? Are there potential positive effects in other sectors from increased investment in prevention?
46894_CH23_679_696.qxd
2/25/10
4:17 PM
Page 685
PROPOSED RESEARCH AGENDAS 685
What would be the expected impact of this policy on population health indicators, on what time scale? Potential research question: Can consequences of similar policies implemented elsewhere be applied, or can models of population impact under different policies be developed and evaluated? Implementation Research Less familiar than the foregoing approaches, “implementation research,” or “implementation science,” was described by staff of the Fogarty International Center, National Institutes of Health, as a tool in the Center’s efforts to strengthen health care especially in developing countries:17 Realizing the need for a quantitative, scientific framework to guide health-care scale-up in developing countries, researchers in health, engineering, and business are building interest in implementation science. Unlike routine applied (or operations) research, which may identify and address barriers related to performance of specific projects, implementation science creates generalizable knowledge that can be applied across settings and contexts to answer central questions. Why do established programs lose effectiveness over days, weeks, or months? Why do tested programs sometimes exhibit unintended effects when transferred to a new setting? How can multiple interventions be effectively packaged to capture cost efficiencies and to reduce the splintering of health systems into disease-specific programs? Answering questions like these will require analysis of biological, social, and environmental factors that impact implementation, both to develop and test communitywide, multisector interventions that are not testable in clinical settings, and to identify how proven clinical interventions should be modified to achieve sustained health improvements in the “real world.” Although this focus is on clinical settings, the questions posed imply that multiple disciplines are needed to address them and that policy and environmental considerations enter into the contemplated research and interpretation of its findings. Public Health Services and Systems Research “Public health services and systems research” is an emerging area of research that might be seen as the public health counterpart of health services research.18 It focuses on the working of the public health system, whereas traditional health services re-
search mainly concerns medical care delivery and financing. The relevance of public health services and systems research to CVD and other chronic disease prevention lies in the importance of upstream, system effects for achieving population-level impact on risk and disease:18, p 171 The continued development of this area is imperative if we are to have any possibility of success in rebuilding and modernizing our public health infrastructure and improving our nation’s health. It is difficult to change major disease indicators one person at a time. It is much more parsimonious to attack these problems on a population basis, where even modest changes can lead to major changes in the incidence of disease. The knowledge of how we build and design systems to provide population-based services efficiently and effectively is key to that effort. Research that answers questions focused on those efforts is vital to the nation’s health.
PROPOSED RESEARCH AGENDAS Proposals for an organized research agenda in various aspects of CVD prevention have been presented and updated from time to time over the past halfcentury by multiple organizations, agencies, and expert groups. The diversity of research topics reflected in several examples highlighted here illustrates the breadth and depth of CVD epidemiology and prevention as a research endeavor. National Heart, Lung and Blood Institute A comprehensive research agenda was presented more than a decade ago in the National Heart, Lung and Blood Institute (NHLBI) Report of the Task Force on Research in Epidemiology and Prevention of Cardiovascular Diseases.19 Its recommendations remain relevant today, more than a decade after its publication. The report addresses the challenge of cardiovascular disease prevention in these terms:19, pp xiv–xv [T]here are now unequaled opportunities for further reducing CVD within all segments of American society. The strategic key, and the greatest opportunity in preventing CVD, is to prevent the development of CVD risk in the first place. . . . The task force thus emphasizes a population-wide strategy aimed at modifying lifestyles and major CVD risk factors beginning in early childhood and continuing throughout
46894_CH23_679_696.qxd
686
2/25/10
4:17 PM
Page 686
CHAPTER 23 EPIDEMIOLOGY AND A CVD PREVENTION RESEARCH AGENDA
the life-span. Pursuit of this strategy will involve observational and analytical studies, basic epidemiological research, randomized clinical trials, and demonstration projects essential to an effective CVD prevention effort. The great importance of this recommended focus of future research in epidemiology and prevention of cardiovascular diseases is that it makes explicit a fundamental shift from prevention of cardiovascular events to prevention of increased risk. Several implications of this concept have been discussed elsewhere.20 The chief emphasis is on the risk factors themselves as the outcome of concern. Their precursors (e.g., dietary imbalance, physical inactivity, and others) become the focus of intervention. The target population for research and demonstration projects begins with early childhood (or earlier) and extends throughout midadulthood—a stage of life in which a large proportion of the population may still exhibit desirable levels of blood lipids, blood pressure, and other risk factors and can still benefit from interventions to prevent the risk factors “in the first place.” Beyond this fundamental approach, the Task Force report proposed six specific priority areas for research (Table 23-1). The first priority expressed the overall goal of prevention of the risk factors. The second addressed control of already-existing risk factors—especially blood pressure, for which greater control should have been achieved through the existing population-wide measures. The third concerned the socioeconomic differences in cardiovascular disease occurrence throughout the United States. The fourth and fifth priorities related to intervention methods and strategies for population monitoring, beginning in youth and extending to whole populations.
Table 23-1
Priority Areas for Research
The priority areas—for both basic epidemiological research on causation and enhanced application of already available knowledge—are: • Prevention of adverse lifestyles and related risk factors • Control of high blood pressure and other established cardiovascular disease (CVD) risk factors • Reduction of CVD events, disability, and death associated with socioeconomic differences • Prevention of hypertension, dyslipidemia, smoking, and atherosclerosis beginning in youth • Improvement of populationwide prevention strategies • Clarification of the insulin-glucose-atherosclerosis association Source: Reprinted from National Heart, Lung and Blood Institute, National Institutes of Health, Bethesda, Maryland.
The sixth priority was to clarify the complex relationships among insulin and glucose regulation, atherosclerosis, and the other major risk factors. In each priority area, a number of more specific research questions were posed. Though proposed more than a decade ago, the areas of emphasis proposed by the Task Force remain valid today. Additional technical resources and developments are needed to support the research agenda proposed by the NHLBI Task Force.19 These include new biostatistical techniques to address complex issues including design and analysis of community trials; improved follow-up methods for population studies especially of low-income, unemployed, or highly mobile persons; enhanced standardization and certification procedures for epidemiologic data collection; long-term storage and utilization of biological samples from population studies; a funding mechanism for short-term pilot studies to explore new laboratory findings or for methodologic investigations; and improved methods for large-scale dietary assessment of populations, based on large samples and repeated comprehensive nutritional assessments. Apart from these methodologic approaches, five broad qualities or characteristics to be considered as criteria in setting research priorities were addressed in the NHLBI Task Force report (Table 23-2).19 First, research that has potential for application early in the processes of atherosclerosis and hypertension should be emphasized. Second, a population-wide focus should be an important criterion for selection of research questions and programs by the Institute in planning its research agenda. Third, specific population groups should be addressed for whom special disease burdens or obstacles to prevention are present. (This provision would reinforce inclusion of women and minorities in epidemiologic research and underscores the importance of populations experiencing health disparities.) Fourth, studies with potential to bridge gaps between laboratory or patient-oriented research and prevention in the community should receive high priority to accelerate translation of research at every level into public health practice. Fifth, studies with potential to build banks of information and materials should be considered especially valuable, since the need for large data sets is increasingly recognized, as is the utility of longterm follow-up of groups with good documentation of behaviors or risk factors prior to the onset of overt cardiovascular disease. Currently, it is the Division of Cardiovascular Sciences within NHLBI that supports populationbased research (as well as clinical studies) on cardiovascular and other diseases. The scope of the
46894_CH23_679_696.qxd
2/25/10
4:17 PM
Page 687
PROPOSED RESEARCH AGENDAS 687
Table 23-2
Strategies for Research
The nation’s research program in epidemiology and prevention of cardiovascular disease (CVD) has the greatest promise of benefiting the public if it is continued and expanded to emphasize research that • Has potential for early application in CVD prevention (especially prevention of elevated risk factors) and for discovering other traits amenable to intervention early in the disease process • Focuses on the whole population, as well as on individuals at high risk, in seeking causes and means of prevention of high rates of disease • Embraces specific population groups whose social circumstances (eg, low socioeconomic status), behavioral patterns (eg, diet, smoking, and physical activity), or other characteristics (eg, age, sex, ethnicity, genetics) impose special burdens of disease or special challenges to prevention • Incorporates approaches and findings from laboratory and clinical science and from developing technologies (eg, noninvasive imaging of atherosclerotic lesions) which enhance the potential for population-based studies to accelerate the translation of laboratory, clinical, and epidemiological research results into public health applications • Builds banks of information and materials to support future studies (eg, from large-scale studies, participant rosters maintained to provide later follow-up data on lifestyle characteristics and risk factors for defined populations, frozen samples of serum or DNA and urine, registries of family sets for future studies of postulated genetic-environmental interactions) Source: Reprinted from National Heart, Blood and Lung Institute, National Institutes of Health, Bethesda, Maryland.
Division’s charge includes epidemiologic research to identify and describe risk factors and study influences on both these factors and disease outcomes; clinical trials of preventive measures; and studies of applied prevention and treatment strategies. Topics of research include genetic, behavioral, sociocultural, and environmental factors and temporal patterns in disease mortality, incidence, prevalence, and morbidity, based in part in ongoing multicenter population studies.21 (Available at www.nhlbi.nih.gov. Accessed December 26, 2009.) Fogarty International Center The Fogarty International Center is the principal component of NIH that addresses needs in global health research. In its strategic plan for 2008–2012, the Center set forth as its first two goals:22,23, pp 1–2 to mobilize the scientific community to address the growing epidemic of chronic, noncommu-
nicable diseases related to increased longevity and changing lifestyles in the developing world . . . [and] to foster implementation research training in order to help reduce the “know-do” gap, which prevents discoveries from being put into practice, particularly in resource-poor countries. The recently increased emphasis by the Center on chronic disease prevention and management is reflected in these goals, with both research and training components. The widely acknowledged failure to follow discovery by application of new knowledge, the “know-do” gap, is to be narrowed specifically through implementation research. The impact as described would be chiefly in the clinical arena, but populationlevel applications are readily identifiable. World Health Organization The World Health Organization (WHO) has addressed research needs in CVD and other chronic disease prevention for several decades, principally through a longstanding effort to convene expert groups by the former CVD Unit at Headquarters in Geneva. Typically such reports presented a review of current science as background to recommendations for policy and research, and several of them are cited in Parts I–III. Following dissolution of the Unit, with reorganization of activities in the area of noncommunicable diseases, a new sequence of reports began to appear. A Scientific Group report in 1994 on research priorities regarding CVD risk factors was followed closely by the more generic report, Investing in Health Research and Development. This was “a review of health needs and related priorities for research and development in the low-income and middle-income countries . . . intended as a resource to assist decision-making by governments, industry and other investors on the allocation of funds to, and within, health R&D.”24,25, p xxi Emphasis was placed on the need to obtain “reliable basic data on prevalence of and trends in noncommunicable diseases and risk factors”:25, p xxxvii Faced with rapidly growing burdens of noncommunicable diseases, low-income and middleincome countries should significantly increase their relevant strategic research in epidemiology, behavioural science and health policy with the aim of reliably monitoring the true prevalence and trends of these conditions in their populations, and understanding their determinants. Basic data on morbidity, mortality and disability are currently inadequate in many regions, as are data on the country-specific and
46894_CH23_679_696.qxd
688
2/25/10
4:17 PM
Page 688
CHAPTER 23 EPIDEMIOLOGY AND A CVD PREVENTION RESEARCH AGENDA
region-specific levels and determinants of environmental and behavioural risk factors. Lowcost methods for collecting reliable data, such as the use of disease surveillance points, must therefore be developed. In contrast to the need for epidemiological and behavioural research, biomedical science relevant to these conditions is already comparatively well supported in the established market economies. However, genuine differences in the characteristics of environments and populations will occasionally require additional biomedical research in some regions—as, for example, in seeking explanations for the observed high risk in South Asians of diabetes and heart disease. . . . The development and evaluation of algorithms and policy instruments for the cost-effective prevention, diagnosis, treatment and rehabilitation of noncommunicable diseases is an immediate priority for support by governments and other investors. A later interest was to support the concept of “life course” research. This aspect of the WHO research agenda for NCD prevention is addressed in Life Course Perspectives on Coronary Heart Disease, Stroke, and Diabetes. The Evidence and Implications for Policy and Research, a WHO report on a 2001 expert consultation.26 The underlying concept is that, more than simple age-specific differences over the life span in manifestations of risk or disease, differences in the influence of health-related exposures characterize certain periods of life, specifically “critical periods” or “sensitive periods”:26, p 3 The term “critical period” implies exposures that must occur in some specified window(s) of time and often involve exposures that alter normal biological development. “Sensitive period” exposures refer to a broader class of influences that may have greater impact on later outcomes if they occur in certain periods than others. In empirical terms, both critical and sensitive period exposures imply time by exposure interactions. Implications of the life-course concept for research in CVD and NCD prevention are to specify several research questions under each of four headings: research on causes and interactions; trends analysis and surveillance; intervention research; and refinement of methodology. Both research topics and design approaches for each category are outlined in the report. At the conclusion of the consultation, it was judged that it represented “a vital starting point for harnessing the potential of the life course perspective to identify the most appropriate and effective prevention policies in different populations.”26, p 30
Most recently, in accordance with its Global Strategy for the Prevention and Control of Noncommunicable Diseases, WHO developed an action plan, adopted by the Sixty-First World Health Assembly in May, 2008, one of whose six objectives is “to promote research for the prevention and control of noncommunicable diseases”:27, pp 14–15 A coordinated agenda for noncommunicable disease research is an essential element in the effective prevention and control of noncommunicable diseases. In establishing such an agenda, the aim is to enhance international collaboration to promote and support the multidimensional and multisectoral research that is needed in order to generate or strengthen the evidence base for cost-effective prevention and control strategies. Priority areas include the analytical, health system, operational, economic and behavioural research that are required for programme implementation and evaluation. The plan calls for the Secretariat of WHO to take action by developing a prioritized research agenda for noncommunicable diseases that “should generate knowledge and help to translate knowledge into action through innovative approaches in the context of low- and middle-income countries.” It is worth considering which if any of the proposed items on the proposed agenda would not be relevant to highincome countries as well:27, pp 15–16 • the assessment and monitoring of the burden of noncommunicable diseases and its impact on socioeconomic development; • the monitoring of the impact of poverty and other indicators of socioeconomic disparity on the distribution of risk factors; • the assessment of national capacity for the prevention and control of noncommunicable diseases and the evaluation of approaches to fill existing gaps in capacity; • the evaluation of impact of community-based interventions on risk factor levels, and on morbidity and mortality associated with noncommunicable diseases in different populations; • the assessment of the cost-effectiveness of clinical and public-health interventions for improving health behaviours and health outcomes; • the evaluation of different strategies for early detection and screening of noncommunicable diseases in different populations, with an emphasis on cancers, diabetes, and hypertension; • the evaluation of interventions for secondary prevention on cardiovascular disease outcomes in different settings;
46894_CH23_679_696.qxd
2/25/10
4:17 PM
Page 689
PROPOSED RESEARCH AGENDAS 689
• the study of the effectiveness of different organizational patterns in health-care institutions in improving health care for chronic conditions, with a special focus on primary health care; • the analysis of research on factors affecting consumer behaviour and dietary choices, including marketing; • the study of approaches for improving access to, and availability of, essential medicines, essential medical technologies and other central elements of health care; and of approaches for improving the development of affordable new drugs for neglected diseases like Chagas disease, and for rheumatic fever, together with vaccines like that against human papillomavirus; • the assessments of the role, efficacy, and safety of traditional medicines in the management of noncommunicable disease [2008–2009]. To implement this action, WHO convened an international meeting in August 2008 to discuss such a prioritized research agenda. Background papers were prepared to address both disease-specific and crosscutting topics. Closing remarks were presented by Richard Horton, Editor of The Lancet, and subsequently published by ProCor.28 (Available at www .procor.org/advocacy. Accessed October 19, 2008.) Reflecting on the proceedings of that meeting, Horton noted the uniqueness of WHO in its role and responsibility to foster research for prevention and control of noncommunicable diseases and the need for WHO to direct more of its resources to this area. He also posed the question how a social movement for research to improve human health is to be created. He suggested that the answer comes not from biomedical science, but from political science:28, p 3 We need to understand the politics of global health. We have to be opportunistic. Is the time right for us? Are the global conditions favorable for us to make progress? They are favorable, but we can make them even more favorable by doing great science, convening the right people at the right events, and through advocacy––all tied to institutional leadership, nationally and globally. International Heart Health Society As part of the International Heart Health Conference in Victoria, British Columbia, and subsequent Heart Health Conferences, calls for policy development and action by governmental and private organizations and agencies included recommended areas of research. For example, the International Heart Health Advisory
Committee included in its Catalonia Declaration, directed to governments throughout the world, several specific research emphases:29, p 79 Research agencies should allocate appropriate resources to accomplish the following: • Develop new methods and approaches that facilitate the dissemination and uptake of existing prevention knowledge and interventions by organizations concerned with heart health at all levels. • Carry out organizational and evaluative research studies to learn about the value and cost of alternative methods for organizing, financing, and managing heart health programmes, including options for private sector funding of the delivery of such programmes. • Implement systems for surveillance of risk factors and cardiovascular disease, particularly in sentinel populations, including young people and people undergoing rapid social and economic change. • Implement systems for monitoring and reporting progress and results of both planned and unplanned interventions. Public Health Action Plan As noted previously, development of a prevention research agenda for cardiovascular diseases is recommended in A Public Health Action Plan to Prevent Heart Disease and Stroke.2,30 From the perspective of the Action Plan, present knowledge must be put into immediate action to prevent cardiovascular diseases. The greatest opportunity at this stage of knowledge lies in development, adoption, and implementation of policies for cardiovascular health promotion and CVD prevention. Because policy and environmental change, system change, and population-wide behavior change have broadest reach, these populationlevel intervention approaches have high priority in a research agenda supporting public health action. A more favorable balance in research between “upstream” and “downstream” areas of the Action Framework is called for. Yet this need is not well recognized. For this reason, a newly articulated research agenda has been considered a high priority. How might such an agenda be framed? For undertaking this task, it appears necessary to focus initially on one selected policy issue, potentially devising in the process a method that might have general applicability in other policy areas. The desired outcome for the research agenda would be to determine whether one or more “critical investigations” could be devised to resolve a policy issue and
46894_CH23_679_696.qxd
690
2/25/10
4:17 PM
Page 690
CHAPTER 23 EPIDEMIOLOGY AND A CVD PREVENTION RESEARCH AGENDA
enable policy development, adoption, and implementation to proceed. For illustration, a candidate focal issue might be the current low level of investment in prevention of cardiovascular and other chronic diseases in the United States, in relation to expenditures for treating these conditions. Proposals to increase investment in prevention have had little success. What are the principal obstacles to progress in this policy area, and are there questions that could be resolved through research? It is anticipated that the National Forum for Heart Disease and Stroke Prevention, as it matures, will develop a proposed research agenda, perhaps initially in the area of economics of prevention, utilizing the strategy of policy analysis outlined previously.
CAPACITY REQUIREMENTS The resource requirements for realizing the research potential in epidemiology and prevention of cardiovascular diseases include sponsorship and funding, training of investigators, and institutionalization. Underlying these capacities is the need for support of political leadership and government, the private sector, and society as a whole. Sponsorship and Funding In the United States, health-related research funding comes largely from the federal government and principally via the National Institutes of Health. The emphasis of this research is concentrated in the biomedical arena, as indicated by the themes of the NIH Roadmap: New Pathways to Discovery, Research Teams of the Future, and Reengineering the Biomedical Research Enterprise.31 Nonetheless, support for population studies and other work in cardiovascular epidemiology continues under the Division of Cardiovascular Sciences at NHLBI. (Available at www.nih.gov/nhlbi. Accessed December 26, 2009.) CDC’s Advancing the Nation’s Health: A Guide to Public Health Research Needs, 2006–2015 links research priorities to the agency’s overall strategic imperatives, among them public health research: “Research supports the scientific foundation of public health policies, programs, and practices [italics in original].”32, p 3 In the area of promoting health to reduce the burden of chronic diseases and disability, CDC’s research priorities include both age-specific goals and research across the life span to enable implementation of effective health promotion strategies and reduction of risk factors and chronic diseases, with special emphasis on disparities. Cross-cutting research priorities potentially relevant to CVD preven-
tion include social determinants of health; health systems; public health science, policy and practice; economics and public health; and others. There is close alignment between the CDC research agenda and the approach outlined previously. It is implemented in part through the Prevention Research Centers Program, linking CDC with multidisciplinary research groups chiefly in schools of public health throughout the United States.10,33 Over time, CDC could develop a greatly expanded extramural public health research enterprise and support a major prevention research agenda for CVD and other chronic diseases. A leading example of a national voluntary organization sponsoring cardiovascular research is the American Heart Association (AHA), whose Task Force on Strategic Research Direction reported in 2002 on identified priority topics in basic science, clinical science, and “population/outcomes/epidemiology/social science.”34, p 2632 Topics proposed by the latter group were: (1) methods to improve utilization and quality of preventive services and the role of the healthcare system and policies; (2) causes of disparities in risk in subpopulations; (3) understanding the relation of lifestyle and metabolic risk factors; (4) psychosocial risk factors; and (5) population genetics and pharmacogenetics. Topic (1) has the closest affinity with the policy-related research emphasized here, although it focuses on clinical preventive practice rather than the broader health system and population health. The AHA Task Force identified two topics for strategic research funding—obesity and other risk factors, and functional genomics and population genetics. These would be given special priority for funding within the overall research program. Of special interest is the mechanism proposed by the Task Force for implementing this new concept of strategic research priorities, by supporting small clusters or teams of investigators:34, p 2630 (a) Multidisciplinary composition––i.e., including intellectual support for all three domains of basic, clinical, and population science (including, as appropriate, bioinformatics and social sciences). The cluster of investigators would be expected to include competence in all three domains. (b) Focus on developing beginning investigators’ abilities in applying basic, clinical, and population science techniques to CVD and stroke problems. (c) Involvement of more than one institution. This might take the shape of the beginning investigator’s receiving training at one institution and bringing these techniques to another. Or it could involve formation of a
46894_CH23_679_696.qxd
2/25/10
4:17 PM
Page 691
CAPACITY REQUIREMENTS 691
research program cluster across two or more institutions. This outline of AHA’s new strategic research direction touches on integration across research domains and disciplines, training for multilevel capacities in new investigators, and multi-institutional collaboration. Training Personnel requirements in health research, especially in epidemiology and prevention, have been emphasized, for example, by the Pew Health Professions Commission.35 In its assessment Critical Challenges: Revitalizing the Health Professions for the TwentyFirst Century, the Commission concluded that “large numbers of health professionals will require retraining in disease prevention, clinical epidemiology, process and systems analysis and managerial epidemiology.”35, p 55 To provide adequate numbers of epidemiologists in cardiovascular disease prevention will require new and expanded training programs, including those with innovative approaches to interdisciplinary research and program development. Fields and topics to be represented in training programs might well include—in addition to epidemiology, biostatistics, and preventive cardiology—such areas as social and behavioral sciences; public health economics; policy development and implementation; and community organization and community participatory research. The NHLBI Task Force report addressed training requirements as well as research priorities.19 Training opportunities for physicians and other health professionals and researchers should be expanded, as these recommendations suggest. Access to both shortterm courses and academic degree programs is needed by large numbers of candidates to meet the current and growing demand for qualified investigators. Opportunities for acquiring and expanding knowledge and skills through web-based tools should be a prominent part of strategies for meeting long-term training needs. Institutionalization The institutional and organizational bases for the needed research depend, as do individual investigators, on continuity of funding for maintenance of key staff, facilities, and equipment to enable the development and implementation of research proposals. Institutionalization of field settings for research is critical for recruitment, training, advancement, and retention of qualified and experienced researchers. Although sponsorship and funding by central agencies is of course indispensable, stable settings in which
to conduct the funded research are equally so. Longterm research programs of the past, such as the Seven Countries Study and the WHO MONICA Project, have provided settings in which new investigators could enter the field, develop credentials and careers, and become leaders. There is need to organize research with such longterm stability in view. This issue is considered at local or national, regional, and global levels, in the Institute of Medicine report Control of Cardiovascular Diseases in Developing Countries. Research, Development, and Institutional Strengthening.36 Although presented in the context of developing countries, the outline of major research and development (R&D) functions seems universally applicable:36, p 61 The immediate goal for cardiovascular R&D is to enhance local capacity through education and training; development of networks where appropriate; and conduct of local research that is comparable with other centers and applicable internationally. For activities in CVD prevention and control to be successful, it is essential that nonhealth sectors––for example, education, agriculture, industry, and environment––are also included in development of the program. Potential partnerships should be identified and integrated into the action plan. The major functions to be undertaken at different levels are as follows: 1. Local or national: Develop and maintain the capacity and resources to plan, implement, and evaluate research and demonstration projects for CVD prevention. These should include: • community-based assessment of the CVD burden; • monitoring of and intervention in risk factors and their determinants; • testing clinical intervention (e.g., EVP [‘essential vascular package’ of effective, low-cost drugs] and algorithms for low-cost clinical care) and, where feasible, • investigating the local determinants of CVD risk. 2. Regional: Support local or national centers as they undertake their activities providing: • technical support; • communication and exchange of information; • funding support from regional and international donors; and
46894_CH23_679_696.qxd
692
2/25/10
4:17 PM
Page 692
CHAPTER 23 EPIDEMIOLOGY AND A CVD PREVENTION RESEARCH AGENDA
• evaluation of the status of CVD prevention activities in the region. 3. Global: Support regional centers in the conduct of their activities: • maintenance of global data on CVD prevention and control programs; • dissemination of current data, protocols, guidelines, and literature on CVD prevention and control; and • convening of advisory groups to assess global needs, evaluate current activities, and recommend additional activities if they are needed. The report concluded that global leadership for R&D in CVD prevention and control should be assumed by WHO. Given assimilation of CVD into the noncommunicable disease agenda by WHO, and the current initiative to develop a comprehensive research agenda in this area, it will be of interest to see the outcome of this process and its applicability to CVD epidemiology and prevention throughout the world.
POPULOMICS: THE POPULATION CONTEXT OF RESEARCH ON HEALTH The current global context of epidemiology of CVD and other chronic diseases suggests that the concept of epidemiology represented in Figure 17-4 and described previously is no longer sufficient. For example, the regional and global scale of population health addressed by the Disease Control Priorities and Global Burden of Disease projects, ultimately concerning the whole of the world population, is far greater than implied by “societies” in the figure.37,38 The object of epidemiologic research has become all people in all places. The dimension of time has expanded with increasing attention to both historicalexplanatory and predictive models of long-term changes in population health. The theory of epidemiologic transition, for example, first published in 1971, gained more recent prominence with growth of interest in the increasing burden of CVD and other chronic diseases in the developing countries.39 Models of past and future trends in CHD mortality are further examples of contemporary epidemiologic research extending backward and forward over decades or longer. These several developments have greatly extended the classic dimensions of person, place, and time in cardiovascular epidemiology. The increasing need for interaction between epidemiology and other health science disciplines, as
well as engagement with sectors beyond biomedical and community health, is also increasingly apparent. This is especially so if epidemiology is to become more strongly oriented toward policy development and public health action. For example, interaction between CVD epidemiology and other social and behavioral sciences gains new importance if social determinants of health are to be addressed more prominently in the research agenda. Similarly, links between CVD and other chronic diseases suggest connections between the respective specialties concerned with these diseases and their causation and prevention. For epidemiology to contribute effectively to policy development, appreciation of other influences on health policy and decision making, including economics, politics, culture, special interests, and others, is required. The research agenda, proposed especially to identify and resolve barriers to policy development and implementation, should also recognize the importance of understanding these influences. The implied connectivity between epidemiology and the array of disciplines concerned with specific levels of biological organization, particular disease processes, and myriad external factors adds a new dimension to our concept of contemporary epidemiology. To put CVD epidemiology in its proper context today requires a holistic view of populations and health. There is an analogy in this circumstance between epidemiology and genetics, as each seeks a more integrated conceptualization of its larger purpose. According to Khoury and others, quoting Guttmacher and Collins, genetics is “the study of single genes and their effects,” whereas genomics is “the study of not just of [sic] single genes but of the functions of and interactions of all genes in the genome.”40, pp 7–8 Much as “genomics” serves the field of genetics as an all-encompassing concept, the term “populomics” would serve the study of population health. In what has been termed “the -omics era,” this is an appealing concept and may have considerable heuristic value.41 Populomics, then, represents a scientific domain that concerns population health and is global in scale, dynamic over long reaches of time, and inclusive of fundamental population processes and their interactions—anthropological, demographic, sociocultural, economic, environmental, and others. Populomics can be understood as an “integrative human ecology,” having population health as its focus and embracing all relevant disciplines, with epidemiology at its core. It is epidemiology transcending the “boxes.” This concept of epidemiology is represented in Figure 23-1.
46894_CH23_679_696.qxd
2/25/10
4:17 PM
Page 693
REFERENCES 693
Contemporary Epidemiology as the Core Discipline of Populomics: Beyond the Box PERSON GLOBAL SOCIETIES COMMUNITES FAMILIES POPULATION
INDIVIDUALS
PLACE ORGANS
CELLS
MOLECULAR AND SUBMOLECULAR PARTICLES
Traditional Epidemiology Clinical Research Pathology, Physiology
TIME
Cell Biology Molecular Biology Contemporary Epidemiology
RELATED DISCIPLINES AND SECTORS
Figure 23-1 Contemporary Epidemiology as the Core Discipline of Populomics: Beyond the Box. Source: Adapted with permission from the Annual Review of Public Health, Vol 1, © 1980, by Annual Reviews, Inc.
CURRENT ISSUES 1. Will the call for a greatly expanded research enterprise to support CVD and other chronic disease prevention be heard by those in a position to provide sponsorship and funding? 2. Can a network of broad-based training programs for research and practice in CVD and other chronic disease prevention become established to strengthen local, national, regional, and global efforts and develop new leaders for today and the coming decades? 3. How shall institutionalization be accomplished to assure growth and sustainability of populomics and its core discipline of epidemiology, as it is applied to the challenge of preventing CVD and other chronic diseases at the population level and on a global scale? 4. What is the compelling research agenda, with what promise of success in leading to societal change for prevention of CVD and other chronic diseases, that will support the case and build on the success of 1, 2, and 3 above?
REFERENCES 1. Remington RD. Role of organized public health in cardiovascular disease prevention. In: Marmot M, Elliott P, eds. Coronary Heart Disease Epidemiology: From Aetiology to Public Health. Oxford (England): Oxford Medical Publications; 1992: 515–524. 2. US Department of Health and Human Services. A Public Health Action Plan to Prevent Heart Disease and Stroke. Atlanta, GA: US Department of Health and Human Services, Centers for Disease Control and Prevention; 2003. 3. Thacker SB. Historical development. In: Teutsch SM, Churchill RE, eds. Principles and Practice of Public Health Surveillance. 2nd ed. Oxford (UK): Oxford University Press; 2000. 4. Birkhead GS, Maylahn CM. State and local public health surveillance. In: Teutsch SM, Churchill RE, eds. Principles and Practice of Public Health Surveillance. 2nd ed. Oxford (UK): Oxford University Press; 2000.
46894_CH23_679_696.qxd
694
2/25/10
4:17 PM
Page 694
CHAPTER 23 EPIDEMIOLOGY AND A CVD PREVENTION RESEARCH AGENDA
5. Teutsch SM, Churchill RE, eds. Principles and Practice of Public Health Surveillance. 2nd ed. Oxford (UK): Oxford University Press; 2000. 6. Centers for Disease Control and Prevention. Framework for program evaluation in public health. MMWR. 1999;48(No. RR-11):1–40. 7. Committee for the Study of the Future of Public Health. The Future of Public Health. Washington, DC: Division of Health Care Services, Institute of Medicine. National Academy Press; 1988. 8. Stallones RA. To advance epidemiology. Annu Rev Public Health. 1980;1:69–82. 9. Rose G. Sick individuals and sick populations. Int J Epidemiol. 1985;14:32–38. 10. Stoto MA, Green LW, Bailey LA, eds. Linking Research and Public Health Practice: A Review of CDC’s Program of Centers for Research and Demonstration of Health Promotion and Disease Prevention. Board on Health Promotion and Disease Prevention. Institute of Medicine. Washington, DC: National Academy Press; 1997. 11. Susser M, Susser E. Choosing a future for epidemiology: I. Eras and paradigms. Am J Public Health. 1996;86:668–673. 12. Susser M, Susser E. Choosing a future for epidemiology: II. From black box to Chinese boxes and eco-epidemiology. Am J Public Health. 1996;86:674–677. 13. Goff DC Jr, Brass L, Braun LT, et al. Essential features of a surveillance system to support the prevention and management of heart disease and stroke. A Scientific Statement from the American Heart Association Councils on Epidemiology and Prevention, Stroke, and Cardiovascular Nursing and the Interdisciplinary Working Groups on Quality of Care and Outcomes Research and Atherosclerotic Peripheral Vascular Disease. Circulation. 2207;115:127–155. 14. Giampaoli S, Capewell S, Shelley E, et al. Foreword. Eur J Cardiovasc Prev and Rehab. 2007;14(suppl 3):S1.
15. Bonita R, de Courten M, Dwyer T, Jamrozik K, Winkelmann R. Surveillance of Risk Factors for Noncommunicable Diseases: The WHO STEPwise approach. Summary. Geneva: World Health Organization; 2001. 16. Fawcett SB, Sterling TD, Paine-Andrews A, et al. Evaluating Community Efforts to Prevent Cardiovascular Diseases. Atlanta, GA: Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion; 1995. 17. Maden T, Hofman KJ, Kupfer L, Glass RI. Implementation science. Science. 2007;318: 1728–1729. 18. Scutchfield FD, Marks JS, Perez DJ, Mays GP. Public health services and systems research. Am J Prev Med. 2007;33:169–171. 19. National Heart, Lung and Blood Institute. Report of the Task Force on Research in Epidemiology and Prevention of Cardiovascular Diseases. Washington, DC: National Institutes of Health, Public Health Service, US Department of Health and Human Services; 1994. 20. Labarthe DR. Prevention of cardiovascular risk factors in the first place. J Epidemiol. 1997; 6(suppl):1–5. 21. National Heart, Lung and Blood Institute. About NHLBI. Division of Prevention and Population Sciences. Available at www.nhlbi .nih.gov/about/dpps/index.htm. Accessed October 19, 2008. 22. Fogarty International Center. Pathways to Global Health Research. Strategic Plan 2008–2012. NIH Publication No. 08-6261. Bethesda: US Department of Health and Human Services, National Institutes of Health; May 2008. 23. Fogarty International Center. Pathways to Global Health Research. Strategic Plan 2008–2012. Available at www.fic.nih.gov/ about/plan/strategicplan_08-12.htm. Accessed July 15, 2008. 24. Report of a WHO Scientific Group. Cardiovascular Disease Risk Factors: New Areas for
46894_CH23_679_696.qxd
2/25/10
4:17 PM
Page 695
REFERENCES 695
Research. Geneva, Switzerland: World Health Organization; 1994. WHO Technical Report Series 841. 25. Ad Hoc Committee on Health Research Relating to Future Intervention Options. Report of the Ad Hoc Committee on Health Research Relating to Future Intervention Options: Investing in Health Research and Development. Geneva, Switzerland: World Health Organization; 1996. 26. Aboderin I, Kalache A, Ben-Shlomo Y, et al. Life Course Perspectives on Coronary Heart Disease, Stroke and Diabetes: Key Issues and Implications for Policy and Research. Geneva: World Health Organization; 2002. 27. World Health Organization. Prevention and control of noncommunicable diseases: implementation of the global strategy. Report by the Secretariat. A61/8 Sixty-First World Health Assembly. Geneva:World Health Organization; 2008. 28. Horton R. Commentary: An NCD research agenda for WHO. Available at www.procor .org/advocacy/advocacy_show.htm. Accessed October 19, 2008. 29. Advisory Board of the Second International Heart Health Conference. The Catalonia Declaration: Investing in Heart Health. Barcelona, Spain: Department of Health and Social Security, Autonomous Government of Catalonia; 1996. 30. US Department of Health and Human Services. Update to a Public Health Action Plan to Prevent Heart Disease and Stroke. Atlanta, GA: Centers for Diseases Control and Prevention; 2008. 31. Zerhouni E. The NIH roadmap. Science. 2003; 302:63–64, 72. 32. Centers for Disease Control and Prevention. Advancing the Nation’s Health: A Guide to Public Health Research Needs, 2006–2015. Atlanta, GA: US Department of Health and Human Services, Centers for Disease Control and Prevention; December 2006.
33. Centers for Disease Control and Prevention. Prevention Research Centers (PRC). Available at www.cdc.gov/prc/. Accessed October 21, 2008. 34. Roberts R, Bonow RO, Loscalzo J, Mosca L. Report of the American Heart Association Task Force on Strategic Research Direction: Executive summary. Circulation. 2002;106: 2630–2632. 35. Critical Challenges: Revitalizing the Health Professions for the Twenty-First Century: The Third Report of the Pew Health Professions Commission. San Francisco, CA: Center for the Health Professions, University of California; 1995. 36. Howson CP, Reddy KS, Ryan TJ, Bale JR, eds. Control of Cardiovascular Diseases in Developing Countries. Research, Development, and Institutional Strengthening. Washington, DC: Institute of Medicine, National Academy Press; 1998. 37. Jamison DT, Breman JG, Measham AR, et al., eds. Disease Control Priorities in Developing Countries. 2nd ed. Washington, DC: International Bank for Reconstruction and Development/The World Bank; 2006. 38. Lopez AD, Mathers CD, Ezzati M, Jamison DT, Murray CJL, eds. Global Burden of Disease and Risk Factors. Washington, DC: The International Bank for Reconstruction and Development/The World Bank; 2006. 39. Omran AR. The epidemiological transition: a theory of the epidemiology of population change. Milbank Q. 1971;49:509–538. 40. Khoury MJ, Little J, Burke W. Human genome epidemiology: scope and strategies. In: Khoury MJ, Little J, Burke W, eds. Human Genome Epidemiology. A Scientific Foundation for Using Genetic Information to Improve Health and Prevent Disease. Oxford (UK): Oxford University Press; 2004. 41. National Heart, Lung and Blood Institute. Working Group Summary: The Next Step. Population Studies in the “-OMIC” Age. Available at www.nhlbi.nih.gov/meetings/workshops/ population.htm. Accessed June 26, 2007.
46894_CH23_679_696.qxd
2/25/10
4:17 PM
Page 696
46394_INDX_697_710.qxd
2/26/10
2:19 PM
Page 697
Index
AAA (Abdominal aortic aneurysm), 120–123 AACVPR (American Association of Cardiovascular and Pulmonary Rehabilitation), 599–600 Abdominal aortic aneurysm (AAA), 120–123 Abell-Kendall method, 273 ABI. See Ankle-brachial index (ABI) ACC. See American College of Cardiology (ACC) ACCORD (Action to Control Cardiovascular Risk in Diabetes) Trial, 384 ACPM (American College of Preventive Medicine), 553 ACS (Acute coronary syndrome), classification of, 64 ACS (American Cancer Society), 409, 523 Action plans, 657–675. See also A Public Health Action Plan to Prevent Heart Disease and Stroke (CDC) in Americas, 658–660 community/population-wide measures, 670–673 in Europe, 660–661 Heart Health Networks, 662–665 high-risk intervention, 673 individual intervention, 673 obstacles to, 669–673 population diversity and, 671–672 resource limitations and, 672–673 in South Asia, 661–662 strategic imperatives, 673–675 Action to Control Cardiovascular Risk in Diabetes (ACCORD) Trial, 384 Active antiretroviral therapy (HAART), 481 Active living concept, 194 Active Living Policy and Environmental Studies, 194, 527 Acute coronary syndrome (ACS), classification of, 64 Acute ischemic heart disease, classification of, 64 Acute myocardial infarction case-fatality for, 72 classification of, 64
diabetes and, 377–378, 381 diet and, 180 risk factors of, 76–77 smokefree laws and, 422 Adolescents. see Children and adolescents Adulthood. See also Age atherosclerosis and, 49–54 health patterns of, 25 obesity in, 236, 239–241, 257–259 Advancing the Nation’s Health (CDC), 690 Aerobics Center Longitudinal Study, 207 Affluent society diet, 160–161, 164, 165 African Americans. see Race/ethnicity African Heart Network, 664 Age. see also specific age groups adjustment and standardization of, 23 ascertainment of, 23 atherosclerosis and, 48, 49–54 blood lipids and, 274 classification of, 22–23 congestive heart failure and, 126–127 coronary heart disease and, 81–82 fetal and neonatal period, 24 high blood pressure and, 317–318 and life stages, 22–26 modifiability of, 22 mortality patterns and, 19–24 population, age 65 and over, 26 and prevalence of risk factors, 12–13 stroke and, 105–107 Agency for Healthcare Research and Quality (AHRQ), 73 AHA. See American Heart Association (AHA) AIDS, 481 Air pollution, as risk factor, 522–525 Alcohol consumption, 431, 433–447 aortic aneurysms and, 122–123 blood pressure and, 343, 435 community/population-wide measures, 445–446 exdrinkers and risk, 434 genetic factors and, 435 HDL-cholesterol and, 152, 434–435 hemostasis and, 435–436 measurement methods, 433–434 myocardial infarction and, 151–152
697
population differences and, 441–444 population distribution and, 437–441 public health issues, 444–445, 447 race/ethnicity and, 437–440 sex and, 437–440, 443 Alcohol Policy Index, 445 ALLHAT (Antihypertensive and LipidLowering Heart Attack Trial), 348 Alliance for a Healthier Generation, 183 AMA (American Medical Association) Assessment and Management of Adult Obesity, 257 Amenable mortality, use of term, 640 American Academy of Pediatrics, 196, 273 American Association of Cardiovascular and Pulmonary Rehabilitation (AACVPR), 599–600 American Cancer Society (ACS), 409, 523 American College of Cardiology (ACC) on congestive heart failure, 124–125, 127 on evidence-based decision making, 575–577 guidelines and policies, 596–597 on influenza vaccination, 481 on peripheral arterial disease, 118 American College of Preventive Medicine (ACPM), 553 American College of Sports Medicine, 192 American Heart Association (AHA) annual statistical updates, 66 Cardiovascular Diseases and Stroke in African-Americans and Other Racial Minorities in the United States, 29 Cardiovascular Health Promotion in the Schools, 609 childhood recommendations, 181, 182, 604–605, 609 Circulation, 12 Committee on Vascular Lesions report, 42 on community-based implementation of recommendations, 183 on congestive heart failure, 124–125, 127 Council on Arteriosclerosis, 42 Council on Epidemiology and Prevention, 63, 92 Criteria for Evaluation of Novel Markers of Cardiovascular Risk, 490–491 on diabetes and stroke, 381
46394_INDX_697_710.qxd
698
2/26/10
2:19 PM
Page 698
INDEX
(American Heart Association (AHA) cont’d) Diet and Lifestyle Recommendations Revision 2006, 180–181 on dietary recommendations, 180, 182 on drug therapy for high risk lipid levels, 296 on evidence-based decision making, 575–577 on food choice determinants, 169 Guide for Improving Cardiovascular Health at the Community Level, 607, 658 guidelines and policies, 596–597, 599–600, 603–607 on HIV/AIDS, 481 on inflammation, 488 on influenza vaccination, 481 on metabolic syndrome, 384 Methodology Manual for ACC/AHA Guideline Writing Committees, 575–577 on obesity, 253 on optimal dietary guidelines, 294 on peripheral arterial disease, 118 Population-Based Prevention of Obesity, 259 on prevention and control, 556–557 on prevention of heart failure, 131 Primary Prevention of Heart Failure, 131 Primary Prevention of Ischemic Stroke, 103 recommendations for women, 603 Relevance of Genetics and Genomics for Prevention and Treatment of Cardiovascular Disease, 143 Report of the Special Writing Group on Cardiovascular Disease in Women, 27 research agendas and, 690–691 scope of, 647–648, 658 on smoking, 414 Taking the Initiative, 607–609, 658 website, 12 American Journal of Clinical Nutrition, 167 American Journal of Preventive Medicine on active living, 194 on physical inactivity, 215 American Medical Association (AMA) Assessment and Management of Adult Obesity, 257 American Stroke Association (ASA), 596–597 Aneroid manometers, 316 Angina pectoris causal pathways, 544 classification of, 64 incidence of CHD, 70, 72–73 unstable angina, 61, 64, 452 Ankle-brachial index (ABI) age-adjusted prevalence of CVD and, 120 atherosclerosis and, 44 PAD assessment and, 114–118 Anthropometry methods, 229, 259 Antihypertensive and Lipid-Lowering Heart Attack Trial (ALLHAT), 348 Antioxidants, 481–485 Antiplatelet Trialists’ Collaboration, 474, 475 Antithrombotic Trialists’ (ATT) Collaboration, 474–475 Aortic aneurysms abdominal, 120–123
background of, 121 cardiovascular diseases and, 111–113, 120–123 classification of, 112 features of individual cases, 120–121 mortality patterns and, 121–122 risk factors of, 122–123 sex and, 121 smoking and, 122–123 surgical intervention and, 123 Appalachia, health patterns in, 30, 32 ARIC. See Atherosclerosis Risk in Communities Study (ARIC) Arrhythmias atrial fibrillation, 135–136 cardiovascular diseases and, 135–136 classification of, 111–113 ventricular, 135–136 Arteriosclerosis, mortality patterns and, 173 Arthogenisis, mechanisms of, 46–47 ASA (American Stroke Association), 596–597 Ashes to Ashes (Kluger), 396 Aspirin treatment, 474–475 Assessing Physical Fitness and Physical Activity in Population-Based Surveys (CDC, 1989), 194 Assessment and Management of Adult Obesity (AMA), 257 Association of State and Territorial Directors of Health Promotion and Public Health Education (ASTDHPPHE), 554–555 Association of State and Territorial Health Officials (ASTHO), 666 ATBC Study, 178, 485 Atherosclerosis, 41–55 blood lipids and, 275–277 causes of, 543–547 development of, schematic view, 44 genetic factors, 152–153 manifestations of, 43–45 measurement methods, 42–43 mechanisms of atherogenesis, 46–47 person, place, and time, 47–49 prevention and control, 54–55, 211 race/ethnicity and, 44–45, 47, 49, 51 sex and, 44, 47–49, 51–53 in specific life stages, 49–54 treatment of, 54 Atherosclerosis Risk in Communities Study (ARIC) on alcohol consumption, 436, 442–443 carotid ultrasonography in, 43 on hemostatic factors, 466, 472–473 on incidence of stroke, 96, 98 on neighborhood of residence, 525–526 scope of, 12, 26, 70 Atherosclerotic plaque, 41–42 Atkins Diet, 165, 179 Atlas of Heart Disease and Stroke Among American Indians and Alaska Natives (CDC), 94 Atlas of Stroke Hospitalizations Among Medicare Beneficiaries, 98 ATP III reports. see National Cholesterol Education Program (NCEP) Atrial fibrillation, 103–104, 111–112, 135–136, 597 ATT (Antithrombotic Trialists’) Collaboration, 474–475
Bacterial infections, 46–47, 480 Barker hypothesis, 518 Behavior. see Psychosocial factors; Type A behavior pattern Behavioral cardiology, 448 Behavioral Risk Factor Surveillance System (BRFSS) on blood pressure, 345 on diabetes/metabolic syndrome, 371–372 on fruit/vegetable consumption surveys, 170–171 on mortality rates, 66 scope of, 12 on stroke mortality patterns, 94, 98–99 Behavior change, use of term, 555, 560 Bell System Operating Companies study, 505 Best Practices for Comprehensive Tobacco Control Programs (CDC), 421 Binge drinking, 433, 437. see also Alcohol consumption Biomedical and community health research, 568–569 Birth weight, social factors and, 515–518 Black box epidemiology, 679, 682 Blood lipids. see also Genes and environment adverse profiles of, 269–304 atherosclerosis and, 275–277 biological mechanisms of, 275–276 community/population-wide measures, 299–301 concepts and definitions, 270–273 determinants of, 274–275 dietary imbalance and, 275 dietary trials, 294 drug trials, 294–296 early trials, 292–294 family history and, 274–275 genetic factors and, 153, 271–272, 274 HDL-cholesterol, 152, 270, 434–435 hypercholesterolemia, 153, 269, 274 LDL-cholesterol, 269, 273, 485 measurement methods, 273–274 population distribution and, 277–283 prevention and control, 292–299 public health issues, 290, 299–302 race/ethnicity and, 274, 277, 280 risk and individual differences, 285–290 risk and population differences, 283–285 secondary hyperlipidemia, 275–276 sex and, 274, 282, 286–290, 296 Blood pressure. see also High blood pressure (HBP) age and, 345 alcohol consumption and, 435 children/adolescents and, 318, 326–327, 329, 342 genes and environment, 153–154 genetic factors and, 153 migration and, 322 salt intake and, 319–321, 343 sex and, 597–599 studies on, 318 Blood Pressure Studies in Children (WHO), 342 Blood Pressure Study of the Society of Actuaries and Association of Life Insurance Medical Directors of America, 336–337
46394_INDX_697_710.qxd
2/26/10
2:19 PM
Page 699
INDEX
Bloomberg Global Initiative to Reduce Tobacco Use, 423–424 BMA (British Medical Association), 232 Body composition, 166, 229, 231. see also Obesity Body mass index (BMI). see also Obesity associated disease risks, 227 birth weight and, 517–518 cardiovascular risk factors, 237 diabetes and, 368 formula for, 223 physical activity and, 208 as standard measure for obesity, 227–228 weight control and, 259 Bogalusa Heart Study on blood pressure, 318 on childhood atherosclerosis, 49–51, 54 on childhood obesity, 229, 234 diabetes/metabolic syndrome and, 369 Brain attacks. see Stroke (CVA) Brain Attack Surveillance in Corpus Christi (TX) Project, 93 BRFSS. See Behavioral Risk Factor Surveillance System (BRFSS) British Medical Association (BMA), 232 British Medical Journal on atherosclerosis, 535–536 on combination pharmacotherapy, 476 on sodium and HBP, 320 British Regional Heart Study on alcohol consumption, 434 on diabetes, 368 on metabolic syndrome, 378 on physical inactivity, 208 CAC (Coronary artery calcium) score, 43 CAD, use of term, 4 Calls to action. see Action plans Canadian Heart Health Initiative, 659 Candidate genes, 152, 542 CARDIA (Coronary Artery Risk Development in Young Adults Study), 12, 26, 54, 454–455 Cardiac rehabilitation programs, 213 Cardiomyopathies, 8, 124, 125, 381 Cardiovascular Disease Prevention and Control (WHO), 664–665 Cardiovascular diseases, major aortic aneurysms and, 111–113, 120–123 arrhythmias and, 135–136 atherosclerosis, 41–55 congestive heart disease, 58–83 congestive heart failure, 111–113, 123–132 peripheral arterial disease, 111–120 pulmonary embolism, 111–113, 132–135 related conditions, 111–136 stroke, 89–107 venous thromboembolism, 111–113, 132–135 Cardiovascular Diseases and Stroke in African-Americans and Other Racial Minorities in the United States (AHA), 29 Cardiovascular Health and Disease in Women report, 27 Cardiovascular Health Promotion in the Schools (AHA), 609 Cardiovascular Health Study (CHS), 12, 26, 44, 115
Cardiovascular Risk in Young Finns Study, 456 Cardiovascular Survey Methods (WHO), 63–64 Carotid artery, ultrasonography and, 43 Carotid IMT (CIMT), 43–44, 50 Case-control-family design, 145 Case-fatality for coronary heart disease, 71–72, 82–83, 105–107 definition, 65 hospitalization and, 127–130 for stroke, 91, 97–98 CATCH. See Coordinated Approach to Child Health (CATCH) Program Causal pie model, 540–541 Causal Thinking in the Health Sciences (Susser), 538 Causes, of cardiovascular diseases, 535–548 for atheroslcerotic diseases, 543–547 causal constructs, 540–543 causal judgment, 537–540 causal pathways, 544–546 causal thinking, 538–539 causes vs. mechanisms, 542–543 etiology vs. pathogenesis, 542–543 for hypertensive diseases, 543–547 from molecules to populations, 541–542 mutifactorial causation, 537–538 Present Reality continuum, 546–547 risk factors and, 543–544, 546 from single agent to n-dimensional complex, 540–541 CDP (Coronary Drug Project), 292 Centers for Disease Control and Prevention (CDC). See also National Center for Health Statistics (CDC); A Public Health Action Plan to Prevent Heart Disease and Stroke (CDC) Action Plan, 689–690 Advancing the Nation’s Health, 690 Atlas of Heart Disease and Stroke Among American Indians and Alaska Natives, 94 Best Practices for Comprehensive Tobacco Control Programs, 421 on combination pharmacotherapy, 476 HDSP Policy Project website, 612 Heart Disease and Stroke Prevention Division, 612, 647 on inflammation, 488 Lipid Standardization Program, 273 on mortality rates, 66 National Diabetes Fact Sheet website, 371 National Office of Genomics and Disease Prevention, 155 Preventing Death and Disability from Cardiovascular Diseases, 666 Racial and Ethnic Disparities in Heart Disease Among Women, 30–31 on resource limitations, 672 scope of, 647 surveillance systems, 12 Cerebral arteries, 89–90. see also Stroke Chagas disease, 124, 689 Chart Book (NHLBI), 70, 96 CHD. See Coronary heart disease (CHD) CHD, use of term, 4 CHF. See Congestive heart failure (CHF)
699
Chicago Heart Association Detection Project in Industry Study, 77, 246, 248 Chicago Peoples Gas Study, 375 Chicago Western Electric Company Study, 176, 435 Child and Adolescent Trial for Cardiovascular Health. see Coordinated Approach to Child Health (CATCH) Program Children and adolescents. See also Age AHA guidelines for, 181, 182, 604–605, 609 alcohol consumption and, 437 atherosclerosis and, 49–54 birth weight and, 515–518 blood lipids and, 261–262, 272–275, 282, 284, 296, 299 blood pressure and, 318, 326–327, 329, 342 developmental body fat changes, 225 diabetes/metabolic syndrome and, 361, 363, 369, 371–372, 374, 383 health patterns of, 24–25 infant death rates, 515 obesity in, 227–228, 234, 236, 237–238, 242, 256–257 physical activity recommendations, 183 physical inactivity and, 201–202, 208, 214–215 prevalence of risk factors, 12–13 prevention and control, 634 school meal choices, 169, 183 smoking and, 400–405, 417, 420–421 social factors and, 512–513, 515 study on marketing of food products to, 261 tracking of risk factors in, 228 Type A behavior pattern, 455 vulnerable developmental periods, 226–227 Chinese box epidemiology, 679, 682 Chlamydia pneumoniae, 46–47, 477–478 Cholesterol, dietary, 172–176, 237. see also Diet, effects of Chronic bronchitis, 131, 416 Chronic diseases, 620–621 Chronic heart failure. see Congestive heart failure (CHF) Chronic ischemic heart disease, classification of, 64 Chronic obstructive pulmonary disease (COPD), 131 Cigarette smoking. See Smoking, and other tobacco use Cigarette Wars (Tate), 396 CIMT (Carotid IMT), 43–44, 50 CINDI PROGRAM (Country-wide Integrated Noncommunicable Diseases Intervention/WHO), 660–661 Circulation (AHA), 12 Circulatory shock, 312 Coagulation factors, 465 Coagulation pathways, 466–470 Cochrane Collaboration, 574–575 Cochrane Database of Systematic Reviews, 575 Cochrane Handbook for Systematic Reviews of Interventions, 574–575 Coffee and tea, effects of, 178–179 Collaborative Lipoprotein Phenotyping Study, 433, 435
46394_INDX_697_710.qxd
700
2/26/10
2:19 PM
Page 700
INDEX
Combination pharmacotherapy, 476, 624–625 Commission on Social Determinants of Health (CSDH), 504 COMMIT (Community Intervention Trial for Smoking Cessation), 420 Committee on Diet and Health, 181–182 Committee on Vascular Lesions report (AHA), 42 Common SNP, 145 Community Pathology Study (CPS), 52 Community Prevention and Control of Cardiovascular Diseases (WHO), 612 Community prevention strategies. see Population-wide prevention strategies; specific communities Community Preventive Services Task Force on diabetes, 384–385 Guide to Community Preventive Services, 570–571, 607 Comparative Risk Assessment Collaborative Group, 642, 645–646 Computed tomography, ultra-fast, 43–44, 54 Congestive heart failure (CHF) background of, 124 cardiovascular diseases and, 111–113, 123–132 classification of, 112 features of individual cases, 124 hospitalization and, 127–128 incidence of, 128–129 mortality patterns and, 125–128, 131 population studies, 124–125 prevalence, 129 prevention and control, 131 race/ethnicity and, 126, 129 right ventricular failure, 131 risk factors of, 129–130 sex and, 126, 129–130 trends and explanations, 130 Contemporary epidemiology scale, 542 Continuing Survey of Food Intakes by Individuals (CSFII), 170 Control of Cardiovascular Diseases in Developing Countries (IOM), 691–692 Coordinated Approach to Child Health (CATCH) Program, 183 on blood lipid control, 300 on physical inactivity, 214, 215–216 on school meals, 634 COPD (Chronic obstructive pulmonary disease), 131 CORIS (Coronary Risk Factor Study), 632, 634 Coronary arteries, 59–60 Coronary artery calcium (CAC) score, 43 Coronary Artery Risk Development in Young Adults Study (CARDIA), 12, 26, 54, 454–455 Coronary Drug Project (CDP), 292 Coronary heart disease (CHD), 59–83 acute myocardial infarction and, 72 air pollution and, 522–524 background of, 62–63 case-fatality and, 71–72, 82–83, 105–107 coronary arteries, 59–60 depression and, 459 diabetes/metabolic syndrome and, 375–378
diagnostic elements, 64 disability and, 73–74 disparities in, 74 family history and, 146 features of individual cases, 60–62 genetic complexity of, 143–144 hemostatic factors and, 473 incidence of, 67–72 mortality patterns, 66–67 occurrence rates, 65–74 population studies, 63–65 prevalence, 65–66, 72–73 race/ethnicity and, 66, 70–73, 77–78 risk factors, 74–79 sex and, 59, 67–68, 70–73, 77–78, 477 smoking and, 410, 413 sudden death and, 61, 67, 69 trends and explanations, 79–83 triggers, 79, 208 Coronary Risk Factor Study (CORIS), 632, 634 Correlations with vascular pathology, 24 Cost-Effectiveness in Health and Medicine, 583 Cotinine. See Smoking, and other tobacco use Council on Arteriosclerosis (AHA), 42 Council on Epidemiology and Prevention (AHA), 63, 92 Country-wide Integrated Noncommunicable Diseases Intervention (CINDI) Program (WHO), 660–661 CPS (Community Pathology Study), 52 C-reactive protein (CRP), 369, 474, 488–489 Criteria for Evaluation of Novel Markers of Cardiovascular Risk (AHA), 490–491 Critical Challenges (Pew Commission), 691 CSDH (Commission on Social Determinants of Health), 504 CSFII (Continuing Survey of Food Intakes by Individuals), 170 CT angiography, 43 CT scans, 43–44, 54 Cultural heritability, 142 CUORE Study, 606 Current issues alcohol consumption, 446–447 arrhythmias, 135–136 chronic heart failure, 123, 131–132 deep vein thrombosis, 134–135 hemostatic factors, 475 peripheral arterial disease, 118, 120 psychosocial factors, 464 pulmonary embolism, 134–135 CVA. See Stroke (CVA) CVD, use of term, 4 DALYs. See Disability-adjusted life years (DALYs) Da Qing IGT and Diabetes Study, 382 DASH Trial (Dietary Approaches to Stop Hypertension) blood pressure and, 321, 342, 343 coronary heart disease risk and, 179–180 optimal dietary pattern and, 163–164 DCCT (Diabetes Control and Complications Trial), 384
Deaths rates. see Mortality patterns Decision making. see Evidence-based decision making Decline Conference reports (NHLBI, 1978), 67, 79–80, 642 DECODE (Diabetes Epidemiology) Study Group, 375, 377–378 Deep vein thrombosis (DVT). see Venous thromboembolism (VTE) Depression, 456–460 Determinants, of cardiovascular diseases adverse blood lipid profile, 269–304 diabetes and metabolic syndrome, 361–388 dietary imbalance, 159–186 genes and environment, 141–156 high blood pressure, 311–352 obesity, 223–262 other personal factors, 431–491 physical activity, 191–217 smoking, 395–424 social and physical environment, 503–528 Determinants of Myocardial Infarction Onset Study, 179 Developing countries. see also Disease Control Priorities in Developing Countries Project (World Bank) community intervention, 632–634 mortality patterns in, 5–6 prevention and control, 561–563 DGA (Food Guide Pyramid), 163–164 Diabetes and metabolic syndrome, 361–388 asymptomatic hyperglycemia and, 375–377 biological mechanisms of, 368–369 blood insulin and, 377 body mass index and, 237 community/population-wide measures, 384–385 concepts and definitions, 363–366 coronary heart disease and, 375–378 determinants of, 367–368 diabetes control, 383–384 genetic factors and, 368 large/small vessel disease, 381–382 measurement methods, 366–367 metabolic syndrome, 363–366, 372–373, 384 population distribution and, 369–375 prevention and control, 382–387 public health issues, 374–375, 382, 385–386 race/ethnicity and, 369–374, 378 screening tests, 366–367, 387 stroke and, 378–381 Diabetes Control and Complications Trial (DCCT), 384 Diabetes in America study (1995), 377 Diabetes mellitus. see Diabetes and metabolic syndrome Diabetes Prevention Program (DPP), 382–383 Diabetes Self-Management Education (DSME), 383, 384–385 Diastolic pressure. see also High blood pressure (HBP) dysfunction of, 124 hypertension and, 313, 314–318 measurement methods, 317
46394_INDX_697_710.qxd
2/26/10
2:36 PM
Page 701
INDEX
as risk factor, 74–75 stroke and, 101–102 Diet, effects of, 172–180 affluent society diet, 160–161, 164, 165 Atkins Diet, 165, 179 childhood recommendations, 181, 182 coffee and tea, 178–179 DASH Trial, 163–164, 179–180, 321, 342, 343 dietary cholesterol, 172–176 dietary fiber, 160–161, 163–165, 178, 506 fats and coronary heart disease, 173–176 fats and total cholesterol, 172 folate, 181, 432, 444, 487–488, 595 high blood pressure and, 319–322, 342, 343–344, 350–351 hunter-gatherer diet, 160–161 Mediterranean diet, 164, 175, 179 mortality patterns and, 179 nutrients and, 179–180 omega-3 fatty acids, 177–178 optimal dietary pattern, 180–182 Ornish Diet, 165, 179 popular diets, 164–165 public health issues, 184–185 saturated fat consumption, 161 Sugar Busters! diet, 165 TLC diet, 165 trans fatty acids, 176–177 Weight Watchers Diet, 165, 179 Zone Diet, 179 Diet, Nutrition and the Prevention of Chronic Diseases (WHO), 184 Diet and Health study (NRC) on alcohol consumption, 434, 437 on changes in dietary patterns, 161 on diet and chronic disease, 172 on dietary assessment methods, 166 on dietary fiber, 178 on high blood pressure, 319–320 on nutrients, 162–163 Diet and Lifestyle Recommendations Revision 2006 (AHA), 180–181 Dietary Assessment Resource Manual, 166 Dietary imbalance, 159–186 blood lipids and, 275 cardiovascular-related effects of diet, 172–180 changes in national dietary patterns, 161–162 community/population-wide measures, 182–184 concepts and definitions, 161–165 dietary composition, 164 dietary prescriptions, 165 evolution of eating patterns, 160–161 food choice determinants, 168–170 foods and, 163–164 measurement methods, 165–168 nutrients and, 162–165, 179–180 optimal dietary pattern, 180–182 popular diets and, 164–165 population distribution and, 170–172 prevention and control, 180–185 Dietary Intervention Study in Children (DISC), 296 Diet-Heart Feasibility Study, 172 Disability-adjusted life years (DALYs) alcohol consumption and, 445
blood lipids and, 290–291 blood pressure and, 338–340, 350 coronary heart disease and, 73–74 depression and, 460 diabetes and, 375–376 global projections for, 14–15, 638 obesity and, 233 physical inactivity and, 211 risk factors and, 16 smoking and, 413, 585 stroke and, 98–99, 107 DISC (Dietary Intervention Study in Children), 296 Disease Control Priorities in Developing Countries Project (World Bank) action plans, 665 on alcohol consumption, 444–446 on blood lipids, 301–302 on chronic diseases, 620 on depression, 460 on incidence of diabetes, 375 on obesity interventions, 261 on prevention and control, 561–562 scope of, 7, 613 on smoking, 403, 413 DPP (Diabetes Prevention Program), 382–383 DSME (Diabetes Self-Management Education), 383, 384–385 Dying Too Young (2005), 646–647 EACPR (European Association for Cardiovascular Prevention and Rehabilitation), 683 Eating pattern determinants, 168–170 EBM (Evidence-based medicine), use of term, 570 EBPH (Evidence-based public health), use of term, 570 ECG-LVH (Left ventricular hypertrophy), 129 Economic issues amenable mortality in U.S., 640 cost-effectiveness of intervention, 301–303, 385–388 cost-effectiveness of prevention, 640–642, 673 decision making and, 571, 583–587 in global public health efforts, 9–10, 385–387 income groups and, 13–14, 17 medical care costs in U.S., 13–14 prevention strategies and, 640–642 research funding, 690–691 Education levels, as risk factor, 27, 99, 506–509 Eight Americas project, 30–31, 33, 510 Elderly, 25–26, 345, 604. See also Age Emergency care/acute case management, use of term, 555–556, 560–561 Emphysema, 131 Ending the Tobacco Problem (2007), 415 End-of-life care, use of term, 556, 561 Endogenous tissue-type plasminogen activator (tPA), 474 Environment, use of term, 503 Environmental factors. see Genes and environment; Social and physical environment The Environment and Disease (Hill), 539–540
701
EPIC (European Prospective Investigation into Cancer and Nutrition), 179, 440–441 Epidemiologic transition theory public health issues and, 7–9, 32 westernization and, 503, 514 Epidemiology of Diabetes and Its Vascular Lesions, 362 Epigenetic epidemiology, 143 Epinephrine, 323 EROS (European Registers of Stroke) Collaboration, 97 ESC (European Society of Cardiology), 577, 597 Essential vascular package (EVP), 476 Estrogen. see Hormone replacement therapy (HRT) Ethnicity. see Race/ethnicity EURAMIC study, 177 European Association for Cardiovascular Prevention and Rehabilitation (EACPR), 683 European Atherosclerosis Society, 597 European Cardiovascular Indicators Surveillance Set (EUROCISS), 683 European Collaborative Trial of Multifactorial Prevention of Coronary Heart Disease, 631 European Prospective Investigation into Cancer and Nutrition (EPIC), 179, 440–441 European Registers of Stroke (EROS) Collaboration, 97 European Society of Cardiology (ESC), 577, 597 European Society of Hypertension, 597 Evaluating Community Efforts to Prevent Cardiovascular Diseases, 684 Evans County Study, 22 Evidence-based decision making, 567–587 ACC/AHA approach, 575–577 clinical intervention, 574–579 Cochrane Collaboration and, 574–575 community intervention, 579–583 economic evaluations and, 583–587 evaluation of evidence, 573–579 external validity, 573, 586 GCPS approach, 579–584 Levy’s arrow, 568–569 nature of evidence, 569–570 randomized controlled trials and, 572–573, 575–576 RE-AIM approach, 583, 585 study design algorithm, 580, 581 Evidence-based medicine (EBM), use of term, 570 Evidence-Based Public Health (2003), 569 Evidence-based public health (EBPH), use of term, 570 EVP (Essential vascular package), 476 Exdrinkers, risk factors of, 434. see also Alcohol consumption Exercise, 192–193, 213, 526–527. see also Physical inactivity Expert Panel on Guidelines for Use of Dietary Intake Data, 167 External validity, 573, 586 Familial aggregation, 24, 144 Familial concordance, 24 Familial risks, genes and environment, 142
46394_INDX_697_710.qxd
702
2/26/10
2:19 PM
Page 702
INDEX
Family Blood Pressure Program, 153–154 My Family Health Portrait assessment tool, 150 Family history blood lipids and, 274–275 coronary heart disease and, 146–149 genes and environment and, 146–150 high blood pressure and, 318–319 high-risk families, 146–147 optimal dietary pattern and, 180 parental genetic influences, 149–150 FAO (Food and Agriculture Organization), 161–162, 175–176 Fat distribution, definition, 226 Fatness, definition, 225 Fats, dietary, 172–178. see also Diet, effects of FCTC (Framework Convention on Tobacco Control), 396, 423–424 FDA. See Food and Drug Administration, U.S. (FDA) Fetal and neonatal development, 24 Fetal origins hypothesis, 318 FHS. See Framingham Heart Study (FHS) Fiber, dietary effects of, 160–161, 163–165, 178, 506 recommendations on, 256, 294, 299, 385 Fibrinogen alcohol consumption and, 434–436 coronary events and, 472–473 genetic factors and, 471 insulin resistance and, 369 risk factors and, 470–471 smoking and, 414, 432 thrombosis and, 465–470 Fibrinolysis alcohol consumption and, 435–436 impairment of, 474 obesity and, 236 physical activity and, 196–197 thrombosis and, 400 Finnish Mental Hospital Study, 292–293 Fogarty International Center goals, 685, 687 Folate, 181, 432, 444, 487–488, 595 Food and Agriculture Organization (FAO), 161–162, 175–176 Food and Drug Administration, U.S. (FDA) dietary guidelines, 167 on tobacco regulation, 422 trans fat content labeling, 177 Food and Nutrition Board (IOM), 177 Food chain elements, 168 Food Guide Pyramid (DGA), 163–164 Foods. see also Diet, effects of; Dietary imbalance coffee and tea, 178–179 fish consumption, 177–178 Food Guide Pyramid, 163–164 Framework Convention on Tobacco Control (FCTC), 396, 423–424 Framingham Heart Study (FHS) on blood lipids, 286–290 on congestive heart failure, 128–130 on coronary heart disease, 81 on diabetes and stroke, 381 40 year follow-up, 124–125 on genetic factors in obesity, 230 on incidence of stroke, 96 population diversity of, 22
on risk factors, 16, 71, 77, 543 risk scores and, 606 scope of, 59, 62, 74 on stroke, 103–104 on sudden deaths, 126 website, 12 on weight and disease, 246 Franklin Cardiovascular Health Program, 626–628, 630–631 Fredrickson classification, of blood lipids, 271–272 Fundamentals of Genetic Epidemiology (Khoury), 142 The Future of Public Health report (1988), 551, 609–610 The Future of the Public’s Health in the 21st Century (IOM), 556–557 Gaining Health (WHO-EURO), 648, 661–663 GBDS (Global Burden of Disease Study), 459–460 GCPS. See Guide to Community Preventive Services (GCPS) Gender. see Sex General Accounting Office (GAO), on lipid measurement standards, 273 Genes and environment, 141–156. see also specific diseases blood lipids, 153, 271–272 blood pressure, 153–154 cardiovascular applications of genomic epidemiology, 152–155 cardiovascular diseases and, 143–144 epigenetic epidemiology, 143 familial risks, 142 family history and, 146–150 genetic complexity of coronary heart disease, 143 genetic epidemiological concepts, 144–145 genetic testing, 156 genome-wide association studies, 145, 152 interaction of, 150–152 linkage analysis, 144–145 population-based family designs, 145 public health issues, 155 research strategies and, 155 stroke, 154–155 Genetic Factors in Coronary Heart Disease (Goldbourt et al.), 152, 368 Genetics and Public Health in the 21st Century, 143 Genetic testing, 156 Genome-wide association (GWA) studies, 145, 152 Geographic information systems (GIS), 30, 525 Geography/place disparities in Appalachia, 30, 32 in eight Americas, 30–31, 33, 510 health patterns and, 29–31 German Cardiovascular Prevention Study, 631–632 GIS (Geographic information systems), 30, 525 Global Burden of Disease and Risk Factors Study (World Bank) on blood lipids, 290 on depression, 460
on diabetes, 382 on global burden of risk, 636, 638–639, 642 on limitations of death registrations, 65 on obesity, 253 on physical inactivity, 211 on public health issues by income/region, 13–14 on risk factors, 16 on smoking, 396 on stroke, 107 Global Burden of Disease Study (GBDS), 459–460 Globalization, social factors and, 503, 514 Global public health issues. see Public health issues; specific issues Global Strategy for the Prevention and Control of Noncommunicable Diseases (WHO), 688–689 Global Strategy on Diet, Physical Activity and Health (WHO), 184–185, 191, 192, 216, 613 Global Tobacco Epidemic, 2008 (WHO), 396, 415, 424 Global Youth Tobacco Survey (GYTS), 404, 405 Glucose metabolism. see Diabetes and metabolic syndrome GNP (Gross National Product), per capita food consumption and, 161–162 Greater Cincinnati/Northern Kentucky Stroke Study, 96 Guide for Improving Cardiovascular Health at the Community Level (AHA), 607, 658 Guidelines and policies, 591–614. see also specific organizations on children and adolescents, 604–605 clinical guidelines, 594–607 community guidelines, 607–609 current recommendations, 593 on developing country populations, 612–613 on elderly persons, 604 in Europe, 597–598 on groups of special concern, 600–605 history of, 592–593 individual total risk and, 601–603 policy models, 642–644 on prevention of CHD, 594–596 on prevention of ischemic stroke, 596–597 public policies, 609–613 risk scores and, 605–607 on secondary prevention, 600 on women’s health, 603 worldwide, 598–600 Guide to Community Preventive Services (GCPS) on evidence-based decision making, 570–571, 579–584 guidelines and policies, 607 GWA (Genome-wide association) studies, 145, 152 GYTS (Global Youth Tobacco Survey), 404, 405 HAART (Active antiretroviral therapy), 481 HALE (Healthy life expectancy), 645 Harvard Six Cities Study, 523
46394_INDX_697_710.qxd
2/26/10
2:19 PM
Page 703
INDEX
HBP. See High blood pressure (HBP) HDFP (Hypertension Detection and Follow-Up Program), 345–346, 348, 624 HDL-cholesterol. see also Blood lipids alcohol consumption and, 434–435 alcohol consumption and metabolism of, 152 coronary heart disease and, 270 Health, United States reports (CDC), 66, 238 Health Affairs, on race/ethnicity disparities, 28 Health Agenda for the Americas, 2008–2017, 659–660 Health and Human Services, U.S. (HHS). see also Healthy People 2010 (HHS) Physical Activity Guidelines for Americans, 192 Report of the Secretary’s Task Force on Black and Minority Health, 29 Health and Retirement Study, 208 The Health Consequences of Involuntary Exposure to Tobacco Smoke (2006), 396 Health educationalist approach, 670–671 Health family trees, 146 Health Professionals Follow-Up Study, 122–123, 178, 436–437, 444 Health promotion, 553, 673 Health-related quality of life (HRQoL), 73, 98–99 HealthStyles surveys, 150, 348 Healthy Eating Index (HEI), 170 Healthy life expectancy (HALE), 645 Healthy People 2010 (HHS) on blood pressure, 327 on diabetes, 372 goals of, 21, 561 guidelines and policies, 611 on heart disease, 127 on high cholesterol, 277 on levels of prevention, 674 physical activity and, 202 scope of, 647, 658, 666 on stages of prevention, 551 on stroke, 99 Healthy weight, definition, 228–229 Heart attacks. see Myocardial infarction Heart Disease and Stroke Prevention Division (CDC), 612, 647 Heart Disease and Stroke Statistics-2009 Update, 135 Heart failure. see Congestive heart failure (CHF) Heartfile, 661–662 Heart Health Network action plans, 662–665 Heart Outcomes Prevention Evaluation (HOPE), 485, 488 HEI (Healthy Eating Index), 170 Helicobacter pylori, 46–47 The Helsinki Multifactorial Primary Prevention Trial, 624 Hemostatic factors, 432, 464–475 alcohol consumption and, 435–436 coagulation factors, 465 coagulation pathways, 466–470 established risk factors and, 470–471 fibrinogen, 369, 414, 432, 434–436, 465–473 fibrinolysis and, 172, 196–197, 236, 400, 435–436, 472, 474
genetic factors and, 471 measurement methods, 465–466 menopause and, 472 population distribution and, 471–472 prevention and control, 474–475 risk factors and, 470–471 studies on, 472–474 theoretical schemes, 466 thrombolytic treatment, 475 Henle-Koch postulates, 539 Heredity. see Genes and environment Herpesvirus group, 46 HHS. See Health and Human Services, U.S. (HHS) High blood pressure (HBP) age and, 317–318 biological mechanisms of, 322–323 body mass index and, 237 classification of, 314–316 community/population-wide measures, 348–350 concepts and definitions, 312–316 as CVD determinant, 311–352 determinants of, 317–322 dietary issues and, 319–322, 342, 343–344, 350–351 family history and, 318–319 incidence of, 333–336 individual differences, 336–338 measurement methods, 316–317 population differences, 336 population distribution and, 323–336 prevention and control, 341–351 public health issues, 338–339, 350–351 race/ethnicity and, 317–318, 323–326, 336 sex and, 317–318 stroke and, 16, 103–104, 336–339 treatment algorithm, 344 High cholesterol, 153, 269, 274. see also Blood lipids High-risk families, 146–147. see also Family history HIV/AIDS, 432, 481 HMG CoA reductase, 294 Homocysteine Studies Collaboration (2002), 486–487 Honolulu Heart Program, 176, 434–435, 460–461 HOPE (Heart Outcomes Prevention Evaluation), 485, 488 Hormone replacement therapy (HRT), 432, 472, 476–479 Hostility and anger, 454–455. see also Type A behavior pattern HRQoL (Health-related quality of life), 73, 98–99 HRT (Hormone replacement therapy), 432, 472, 476–479 Human Body Composition, 225 Human Genome Epidemiology (Khoury et al.), 143 Human Genome Project, 156 Hunter-gatherer diet, 160–161 Hypercholesterolemia, 153, 269, 274. see also Blood lipids Hyperglycemia, 52, 361. see also Diabetes and metabolic syndrome Hyperhomocysteinemia, 432, 476, 485–488 Hyperlipidemia. see also Blood lipids
703
causal pathways, 544 Fredrickson classification of, 271–272 impaired fibrinolysis and, 472 prevention and control, 610, 620–621 screening tests, 384, 604 secondary, 275–276 Hypertension. see also High blood pressure (HBP) essential vs. secondary, 313–314 use of term, 54 Hypertension Control (WHO), 348–349 Hypertension Detection and Follow-Up Program (HDFP), 345–346, 348, 624 Hypertension Primer, 322–323 IAP (International Atherosclerosis Project), 47–49 ICD 10. See International Statistical Classification of Diseases and Related Health Problems (ICD 10) IHD, use of term, 4 IMPACT model, 81–82, 642 Implementation research, 685. see also Prevention research agendas IMT (Intimal-medial thickness), 43, 50 Incidence & Prevalence (NHLBI) on congestive heart failure, 128–129 on current studies, 69, 72 on incidence of stroke, 96 Income, mortality patterns and, 506–509 Indian Polycap Study (TIPS), 476 Indirect auscultatory method, 316 Infant death rates, social factors and, 515 Infection, as risk factor, 477–478, 480–481 Inflammation, as risk factor, 46, 488–489 Influenza vaccinations, 481 Institute of Medicine (IOM) on childhood obesity, 256–257, 262 on combination pharmacotherapy, 476 Control of Cardiovascular Diseases in Developing Countries, 691–692 Diet and Health report, 181 Ending the Tobacco Problem, 415 The Future of the Public’s Health in the 21st Century, 556–557 Preventing Childhood Obesity, 256–257 Progress in Preventing Childhood Obesity, 262 research agendas and, 691–692 Social-Ecological Framework, 557 on social environments, 504 on trans fatty acids, 177 Insulin-glucose regulation. see Diabetes and metabolic syndrome Insulin resistance, obesity and, 236 Insulin Resistance Atherosclerosis Study (IRAS), 369 INTERHEART Study on acute myocardial infarction, 76 on blood lipids, 290 on community intervention, 671 on dietary patterns, 180 on global burden of risk, 639, 640 on obesity, 250–252 on psychosocial factors, 462 on smoking, 410–413 INTERMAP study, 167, 321–322 Intermittent claudication, use of term, 114 International Action on Cardiovascular Disease, 648–649, 662–664
46394_INDX_697_710.qxd
704
2/26/10
2:19 PM
Page 704
INDEX
International Atherosclerosis Project (IAP), 47–49 International Heart Health Society, 648, 662–664, 689 International Journal of Epidemiology, on genetic epidemiology, 143 International Obesity Task Force (IOTF), 243, 256, 261 International Physical Activity Questionnaire (IPAQ), 202–204 International Society of Hypertension, 351 International Statistical Classification of Diseases and Related Health Problems (ICD 10) atherosclerosis and, 44 circulatory system diseases, 4–5 classification of diabetes, 363 coronary heart disease and, 64 related conditions, 111–112 stroke and, 92–93 International Studies of Infarct Survival (ISIS) Study, 471 International Symposium on Moderate Drinking and Health, 445 International Textbook of Diabetes Mellitus, 363 International Union for Health Promotion and Education (IUHPE), 553 INTERSALT Study on alcohol consumption, 437, 440 on blood pressure, 317–318 on body mass index, 243 on sodium and HBP, 321 Inter-Society Commission for Heart Disease Resources guidelines and policies, 610–611 Primary Prevention of the Athersclerotic Diseases, 610 Intimal-medial thickness (IMT), 43, 50 Investing in Health Research and Development (WHO), 687–688 IOM. See Institute of Medicine (IOM) IOTF (International Obesity Task Force), 243, 256, 261 IPAQ (International Physical Activity Questionnaire), 202–204 IRAS (Insulin Resistance Atherosclerosis Study), 369 Ireland-Boston Study, 176 Ischemic heart disease alcohol consumption and, 444 diabetes and, 378 genetic factors, 153 obesity and, 254–255 physical inactivity and, 211 risk factors of, 16–17 ISIS (International Studies of Infarct Survival) Study, 471 IUHPE (International Union for Health Promotion and Education), 553 Jenkins Activity Survey (JAS), 454 Job strain, 455–458 Journal of Chronic Diseases, on HenleKoch postulates, 539 Justification for Use of Statins in Prevention (JUPITER), 489 Kenyan Luo Migrant Study, 520 Know Your Body Program, 634
Korotkov sounds, 316 Kuopio Ischemic Heart Disease Risk Factor Survey, 471 The Lancet on alcohol consumption, 444–445 on chronic diseases, 620, 641, 649–650 on genetic epidemiology, 144 on global public health impacts, 16 on health promotion in clinical practice, 673 The Lancet Neurology, on stroke, 107 Large-vessel PAD (LV-PAD), 114–115, 119 LDL-cholesterol, 269, 273, 485. see also Blood lipids Lectures on Angina Pectoris and Allied States (Osler), 62 Left ventricular hypertrophy (ECG-LVH), 129 Leisure time physical activity (LTPA), 193, 208 Levy’s arrow (research continuum), 568–569 Life Course Perspectives on Coronary Heart Disease, Stroke, and Diabetes (WHO), 688 Life course research, 688 Life-course view, on obesity, 230–232 Life expectancy at birth, 8–9 Lifestyle Modification for the Prevention and Treatment of Hypertension, 343 Linkage analysis, genes and environment, 144–145 Lipids. see Blood lipids Lipid Standardization Program (CDC), 273 Lipoprotein molecules, 270–271 London School of Hygiene questionnaire, 65, 114 Longitudinal Investigation of Thromboembolism Etiology, 132–133 LTPA (Leisure time physical activity), 193, 208 LV-PAD (Large-vessel PAD), 114–115, 119 Magnetic resonance imaging (MRI), 43, 92, 93–94 Marketing influences, on smoking behaviors, 398 Measurement methods. see specific methods and diseases Medical Expenditure Panel Survey (MEPS), 73, 98–100 Medical Nutrition Therapy (MNT), 383 Medicine and Science in Sports and Exercise, on assessment questionnaires, 194 Mediterranean diet, 164, 175, 179 MED PED program, 146, 274 Men. See Sex Menopause hemostatic factors and, 472 HRT and, 432, 472, 476–479 MEPS (Medical Expenditure Panel Survey), 73, 98–100 MESA (Multi-Ethnic Study of Atherosclerosis), 43, 232, 524 Metabolic syndrome (MetS). see Diabetes and metabolic syndrome
Methodology Manual for ACC/AHA Guideline Writing Committees, 575–577 Metropolitan Life Insurance Company, on desirable weight tables, 225 MetS (metabolic syndrome). see Diabetes and metabolic syndrome Mexican-Americans. see Race/ethnicity Microalbuminuria, 381 Migration blood pressure and, 322 social factors and, 513, 518–520 Minnesota Heart Health Program, 626–630 Minnesota Heart Survey, 64 Minnesota Leisure Time Physical Activity questionnaire, 206 Minnesota Multiphasic Personality Inventory (MMPI), 449, 454 MNT (Medical Nutrition Therapy), 383 Modified Framingham Stroke Risk Profile, 103 MONICA Project (WHO) on alcohol consumption, 151–152, 434, 435 on antioxidants, 481–482 on blood pressure, 330–333, 335 on coronary heart disease, 59, 82–83 diagnostic elements, 63–64 on global cholesterol levels, 281–282 population diversity of, 71–72 research agendas and, 691 on stroke, 92, 94–96, 98, 105–107 on worldwide mortality rates, 67–69, 94–96, 105–107 MONItoring Trends and Determinants in CArdiovascular Disease. see MONICA Project Mortality patterns. See also specific studies age and, 20–21, 23–24 aortic aneurysms and, 121–122 arteriosclerosis and, 173 congestive heart failure and, 125–128, 131 coronary heart disease and, 65–67 in developing countries, 5–6 diet and, 179 education and, 506–509 in eight Americas, 30–31, 33, 510 globally, 5–6 income and, 506–509 in industrial countries, 5–6 myocardial infarction and, 66 peripheral arterial disease and, 115 physical inactivity and, 209–211 stroke and, 91–96, 98–99, 104–107 subclinical disease and, 44–45 in U.S., 6–7, 12–13, 19–22, 66, 253, 510 venous thromboembolism and, 133–134 WHO data (1950–1987), 32, 34–35 in women, ages 35 and older, 31 World Bank estimates (1988–1998), 10–11 World Bank estimates (2001), 13–14 MPOWER Package (smoking prevention), 396, 415, 424 MRFIT. See Multiple Risk Factor Intervention Trial (MRFIT) MRI (Magnetic resonance imaging), 43, 92 stroke and, 93–94
46394_INDX_697_710.qxd
2/26/10
2:19 PM
Page 705
INDEX
Multi-Ethnic Study of Atherosclerosis (MESA), 43, 232, 524 Multi-factor primary prevention, 621–636 intervention in communities, 625–634 intervention in individuals, 622–625 intervention in youth, 634 in North Karelia Project, 625–628 outcomes, 634–636 trials, 622–625 The Multifactor Primary Prevention Trial, 623–624 Multi-infarct dementia, 91 Multinational Study of Vascular Disease in Diabetics (WHO), 381 Multiple Risk Factor Intervention Trial (MRFIT) on blood lipids, 285–286 on coronary heart disease, 75–76 on diabetes, 378, 380 on dietary assessment methods, 167 intervention in individuals, 622–623 mortality patterns and, 639 on physical inactivity, 206–207 on smoking, 414 on stroke, 102–103 on Type A behavior pattern, 454 Muscatine Study, 54 Myocardial infarction. see also Atherosclerosis alcohol consumption and, 151–152, 435–436 atherosclerotic plaque and, 41–42 causal pathways, 544 classification of, 64 elderly survival of, 25–26 heart failure and, 124 high-risk families and, 146–149 job strain and, 456 mortality patterns and, 66 physical activity and, 208 rehabilitation with exercise, 213 thrombosis and, 62 National Action Plan for Prevention and Control of Non-Communicable Diseases and Health Promotion in Pakistan, 661–662, 664 National Bureau of Standards, 273 National Center for Health Statistics (CDC), 12 Assessing Physical Fitness and Physical Activity in Population-Based Surveys, 194 Health, United States reports, 66, 238 on stroke, 93–94 National Cholesterol Education Program (NCEP) on blood lipid control, 270, 294–295, 300–301, 303–304 on diabetes and coronary heart disease, 362 on global burden of risk, 636–638 guidelines and policies, 594–596, 600, 603 guidelines for children, 272–273 on HIV/AIDS, 481 on Therapeutic Lifestyle Changes, 165 National Cooperative Pooling Project, U.S. blood lipids and, 285, 287 population diversity and, 22 on risk factors, 74–76
on smoking, 407–409 on weight and disease, 245–246 National Council of State Legislatures, 612 National Diabetes Fact Sheet website (CDC), 371 National Forum for Heart Disease and Stroke Prevention, 611–612, 658, 668–670 National Health and Nutrition Examination Surveys (NHANES) on ABI categories, 115 on blood pressure, 317, 335–336, 348 on congestive heart failure, 126–127, 129 continuous data collection, 12 on diabetes, 361, 369–371 on global burden of risk, 636–638 on high blood pressure, 323–324 on incidence of stroke, 98 on metabolic syndrome, 372–373, 378 on modifiable dietary factors, 179 on mortality decline (II Study), 81 on mortality decline (I Study), 81 on nutrition, 170 on prehypertension, 314, 326 on prevalence of high cholesterol in U.S., 277–280 on pro-oxidants, 485 proposed goals of, 685–687 research agendas and, 685–687 on risk factors (III Study), 77 on smoking, 400, 401–403, 414 on sodium and HBP, 320 National Health Education Committee, Inc., A Statement on Arteriosclerosis, Main Cause of “Heart Attacks” and “Strokes,” 543 National Health Interview Survey (2004) on alcohol consumption, 437 on ischemic heart disease, 378 on prediabetes, 383 on race/ethnicity disparities, 28 on smoking, 400 National Health Service (UK), 294, 296 National Heart, Lung and Blood Institute (NHLBI), 12. see also National Cholesterol Education Program (NCEP) Cardiovascular Health and Disease in Women report, 27 Chart Book, 70, 96 on CRP, 488–489 Decline Conference reports (1978), 67, 79–80, 642 guidelines and policies, 594–596, 604 on heart failure mortality rates, 126 Incidence & Prevalence, 69, 72, 96, 128–129 on incidence of stroke, 96 research agendas and, 685–687, 691 Task Force on Research in Epidemiology and Prevention of Cardiovascular Diseases, 685–687 Working Group on Genome Wide Association in NHLBI Cohorts, 155 National High Blood Pressure Education Program, 326, 341–342 National Hospital Discharge Surveys, 377–378 National Human Genome Research Institute, 155 National Institutes of Health (NIH) as funding source, 690
705
on health disparities, 20 on obesity, 224, 225 on population diversity, 22 research agendas and, 685, 687 on smoking, 415 National Longitudinal Mortality Study (1979–1989), 29, 506–508 National Mortality and Morbidity Air Pollution Study (NMAPS), 523 National Mortality Follow-Back Survey (1986), 378 National Office of Genomics and Disease Prevention (CDC), 155 National Research Council (NRC). see Diet and Health study (NRC) National Survey on Drug Use and Health, 437 Nationwide Food Consumption Survey (NFCS), 170 NCEP. See National Cholesterol Education Program (NCEP) n-dimensional complex, 540–541, 682 Neighborhood characteristics, as risk factor, 524–527 NEMS-S (Nutrition Environment Measures Survey-Stores), 169 NHANES. See National Health and Nutrition Examination Surveys (NHANES) NHLBI. See National Heart, Lung and Blood Institute (NHLBI) Nicotine addiction, 396, 400. See also Smoking, and other tobacco use NIH. See National Institutes of Health (NIH) Ni-Hon-San Study, 62, 519 NMAPS (National Mortality and Morbidity Air Pollution Study), 523 Norepinephrine, 323 North Karelia Project, 625–628 Northwick Park Heart Study (NPHS), 471–472 Norwegian Adolescent Follow-Up Study, 248–249 Norwegian Vitamin (NORVIT) Trial, 488 Novel risk factors, 489–491 Nuclear lung scans, 134 Nurses Health Study on alcohol consumption, 436–437 on body mass index, 236 on smoking, 409 on smoking and BMI, 246–247 on trans fatty acids, 176 Nutrition Environment Measures SurveyStores (NEMS-S), 169 Obesity, 223–262 biological mechanisms of, 233–236 community/population-wide measures, 259–261 concepts and definitions, 224–229 determinants of, 229–232 genetic factors and, 230–231 measurement methods, 229 population distribution and, 236–244 prevention and control, 253–262, 572 public health issues, 241, 243–244, 253, 261–262 race/ethnicity and, 238, 241 risk factors of, 230 sex and, 243, 248–249
46394_INDX_697_710.qxd
706
2/26/10
2:19 PM
Page 706
INDEX
(Obesity cont’d) studies on, 244–252 treatment algorithm, 257 validity of surveys and, 170 Obesity: Preventing and Managing the Global Epidemic (WHO), 259–261 Obesity Education Initiative, 226 Occupational stress, 455–458 Omega-3 fatty acids, 177–178. see also Diet, effects of Ontario Survey on the Prevalence and Control of Hypertension (2006), 332–333 Ornish Diet, 165, 179 The Oslo Study, 623 Overweight, definition, 225–227. see also Obesity Oxygen supply, 196–197 PAD. See Peripheral arterial disease (PAD) Palpitations. see Arrhythmias Partnership for Prevention, 658 Partnership to Fight Chronic Disease, 658 Pathobiological Determinants of Atherosclerosis in Youth (PDAY) Study, 49–54, 249, 412–413 Paul Coverdell National Acute Stroke Registry, 93 Pawtucket Heart Health Program, 626–628, 630 PE. See Pulmonary embolism (PE) Peasant agriculturalist diet, 160–161 Pediatrics, on childhood obesity, 256 Perinatal Collaborative Project, 318 Peripheral arterial disease (PAD) ankle-brachial index, 120 ankle-brachial index and, 114–118 background of, 114 cardiovascular diseases and, 111–120 classification of, 112 hospitalization and, 112 mortality patterns and, 115 population studies, 114 race/ethnicity and, 115, 117–118 risk factors of, 115–119 sex and, 117–118 Person, place, and time factors atherosclerosis and, 47–49 health patterns and, 32–35 scales of time, 32 WHO data, 32, 34–35 Personal factors, in cardiovascular disease risk, 431–491 adverse psychosocial factors, 431–432, 447–464 alcohol consumption, 433–447 emerging factors, 476–491 hemostatic factors, 432, 464–475 Pew Health Professions Commission, 691 Physical activity, definition, 192 Physical Activity and Health, 191–192 Physical Activity and Public Health, 213 Physical Activity Guidelines for Americans (HHS, 2008) on physical inactivity, 192, 206, 212–213 physical inactivity and, 211 Physical environment. see Social and physical environment Physical fitness, definition, 193–194 Physical inactivity, 191–217 biological mechanisms of, 196–198
cardiovascular-related effects of, 204–211 community/population-wide measures, 214–216 concepts and definitions, 192–194 determinants of, 194, 196 measurement methods, 194 population distribution and, 197, 199–204 prevention and control, 211–217 public health issues, 211 race/ethnicity and, 197–202 sex and, 197–202, 207 social factors and, 526–527 Physicians Health Study, 435–436, 442, 474 PIOPED (Prospective Investigation of Pulmonary Embolism), 134 Policies. see Guidelines and policies; specific organizations Policy analysis, 684–685 Policy Framework Statement for Regional and Global Partnerships, 611–612 Polypills. see Combination pharmacotherapy Popular diets, 164–165 Population-based family designs, 145 Population-Based Prevention of Obesity (AHA), 259 Population-wide prevention strategies action plans, 670–673 alcohol consumption and, 445 blood lipids and, 299–301 community intervention in North Karelia, 625, 626–628 community intervention in U.S., 626–634 diabetes/metabolic syndrome and, 384–385 heart healthy nutrition and, 182–184 high blood pressure and, 348–350 physical inactivity and, 214–216 smoking and tobacco use, 419–422 walkability indexes, 216 Populomics, 680, 692–693 Prediabetes, use of term, 363, 368. see also Diabetes and metabolic syndrome Pregnancy, 230, 385, 423, 445, 578 Present Reality continuum, 546–547, 552, 560, 562 Preventing Childhood Obesity (IOM), 256–257 Preventing Chronic Diseases (WHO), 620 Preventing Coronary Heart Disease in South Asia (SAARC), 662 Preventing Death and Disability from Cardiovascular Diseases (CDC/HNLBI), 666 Preventing Tobacco Use Among Young People, 398 Prevention, evidence for, 619–650 counter-arguments, 621, 649 economic issues and, 640–642 future trends, 646–649 global burden of risk and, 636–640, 642–647 intervention in communities, 625, 628–634 intervention in individuals, 622–625 intervention in youth, 634 low risk populations and, 639–640
multi-factor primary prevention and, 621–636 policy models, 642–644 predictive models, 642–647 Prevention and Control of Noncommunicable Diseases, 665 Prevention Effectiveness, 583 Prevention in Childhood and Youth of Adult Cardiovascular Diseases (WHO), 301, 342, 423 Prevention of Cardiovascular diseases (WHO), 350 Prevention of Coronary Disease (WHO), 612 Prevention of Diabetes Mellitus (WHO), 363–364 The Prevention of Stroke (Gorelick and Alter), 92 Prevention research agendas, 679–693 capacity requirements, 690–692 CDC Action Plan, 689–690 concepts of epidemiology, 681–682 Fogarty International Center goals, 687 goals of, 682–683 implementation research, 685 institutionalization and, 691–692 International Heart Health Society, 689 NHLBI goals, 685–687, 691 outcomes, 682–683 personnel training, 691 policy analysis, 684–685 populomics, 692–693 program evaluations, 684 sponsorship and funding, 690–691 surveillance strategies, 683 systems research, 685 WHO goals, 687–689 Prevention strategies, 551–565 for atherosclerosis, 54–55 for blood lipids, 292–299 concepts and language of, 552–557 for congestive heart failure, 131 developing country perspective, 561–563 for diabetes/metabolic syndrome, 382–387 for dietary imbalance, 180–185 emergency care/acute case management, 555–556, 560 end-of-life care, 556, 561 in global public health, 15–17, 216–217 hemostatic factors and, 474–475 for high blood pressure, 341–351 high-risk approach, 563–564 intervention approaches, 560–561 lifestyle changes, 556 multi-factor intervention, 556 for obesity, 253–262, 572 outcomes, 557–560 pharmacotherapy, 556 for physical inactivity, 211–217 policy and environmental change, 560 psychosocial factors and, 462–464 public health framework, 557–558 rehabilitation/long-term case management, 556, 561 risk factor detection and control, 555, 560 settings of, 556–557 single-factor intervention, 556 for smoking, 414–424 Preventive Medicine on children’s physical activity, 192 on physical inactivity, 214
46394_INDX_697_710.qxd
2/26/10
2:19 PM
Page 707
INDEX
Preventive Services Task Force (USPSTF) on adult obesity, 257–258 on behavioral counseling, 182, 213, 445 on blood lipids, 303–304 on blood pressure, 342 on childhood obesity, 256 on diabetes, 384 on evidence-based decision making, 577–578 on hemostatic factors, 474 on HRT, 477 on obesity, 118 on smoking, 417 Primary Prevention of Essential Hypertension (WHO), 341, 612–613 Primary Prevention of Heart Failure (AHA), 131 Primary Prevention of Ischemic Stroke (AHA), 103 Primary Prevention of the Athersclerotic Diseases, 610 Primordial prevention, 551, 554, 558–559 PROCAM Study, 606 Progress in Preventing Childhood Obesity (IOM), 262 Project HeartBeat!, 234, 236, 282, 284 Pro-oxidants, 485 Prospective Diabetes Study (UKPDS), 384 Prospective Investigation of Pulmonary Embolism (PIOPED), 134 Prospective Studies Collaboration, 101–103, 337–338 Psychosocial factors, 431–432, 447–464 depression, 456–460 favorable attributes, 452–453 measurement methods, 449 occupational stress, 455–458 prevention and control, 462–464 public health issues, 464 race/ethnicity and, 450 sex and, 450 social support, 460–462 stress research, 449–452 theoretical background of, 448–449 Type A behavior pattern, 453–455 A Public Health Action Plan to Prevent Heart Disease and Stroke (CDC) background of, 657, 665–666 components of, 666–667 implementation of, 668–669 institutionalization of, 669 research agendas and, 680, 683, 684, 689–690 scope of, 611 strategic imperatives, 658, 673–675 summary of determinants, 547 Public health issues, 3–35. see also Population-wide prevention strategies; Prevention strategies age and life stages, 22–26 blood lipids, 290, 299–302 cartography of, 29–30 diabetes/metabolic syndrome, 374–375, 382, 385–386 dietary patterns, 170, 184–185 disability-adjusted life years (DALYs), 14–15, 638 economic issues, 9–10, 385–387 epidemiologic transition theory, 7–9, 32, 503, 514
genes and environment, 155 geography and place of, 29–30 global concerns, 5–10, 14–15 health disparities, 20–22 high blood pressure, 338–339, 350–351 international classification, 4–5 International Physical Activity Questionnaire, 216–217 list of characteristics, 490 mortality patterns in U.S., 6–7, 12–13, 19–22, 30–31, 33 obesity, 241, 243–244, 253, 261–262 occurrence rates, 10–11 person, place, and time, 32–35 physical activity questionnaire, 202–204 prevention and control, 15–17, 216–217, 636–640, 642–647 psychosocial factors, 464 race/ethnicity and, 27–29 salt intake, 350–352 scope of, 3–5 sex and, 26–27 smoking, 403–404, 413–414, 422–424 treaties on, 396, 423–424 world income groups/regions, 13–14 years of life lost (YLL), 14–15, 73–74, 290–291 Public Health Strategies for Preventing and Controlling Overweight and Obesity in School and Worksite Settings, 261 Pulmonary embolism (PE), 111–113, 132–135. see also Venous thromboembolism (VTE) Quetelet’s index. see Body mass index (BMI) A Race Against Time study (2004) on chronic diseases, 16 on economic investment in prevention, 641 on future trends, 646–647 on government intervention, 562–563 on productive years of life lost, 15 Race/ethnicity alcohol consumption and, 437–440 ascertainment of, 28 atherosclerosis and, 44–45, 47, 49, 51 blood lipids and, 274, 277, 280 classification of, 27–28 congestive heart failure and, 126, 129 coronary heart disease and, 66, 70–73 diabetes/metabolic syndrome and, 369–374, 378 high blood pressure and, 317–318, 323–329, 336 modifiability of, 22 and morality patterns in U.S., 19–22 mortality patterns and, 20 obesity and, 238, 241 peripheral arterial disease and, 115, 117–118 physical inactivity and, 197–202 prevalence of risk factors and, 12–13 psychosocial factors and, 450 smoking and, 400, 406 stroke and, 89, 98–99 Racial and Ethnic Disparities in Heart Disease Among Women (CDC), 30–31
707
Randomized controlled trials (RCTs), 572–573, 575–576 RE-AIM approach, on evidence-based decision making, 583, 585 Recommendations. see Guidelines and policies; specific organizations Reducing Tobacco Use (2000), 417 Rehabilitation/long-term case management, use of term, 556, 561 Relative weight, measurement methods, 229 Relevance of Genetics and Genomics for Prevention and Treatment of Cardiovascular Disease, 143 Report of a Joint FAO/WHO Expert Consultation, 184 Report of the Secretary’s Task Force on Black and Minority Health (HHS), 29 Report of the Special Writing Group on Cardiovascular Disease in Women (AHA), 27 Report on Tobacco Control in India, 423 Research. see Prevention research agendas; specific organizations and studies Research-policy interface, 569 Response-to-injury hypothesis, 46 Rheumatic fever/heart disease, 4, 7–8, 30, 124, 665, 689 Rich diet, 160–161, 164, 165 Right ventricular failure, congestive heart failure and, 131 Risk factor, use of term, 489 Robert Wood Johnson Foundation, 194, 527 Robin Hood Index, 508 Rose questionnaire, 65, 114 Rotterdam Study, 122 SAARC (South Asian Association for Regional Cooperation), 662 Salt, Diet & Health, on high blood pressure, 319 Salt intake blood pressure and, 319–321, 343 global reduction strategies, 350–352 increases in, 170 mortality patterns and, 179 Schools. see also Population-wide prevention strategies Know Your Body Program, 634 meal choices, 169, 183, 634 smoking prevention programs, 420–421 Science on genetic testing, 156 on high-risk family factors, 146, 149 on physical inactivity, 217 Secondhand smoke exposure, 396–397. See also Smoking, and other tobacco use Sedentary lifestyle. see Physical inactivity Sensible Drinking report, 445, 446 Seven Countries Study, 62, 70–71 blood lipids and, 282–285 on blood pressure, 336 on dietary assessment methods, 166–167 on dietary cholesterol intake, 176 on fats and coronary heart disease, 174–175 obesity and, 244 physical activity and, 204–206 scope of, 59 on smoking, 404, 414 on trans fatty acids, 176–177
46394_INDX_697_710.qxd
708
2/26/10
2:19 PM
Page 708
INDEX
Sex alcohol consumption and, 437–440, 443 aortic aneurysms and, 121 ascertainment of, 26 atherosclerosis and, 44, 47–49, 51–53 blood lipids and, 274, 282, 286–290, 296 blood pressure and, 597–599 classification of, 26 congestive heart failure and, 126, 129–130 coronary heart disease and, 59, 67–68, 70–73, 77–78, 477 health patterns of, 26–27 high blood pressure and, 317–318 modifiability of, 22 and morality patterns in U.S., 19–22 mortality patterns and, 20 obesity and, 243, 248–249 peripheral arterial disease and, 117–118 physical activity and, 197–202, 207 psychosocial factors and, 450 public health issues and, 26–27 smoking and, 406 stroke and, 96 SHEP (Systolic Hypertension in the Elderly Program), 345 Silent infarction, definition, 61 Single-nucleotide polymorphism (SNP), 145 Skinfold thickness, 224 SMARTRAQ (Strategies for Metropolitan Atlanta’s Regional Transportation and Air Quality), 216 Smokefree laws, 422 Smoking, and other tobacco use aortic aneurysms and, 122 biological mechanisms of, 398–400 body mass index and, 237, 246–247 community/population-wide measures, 419–422 concepts and definitions, 396–397 coronary heart disease and, 75–76, 151 harm reduction and, 424 individual differences, 407–413 initiation stages, 398, 399 intervention options, 418 marketing and, 398 measurement methods, 397 other tobacco use and, 395–424 population differences, 404–407 population distribution and, 400–404 pregnancy and, 578 prevention and control, 414–424 public health issues, 403–404, 413–414, 422–424 race/ethnicity and, 400, 406 secondhand smoke, 397 sex and, 406 smokeless tobacco and, 396, 400 treatment algorithm, 419 Smoking and Health report, 539 SNP (Single-nucleotide polymorphism), 145 Social and physical environment, 503–528 birth weight and, 515–518 changes in social conditions, 512–522 diet/physical activity and, 525–526 economic development and, 514 globalization and, 514 infant death rates, 515
migration and, 513, 518–520 neighborhood characteristics, 524–527 particulate air pollution, 522–525 programming mechanism, 518 social change concepts, 512–513 social development and, 515–518 social status, 505–512 societal changes over time, 513–514 socioeconomic status and social change in U.S., 520–522 westernization of population groups, 503, 514 Social disadvantage, 509–512 Social-Ecological Framework (IOM), 557 Social environment. see Social and physical environment Social status, 505–512 disparities in health and, 510 income and education, 506–509 occupational class and, 505–507 Robin Hood Index, 508 social disadvantage and, 509–512 Social support, 460–462 Societal influences, on smoking behaviors, 398 Socioeconomic Inequalities in Health Working Group (EU), 506 Sodium intake. see Salt intake South Asian Association for Regional Cooperation (SAARC), 662 Standards for Medical Care in Diabetes2009, 383 Stanford Five-City Program, 72, 626–630, 671 Stanford Three-City Program, 626–629 A Statement on Arteriosclerosis, Main Cause of “Heart Attacks” and “Strokes,” 211, 543 State programs, (US), 554–555, 612, 666. see also Population-wide prevention strategies Statins, 294–296, 489 Status syndrome, 511–512 Strategies for Metropolitan Atlanta’s Regional Transportation and Air Quality (SMARTRAQ), 216 Strategies of prevention. see Prevention strategies Strategy for the Prevention and Control of Noncommunicable Diseases and Injuries in the Russian Federation, 660 The Strategy of Preventive Medicine (Rose), 552 Stress research, psychosocial factors and, 449–452 Stroke belt region, 92, 94, 105 Stroke (CVA), 89–107 background of, 91–92 body mass index and, 248, 250 case-fatality and, 91, 97–98 cerebral arteries, 89–90 diabetes/metabolic syndrome and, 378–381 diagnostic elements, 92 disability and, 98–100 disparities in, 99 features of individual cases, 90–91 genetic factors and, 154–155 high blood pressure and, 16, 17, 336–339 incidence of, 96–97
mortality patterns and, 91–96, 104–107 population studies, 92–93 prevalence, 98 race/ethnicity and, 89, 93–94, 98–99 risk factors of, 16–17, 101–104 sex and, 96 trends and explanations, 104–107 Sudden death, 65, 208 Sugar Busters! diet, 165 Surgeon General Reports, U.S. on health promotion and disease prevention, 553 on nutrition and health, 184 on smoking, 396, 398, 404, 411–412, 414, 417, 419, 422 Surveillance strategies, 683 Sydney Principles (IOTF), 261 Syndrome X, 364. see also Diabetes and metabolic syndrome Systolic Hypertension in the Elderly Program (SHEP), 345 Systolic pressure, 123–124, 316, 317. see also High blood pressure (HBP) TAAG (Trial of Activity for Adolescent Girls), 208 Taking action. see Action plans Taking the Initiative (AHA), 607–609, 658 Task Force on Community Preventive Services on evidence-based decision making, 579–584 on obesity, 261 on physical inactivity, 214 on school-based intervention programs, 183 Task Force on Research in Epidemiology and Prevention of Cardiovascular Diseases (NHLBI), 685–687 Tecumseh Study, 22 Therapeutic Lifestyle Changes (TLC) on blood lipid control, 299 blood pressure and, 342–348 diabetes/metabolic syndrome and, 382–383 TLC diet, 165 Thin-cap fibroatheroma, 42 Third Joint Task Force Recommendations (Europe), 597–598 Thrifty gene hypothesis, 367–368 Thrombolytic treatment, 475 TIA (Transient ischemic attack), 91, 92, 96 TIPS (Indian Polycap Study), 476 TLC. See Therapeutic Lifestyle Changes (TLC) TLC diet, 165 Tobacco use. See Smoking, and other tobacco use TOHP (Trials of Hypertension Prevention), 343 tPA antigen, 474 Tracking obesity and, 225, 228 of risk factors in childhood, 24 Trans fatty acids, 176–177. see also Diet, effects of Transient ischemic attack (TIA), 91, 92, 96 Treaties, on global health, 396, 423–424
46394_INDX_697_710.qxd
2/26/10
2:19 PM
Page 709
INDEX
Treatment. see specific cardiovascular diseases Trial of Activity for Adolescent Girls (TAAG), 208 Trials of Hypertension Prevention (TOHP), 343 Triglycerides, classification of, 271. See also Blood lipids Trust for America’s Health, 658, 672–673 T2DM mellitus. see Diabetes and metabolic syndrome Type A behavior pattern, 453–455 development of concept, 453 hostility and anger, 454–455 initial findings, 453 modifiability of, 454 Western Collaborative Group Study results, 453–454 UCLA Health Forecasting Project, 241 Ultra-fast computed tomography (UFCT), 43–44, 54 Ultrasonography aortic aneurysms and, 121 on carotid arteries, 43 PAD assessment and, 114 United States. See also specific organizations and studies health related economic issues in, 13–14, 640 medical care costs (2009), 10 mortality patterns in, 6–7, 12–13, 19–22, 66, 253, 510 socioeconomic status and social change in, 520–522 Unnatural Causes: Is Inequality Making Us Sick? (film), 510 Unstable angina, 61, 64, 452 USPSTF. See Preventive Services Task Force (USPSTF) Västerbotten Intervention Program, 632–633 Venography, 134 Venous thromboembolism (VTE) background of, 132–133 cardiovascular diseases and, 111–113, 132–135 classification of, 112 features of individual cases, 132 mortality patterns and, 133–134 population studies, 133 risk factors of, 134 Ventricular arrhythmias, 135–136 Ventricular fibrillation (VF), 135 Ventricular tachycardia (VT), 135 VERB Campaign, 214–215 Veterans Administration, U.S., 345 The Victoria Declaration on Heart Health, 648 Viral infections, 46, 480 A Vital Investment (WHO), 648 Vitamin B complex, 485–487 Vitamin E, 481–482 VTE. See Venous thromboembolism (VTE)
Waist-hip ratio (WHR) diabetes and, 368, 372 measurement methods, 224, 229, 236, 244, 251 Walking, as exercise, 526–527 Web of causation, 540–541 Weight. see Obesity Weight Watchers Diet, 165, 179 Western Collaborative Group Study, 453–454 Western Diseases (Trowell and Burkitt), 514 Westernization, of population groups, 503, 514 WHA. See World Health Assembly (WHA) WHI. See Women’s Health Initiative (WHI) WHO. See World Health Organization (WHO) WHR. See Waist-hip ratio (WHR) William J. Clinton Foundation, 183 Women, AHA guidelines for, 603. see also Sex Women’s Health Initiative (WHI) on air pollution exposure, 523–524 on aspirin treatment, 475 on HRT, 476–477 risk scores and, 606 Women’s Health Study (WHS), 248, 488 Worcester DVT Study, 132–133 Work, stress and, 455–458 Working Group on Genome Wide Association in NHLBI Cohorts, 155 World Action on Salt & Health, 351 World Bank. See also Disease Control Priorities in Developing Countries Project (World Bank); Global Burden of Disease and Risk Factors Study (World Bank) on government intervention, 562 on mortality rates by region, 10–11, 13–14 on per capita food consumption, 161–162 World Health Assembly (WHA) on alcohol consumption, 444 on coronary heart disease, 301 on diabetes, 385 on dietary recommendations, 184–185 on physical inactivity, 192 on prevention and control, 664–665 on smoking, 423 World Health Organization (WHO). see also MONICA Project (WHO) on alcohol consumption, 445–446 atherosclerosis studies, 48–49 Blood Press Studies in Children, 342 on BMI, 226 on burden of risk, 638 Cardiovascular Disease Prevention and Control, 664–665 Cardiovascular Survey Methods, 63–64 CINDI Program, 660–661
709
Community Prevention and Control of Cardiovascular Diseases, 612 Diet, Nutrition and the Prevention of Chronic Diseases, 184 on estimated lost income, 10 on evidence-based decision making, 579–580 Gaining Health, 648, 661–663 Global Strategy for the Prevention and Control of Noncommunicable Diseases, 688–689 Global Strategy on Diet, Physical Activity and Health, 184–185, 191, 192, 216, 613 Global Tobacco Epidemic, 2008, 396, 415, 424 guidelines and policies, 598, 601–603, 612–613 Hypertension Control, 348–349 International Physical Activity Questionnaire, 202–204 Investing in Health Research and Development, 687–688 Life Course Perspectives on Coronary Heart Disease, Stroke, and Diabetes, 688 map of distribution of DALYs, 15 Multinational Study of Vascular Disease in Diabetics, 381 on obesity, 243–244, 259 Obesity: Preventing and Managing the Global Epidemic, 259–261 on physical inactivity, 216 Preventing Chronic Diseases, 620 Prevention in Childhood and Youth of Adult Cardiovascular Diseases, 301, 342, 423 Prevention of Cardiovascular Diseases, 350 Prevention of Coronary Disease, 612 Prevention of Diabetes Mellitus, 363–364 Primary Prevention of Essential Hypertension, 341, 612–613 research goals, 687–689 on smoking, 414–415 STEPS Stroke System, 97 Study Group Report, 25 Surveillance STEPS program, 683 A Vital Investment, 648 website, 15 World Health Report, 2004, 67–69 World Heart and Stroke Forum, 593 World Heart Federation, 593, 664 Worldwide Efforts to Improve Heart Health, 634–635, 664 Years of life lost (YLL), 14–15, 73–74, 290–291 Youth Media Campaign Longitudinal Survey, 215–216 Youth Risk Behavior Survey (YRBS), 171, 208, 400 Zone Diet, 179