This Page Intentionally Left Blank
Frontispiece The Cross by Marion Patrickc. 1967; oil on hardboard. Reproduced with...
191 downloads
4769 Views
10MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
This Page Intentionally Left Blank
Frontispiece The Cross by Marion Patrickc. 1967; oil on hardboard. Reproduced with the kind permission of Bethlem Royal Hospital Archives and Museum, and the Ridehalgh family on behalf of the late Marion Patrick.
This Page Intentionally Left Blank
BM, BGh FRCP, FRGPsych Dean, Institute of Psychiatry; Denmark Hill, London
~l~ckwell Science
0 1998by
DISTRIBUTORS
Blackwell Science Ltd EditorialOffices: Osney Mead, Oxford0x2 OEL 25 John Street, Londonw C l N ~ B L 23 Ainslie Place,Edinburgh E H 6AJ ~ 350 Main Street,Malden MA 021485018, USA 54 University Street,Carlton Victoria 3053, Australia 10rue Casimir Delavigne 75006 Paris, France Other EditorialOffices: Blackwell Wissenschafts-VerlagGmbH Kurfurstendamm 57 10707 Berlin,Germany Blackwell ScienceJKK MG Kodenmacho Building 7-10 Kodemacho N~ombashi Chuo-ku, Tokyo 104,Japan All rights reserved.No part of this publication may be reproduced, stored ina retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by the UK Copyright, Designs and Patents Act 1988, without the prior permission of the copyright owner.
Marston Book Services Ltd PO Box 269 Abingdon Oxon 0x14 ~ Y N (Orders: Tel: 0x235 465500 Fax: 01235 465555) USA Blackwell Science, Inc. Commerce Place 350 Main Street Malden, MA 02148 5018 (Orders: Tel: 800 759 6102 781 388 8250 Fax: 781 388 8255) Canada Copp ClarkProfessional 200 Adelaide St, West,3rd Floor Toronto, Ontario, M ~ 1w7 H (Orders: Tel: 416 597-1616 800 815-9417 Fax: 416 597-1617) Australia Blackwell SciencePty Ltd 54 University Street Carlton, Victoria 3053 (Orders: Tel: 3 9347 0300 Fax: 3 9347 5001) A catalogue record for this title is available from the British Library
First published 1998
ISBN 0-86542-987-1
Set by Setrite Typesetters,Hong Kong Printed and bound in Great Britain at theUniversity Press,Cambridge
Library of Congress Cataloging-in-publication Data
The Blackwell Science logo is a trade markof Blackwell Science Ltd, registered at the United Kingdom Trade Marks Registry
97-6410
Management of depression / edited by Stuart Checkley. p. cm. Includes bibliographical references and index. ISBN 0-86542-987-1 1. Depression, Mental. I, Checkley, Stuart. RC537.M341997 616.85’270651-dc21 GIP
List of contributors, vii Introduction, x § ~ u a Checkley r~ 1
2
The assessment and epidemiology of affectivedisorder, 1 ~ a~ e~~ b li n g ~ o n Depression presenting at different levels of care,27 David oldb berg i o l o ~ i c models a~ of depression and response to an~depressant t r e a ~ e n t s42 , §~uart C~ec~ley
4 A"psychosocia1 model of depression: i~plicationsfor mana~ement,70 ~ i r r i 1 0 .arris
5 A cognitive theory of depression, 94 ~ o ~ i nLam ic 6 A psychod~amic-interpersonalmodel of depression based on attachment theory, 125 ~ i l ~ i E. a n~ a r d y 7 Efficacy of psychodynamic, interpersonal and experiential t r e a ~ e n t of s depression,142 ~ a vA. i ~~ ~ a ~ i r o The efficacy ofc o g ~ i ~ v e - b e h a ~ i o utherapy, ral 165 E d ~ a r d~ ~ ~ kand i n~ su t~h i l l i a ~ s
188 9 The efficacy of antidepressant medication, ~ a l c oN. l ~Lader 10
The adverse effects of antidepressant medication, 212 ~ aTaylor ~ i ~
11
Efficacy oft r e a ~ e n t for s resistant depression, 234 hil lip J.Cowen
12
The efficacy of electroconvulsive therapy, 252 Allan 1.E Scott
s recurrent affective 13 Lithium and other drugt r e a ~ e n t for disorder, 275 ~ o h ncook so^ 14 The management of depression in general practice, 307 Scott Weich and Anthony ann 15
The management of depression in the c o m m ~ i325 ~, ~eraldine Stra~hdee and ~ ~ a l Bhui d e ~
16 The managementof depression in children and adolescents,345 Eric ~ o ~ b o n n e
37 Depression after childbirth, 364 ~ a u r e e narks and R. Channi~ u ~ a r 18 The management of depression in the elderly, 383 Lucy C. Aitken and RobertC. B a l a ~ i n
in physical illness, 410 19 The management of depression att thew Notopf and SimonWessely 20
vi
The management of resistant depression, 430 Stuart C ~ e c k l ~
LUCY C . A I T E N MRCPsych, Senior Registrar in Old Age Psychiatry, Withington Hospital, Ne11 Lane, Manchester M20 2LR
W IN I DM, FRCP, FRCPsych,Old Age Service, P s y c h i a t ~ Directorate, Central Manchester Healthcare NHS Trust, York House, York Place, Oxford Road, Manchester Mz3 9WL N I NLA,PhD, FRCP, FRCPsych,Professor of Social and C o ~ ~ u n i t y P s y c and h i a Behavioural t~ Sciences, University College London, Archway Wing, WhittingtonHospital, Highgate Hill, London Nz9 5NF
H U I MSc, MRCPsych, Wellco~eResearch Fellow, ~nstituteof Psychiat~, De Crespigny Park, D e n ~ a r Hill, k London SE5 8AZ
L E Y BM, BCh, FRCP, FRCPsych, Dean and Professor of Psychoneuroendocrinolo~, ~ n s t i t ~oft ePsychiatry, De Crespigny Park, D e n ~ a r Hill, k London SE5 8AF 0N BM, DPhil, FRCP, FRCPsych, Consultantand Honorary Senior Lect~rerin Psychiat~, Royal London Hospital (St C l e ~ e n t s2A ~ , BowRoad, London E3 4LL
E N MD, FRCPsych,Professor
ofP s y c h o ~ h a r ~ a c o l o ~ ,
Psychophar~acolo~ ~esearch Unit, Warn~ord Hos~ital, Headington, Oxford OX3 7JX
ERIC F
N N E MD,MRCPysch, Readerin Epide~iologicalChild P s y c h i a t ~ and H o n o r a ~ Consultant, D ~ a r t ~ eofn Child t and Adolescent Psychiatry,~nstituteof Psychiatry, De C r e s p i ~ Park, y D e n ~ a rHill, k London SE5 8AF L P)B E R G DM, FRCP, FRCPsych,Professor of Psychiatry,~nstituteof Psychiatry, De C r e s p i ~ Park, y D e n ~ a r Hill, k London SE5 8AF
vii
Y MSc, Clinical Research Fellow, Psychological Therapies ~esearchCentre, ~niversityof Leeds, 17 Blenhei~Terrace, Leeds LS2 9/11
A,Senior ResearchFellow, Royal Holloway, ~niversity of London, 11 Bedford Square, London WCxD 3 M
I-$ F B§c, MBB§ MRCPsych, MSc,Clinical Research Fellow
and Lecturer in Psychiat~, Depart~ent of Ps~choZog~cal Medicine, ~nstituteof Psychiat~, De Crespigny Park, D e n ~ a rHill, k London SE5 8AZ MD, PhD, FRCPsych,Professor of Perinatal P s y c h i ~ t ~ , D ~ a r t ~ eofPerinata1 nt Psychiatry,~nstituteof Psychiat~, De Cres~igny Park, D e n ~ a r k Hill, London SE5 8AF DSc, PhD, MD, FRCPsych, Professor of Clinical P s y c ~ p ~ a r ~ u c o ~ o ~ of , P~s yncshti~~ ttDe ~ ,tCrespigny e Park, D e n ~ a r Hill, k London SE5 8AF PhD, Senior Lecturer in Clinical psycho lo^, D ~ a r t ~ eofn t
P s y c ~ o Z o ~ , ~ n of s t Psyc~ziat~, it~te De Crespigny Park, D e n ~ a rHill, k London SE5 8AF MD, FRCP, FRCPsych, of Professor
A
Epide~iological
Psychiat~, ~nstituteof Psychiat~, De C r e s p i ~ Park, y D e n ~ a rHill, k London SE5 8AF
S
Dphil, Senior Lecturer,D ~ a r t ~ eofn Perinata~ t Psych~at~,
~nstituteof Psychiat~/ De Crespigny Park, Den~zarkHill, ond don SE5 8AF
T T BSc, MPhil, MD, MBA, MRCPsych,Consultunt Psychiatrist,Andrew Duncan Clinic, Royal ~ d i n ~ u rHospital, gh Morningside Terrace, E d i n ~ ~ r EHzo g h 5HF BA, MSc, PhD,FBPsS, Professor of CZinicaZ psycho lo^
and Director,Psychological Therapies Research Centre, ~niversityof Leeds, 17 Blenhei~ Terrace, Leeds, LS2 9JT
E
MBBCh, MRCPsych,Head of Service
Develop~ent, Sains~ Centre u ~ , sfor Mental Health, 134-138 Borough HighStreet, ond don SE1 1LB
AVID TAYLO ,BSc MSc, MRPharms, Chief Phar~acist,
Maudsl~ Hos~ital,
D e n ~ a r Hill, k London SE5 8AZ
S
MSc, CPsychol,~nst~tute of Psychiut~,De C r e s ~ i ~ y
Park, D e n ~ a rHill, k London SE5 8AF
viii
T MRCPsych, MSc, Senior Lecturer in P s ~ c h i a t A ~ ,c a d e ~ i c ~ e p a y t ~ofeP~st ~ c ~ i Royal a t ~ ,Free Hospita~~ e d i c School, a ~ ~~~~n~ Hill Street, ~ o n ~ o ~ ~~3
2PF
,MA, BM, BCh, MSc,MD,FRCP, RCPsych, Professor of E p ~ ~ e ~ i o l o gand i c a~l i ~ i s o n ~ s King’s ~ c h i College a t ~ , School of ~ e d i c i nand e ~nstituteof P s ~ c ~ i De at~ Crespigny , Park, ~ e n ~ aHill, r k ~ o n ~SE5 o n 8AF MA, CPsychol,Senior Lecturer irr Clinical ~ s y c h o l o ~ , ~nstituteo ~ P s y c ~ i aDe t ~Crespign~ , Park, ~ e n ~ aHill, r k ~ o n d SE5 ~ n 8AF
ix
The purpose of this book istobr together from different disc what is knownaboutthemanage of depression. A.greatdeal h learnt duringthe last 30 years concerning the effectiveness of the main physical and psychological treatments for depression. Important advances have also been made in understanding the biological and psychosocial causesof depression and themecha~sms by which physical and psychological treatments work. Yet most of these advances have been made in isolation from each other. There has been some integration between the psychological and social studiesof depression but little between the psychosocial and biological. Most casesof depression in the communi^ can be treated satisfactorily either with psychological or with physical treatment,little yetis known about interactions between the two, and in everyday practice most clinicians operate within an exclusively biological or psychosocial framework. The intention of this book is to bring these different approaches together whenever possible. The first section of the book deals with epidemiology and aetiological t depression. For a much models in so far as they relate to the t r e a ~ e n of fuller account of aetiology the reader is referred toPaykel(~992)~ and more specialist reviews (Bloom 8r Kupfer 1995)." In order to discuss the mana ment of depression in the different levels of carein ~ ~it ~resents, c h it is first necessary to have an ~ d e r s t a n d i nof~the epidemiolo~of this extremely c o ~ o disorder n (Chapter l), and second to erst stand how depression presents at differentlevels of care and, also, which filters re of care (e.g. primary or secondary) to another ment of patients from one level (Chapter 2). The next four chapters summarize, froma clinical perspective, the biological, social and psycholo~ical models of depression. Genetic studies have defined both the environmental and enetic contribution as lifetime lPaykel, E.S.(1992)~ ~ ~ ~ ~ O o ~~~ os o~r ~A 2nd e r~s ,edn. e c fChurchill ~ ~ e Livingstone, Edinburgh. 2Bloorn,F.E. & Kupfer, D.J. (1995)P s ~ c h o p ~ r ~the ~ Fourth c o ~G o ~e ~~ e ~o~P~ogress. ~ f ~ oRaven ~
Press, New York. X
vulnerability toall forms of depression, and have clearly shown the different balance between the influence of nature (genetics) and nurture (enviroment) in different forms of depression (unipolar and bipolar). The psychosocial model describedin Chapter 4 accounts for most of the variance in onset in the majority of episodes of unipolar depression, and for much of the variance in rates of recovery. The model is extremely well cross-referenced throughout the book and clearly provides one unifying theme. Anotheris provided by the cognitive model of depression (Chapter 5), which seeks to explain the mechanisms of action of cognitive therapy but also provides a good understanding of the psychological mediators between the effects of distal and proxi4). mal social processes on the onset of depression (Chapter Another unifying theme is attachment (Chapter6), which, as well as being central to any understanding of interpersonal psychotherapy, is also implicated in mechanisms of social support. Few links currently exist between these psychosocial models and the biology of depression, but a neuroendocrine mechanism which might 3. Biological models underlie the psychosocial model is described in Chapter come into their own in understanding the mechanisms of action of currently established antidepressant treatments and in suggesting new forms of 3). treatment for the future (Chapter The second sectionof the book reviews whatis known of the wanted and unwanted effectsof the main antidepressant treatments. Few, if any, of these treatments produce a change in depression ratings whichis greater than the standard deviationof the baseline score. For this reason it is necessary tobe certain that all treatments are given in an optimal manner: where possible each chapterin this section indicates what is thoughtbeto the optimal useof the treatment in question. Available information on the unwanted as as well the wanted effects of antidepressant treatmentis reviewed particularly in the case of antidepressant medication (ChapterXI), for which most prescribing decisions withinany class of drug are made on the basis of unwanted rather than wanted drug effects. The final sectionof the book describes the management of depression in the very different settings and age groups atit presents. which Although there are some striking and unexplained differences (e.g., the that fact antidepressant drugs do not appear tobe effective in childhood depression), the similarities between these final chapters are much greater than the differences: all adopt a multidisciplinary model; most refer to the psychosocial model mentioned in Chapter 4; most use cognitive models and/or cognitive therapy; and yetall have expertisein the appropriate useof physical treatments. It is a pleasure to acknowledge the timely and excellent chapters which my colleagues have written for this book. Manyof us have collaborated in clinical and academic matters over the years and it is hoped that this willbe an advantage rather than a disadvantage to the overall of theaim book; which
xi
is to integrate different approaches to the treatmentof depression. I would also like to acknowledge particularly con~ibution the made by who provided editorial assistance in the preparationof the m a n ~ s c r i ~ t s . Stuart Checkley
xii
The aimof this chapter is to review the epidemiology of affective disorder in the lightof constra~tswhich arise from problems of measurement. The central problem is that epidemiology is essentially a medical discipline and therefore exploits the conceptof disease. Diseases are primarily categories. While, for many diseases, what starts off as a categorical concept ends by revealing underlying continuities, the continuous distribution of depressive symptoms in the general populationis obvious from the outset. There is little evidence a disease-based approach must therefore be offset by for valid cut-points, and a robust scepticism: the categories belovedof the epidemiologist shouldbe treated as particularly tentative fictions. Such fictions may assist the progress of knowledge, butif we are to rely on them, they must be used consistently.
The purposeof any classification is to aidc o ~ u n i c a t i o nIn . psychiatry we use mental phenomena (and sometimes physical correlates) as the basis of classification sothat diagnostic classes and also series of exemplars of a given class canbe compared. This is done in order to test ideas relating to the cause, outcome and treatmentof the disorders that are represented by the class. Diagnostic classes depend on two sets of definition: that relating to the class itself and that relating to the elements making up the criteria for class reco~ition.In psychiatry the first typeof definition relates to thes ~ ~ ~ ~ o ~ e s that form the basis of disorders, while the second relates to cli~ical p ~ e ~ o ~ e ~ ~ , mainly symptoms. Symptoms are used according to given rules to construct syndromes. In recent years these rules have been set out explicitly in the Diagnostic Criteria for Research (DCR) attached to specific classi~cationssuch (APA), 1987), DSM-IV(APA, -111-R (American Psychiatric Association 1994) and ICD-10, so precisely that it is possible to incorporate them into computer algorithms such as GATEGO (Wing et al., 1990) and OPCR~T et al., 1991). (McGuff~ Once the presence of symptoms has been established, the information can be entered into the computer program in order to provide a diagnostic 3
classification. Human idiosyncrasy can be reduced to an absolute minimum in this process. However, researchers must then decide how carefully the underlying symptoms should be identified. The choices include unstructured clinical assessment, responses to questionnaires, ands e ~ i - s ~ c t u r research ed interviews; each hasits own problems. x U n s ~ c t u r e dclinical judgement introduces variability into the processof case allocation, since researchers are relying merely on their devotion to a common educational tradition.The identification of clinical symptoms may consequentlybe biased by prior prejudicesin favour of a particular diagnosis. This situation is even worse when the judgements of others (for example the treating physician) areused, as with the diagnostic information recorded in case registersor in national statistics. 2 In order tobe practicable, questionnaires should comprise simple responses to unelaborated questions. However, symptoms are traditionally recognized through an assessment of mental experiences that demand quite elaborate enquiry They are usually established by a processof clinicalcross-exa~ation. This process is rather complicated since it requires the questioner to frame further questions in a flexible wayin the light of the answers given by the subject. Whileit might be possible to ncapsulatethis procedure in a branching algorithm, it would be exhaustive and exhausting. It might require paths comprising over a dozen questions just to establish the presence of depressed mood. In these circumstances there are clearly limits to the process of standardization, and we are probably better off relying on the shortcuts available from using the skills of trained clinicians. Since diagnosis is built around symptoms defined and elicited in this manner, redefinitionin terms of answers to much more limited questions would involve changing the concept of diagnosis itself. No-one has seriously suggested that the conceptualization of psychiatric symptoms should be changed; thereforeif a questionnaire is used, phenomena may be recorded as present when subsequent clinical enquiry might reveal otherwise, and vice versa. Nevertheless, structured questionnaires allow lay interviewers to be used, with considerable cost savings. 3 Semi-structured research interviews are costly in clinical time, and the way in which symptoms are established makes it impossible to standardize the 1995).Becauseof the reliance on clinical judgement procedure entirely (Robins, and the effectthis has on the choiceof follow-up questions, some variability will remain. This is the price paid for greater validity, that is, the closer approximation to the clinical consensus about the nature of given symptoms. Classifications in psychiatry were originally developed on basis the of an informal process of abstraction, involving the exercise of considerable intellectual effort; revision of classificatory schematain the earlydays relied almost wholly onclinical reflection,However, since the classifications are set 2
up primarily for scientific purposes, they should be properly modified in the light of empirical research whichpermits definitive statements about their utility, and the st~dardized and operationalized classifications which are now available offer an opportunity for using research in this way. Unfortunately, much of the pressure for change has continued to originate from clinical and political demands. Revisions have sometimes had the appearanceof tinkering in order tocapture some imagined essence of the included disorders (Birley, 1990). What looks like fine-tuning can nevertheless make considerable differences to whether disorders meet criteria or not, and thus affects prevalence data disproportionately. Classifications should only be jettisoned rounds of ~ a d e ~ u ascientific te utility and as seldom as possible, since rapid too revision defeats objective the of comparison. ons arecreatedby c o ~ i t t e e sand , the natural tendencytohorse-traeen experts selected precisely because they are powerfuland opinionated leads to an overelaborate structure,an excess of allowable classesand subclasses, and complicated defining criteria. All these tendencies are apparent in the classificationof affective disorders ealth Organization (WHO), 1992).Thus in D 111-R there are potentially 14 categories to consider before allocating som ssed mood,and in ICD-lo (WHO, 1992) there are 22 categories. It r’s view that greater utility would accrue from limiting the primary categories to three (bipolar disorder, unipolar depressive psychosis and unipolar non-psychotic depression),and epidemiological research often uses re clearly the major accepted classi~cations.What are their implications for case definition and identification? For the purpose of illustration, Table 1.1 compares the criteria relating to major depressive DD) and depressive episode ( .It will be seen that, although there is a welcome convergence between and ICD, the categorization is far from identical. This has serious implications for of affective studies disorder, particularly those seeking to establish prevalence and cornorbidity. The categories are far enough apart to cause discrepancies in identification, but probably too close together for empirical studies to establish their relative and several ICD categories lack existence of such rag-bag categories takes away from the comparability that for. the S stem as a whole strives This is reflected e disorderswith explicit DC are the more severe ones. (ECA) rted prevalences. Thus inEpidemiologicCatchmentArea eties that the instrument used (the Diagnostic Interview erestimated prevalence:the prevalence of MDD was only x. Regier, 1991). In its mild form, the ICD category DE has a slightly lower threshold than MDD, but the 1-month
3
DSM-I~-~/~SM-~
ICD-10
Symptoms present nearly every dayin the same 2-week period
Episode must have lasted at least 2 weeks with symptomsnearly every day
Change from normal f ~ c t i o n i n g
Change from noma1 functioning
Key s y ~ p (n~= 2) ~ s Depressed mood Anhedonia
Key s y r n p ~ o(n ~ s= 3) Depressed mood Anhedonia Fatigue/loss of energy
Ancilla~ syrnpto~sfn= 7)
Ancilla~ s y ~ p t o f~n = s 7)
Fatigue/loss of energy ~eight/appetiteloss/gain ~omnia/hyperso~a Observed agitation/r~ardation Low s e ~ f - e s t e e m / ~ u ~ ~ t Impaired t h l n k ~ g / c o ~ c e n t r a t i o ~ Suicidal thoughts
Weight and appetitechange Sleep disturbance Subjective or objective agita~on/retardation Low self-esteem/co~idence Self reproach/guilt Impaired thinking/concentration Suicidal thoughts
Criteria: 2 key, 5 s y r n p t ~ in ~ stotal
nt distress airmen%
Criteria:
Mild episode: 2 key, 4 symptoms in total Moderate episode: 2 key, 6 symptoms in total Severe episode: 3 key 8 symptoms in total
~xcl~sions
~xcl~sio~s
Not mixed episode
No history (ever) of manic symptoms Not substance related Not organic
Not organic Not bereavement Not psychotic
all forms of depressive episode rep ational S w e y of Psychiatric te’ and ’severe’ episodes are cor ity surveys have used the 9th ), which, irt so far as it approx ith disorders of somewhat l lying between the twot ~ e s h o l d would s fall into other gories. These include eit fluc~ating,o epression, which lacks DC the ICD s u g ~ e s t i o and ~ co~structedour own criteria for this of a failure to meet criteria for 0th 12 or more on the revisedClinical Intervie he disorder had a natio rably lower than that ~ l i n is ~~ l ~ ~ of idepressive es disorder meet criteria for DD or DE. However,in the case of c o ~and ~ ~r ~ i i ies, we are faced with a serious dilemma. It is clear theclassificator~
4
Prevalence ofCATEGO depressive classes(YO). Site
Female
Group 1
Canberra (Hendersonet al., 1979)" Camberwell (Bebbingtonet al., 1981)" Edinburgh (Surteeset al., 1983) Nijmegen (Hodiamontet al., 1987)" Finland (Lehtinenel al., 1990)t Cambewell (Bebbington & Marsden, 199%)"
Group 2
Athens (Mavreaset al., 1986) Santander (Vazquez-Barqueroet al., 1987)* Cambewell (Cypriot) (Mavreas& Bebbington, 1987) Sardinia (Cartaet al., 1991)
Group 3
Uganda (Orley& Wing, 1979) Palembang (Bahar, 1989) Dubai (Ghubashet al., 1992)
Male
n 756 800 576 3232 742 760
(157) (310) (486)
2.6 4.8
-
-
Total 6.7 9.0 6.9 -
4.8 7.0
-
5.5 4.6 5.4
(408)
2.4 5.7
6.5 5.2
489 1223 (452) 307
4.3 4.5 4.2
10.1
7.8 7.1
7.4 6.2 5.6
5.2
11.0
8.3
17.0
21.0
18.9 6.5
374 206 839 (100) 300
-
-
13.7
*Two-stage survey-numbers givenPSE at second stage in brackets. Prevalences weighted to represent original population. 1.Age adjusted figures. Studies that do not give a breakdown in terms of CATEGO classes are not included.
view to providing standard reporting of data. The range of prevalences given in Table 1.4 is difficult to explain, as it does not correspond to the obvious cultural differe~cesbetween the locations of the surveys. There are, for example, very high rates in Europe compared with the USA, but the studies from Canada and New Zealand are also high. High in rates eirut are perhaps un~erstan~able, and have similarities to the PSE data from another hightrauma site, the Uganda of the mid-zg[iros(Orley in the overall interpretation of these results suggest differenc~s inthe application of the interview. Since these surveys were undertaken, data have been published from two 1 probabili~samples, the American -scale i n v e s t i g a ~ based o ~ on na onalComorbiditySurvey (NC ssler et al., 1993,1994a,b)andthe ritish ~ a t i o n aSurvey l of Psychiat idity (Meltzer et al., 199 et al., 1997b). They involved interviews with 8000 and 10'000 Table 1.5).Using the Composite Intern~tional ins et al., 1988) in its U ~ v e r sofi ~ ivari c ~ and colleagues (1993) obtained a noticeably high prevalence for ~ S M - 1 1 1 - ~ inst~ment -the NCS is oneof the first based on CIDI, and the of this use interviewers may result i ower thresholds than its predecessors. The et- al., ~ 9 9also ~used ) lay tional Survey of Psychiatric ltzer interviewers. It was based on the revised versionof the CIS-R (Lewis et al.,
C ~ ~ ~I : ~t fef er c t i ~~ei s o r ~ e r Table 1.3 Six-month prevalence of major depressive episode(%). Site
Total
Female
Male
n
ECA Studies
Baltimore (Myerset al., 1984) New Haven (Myers et al., 1984) St Louis(Myers et al., 1984) Piedmont (Blazer et al., 1985)* Los Angeles(Burnham et al., 1987)
3481 3058 3004 3921 3125
2.21.3 3.52.2 3.21.7
3.0 4.6 4.5
1551 3258
3.02.4 3.22.5
3.3 3.9
5004 3005 2995 5.33.4 1498 1366 (501) -
-
-
_ .
-
1.7 3.1
_ .
Other sites
Puerto Rico (Canino, et al., 1987) Edmonton (Bland et al., 1988) Taiwan (Hwu et al., 1989)" Taipei Small town Rural Christchurch (Oakley-Browne et al., 1989) Munich (Wittchenet al., 1992)$
_ .
'0.6 1.1 0.8
7.1
-
3.0
*Weighted to take accountof over-sampling of elderly.
t One-year prevalence. $Two-stage survey-numbers given DIS at second stagein brackets. Prevalences weighted to represent original population.
Table 1.4 Recent
surveys based on national probability samples.
Prevalence Study National Comorbidity Study (Kessler et al., 1994a,b)
n 8098 48
108 10 National Survey of Psychiatric Morbidity (Jenkins et al., 199%)
Location contiguous United States Great Britain 18-64 episode (except Highlands and Islands)
Age range
Period
Disorder FemaleMale
15-54
1 year
12.9 Major 7.7 depressive disorder
1 week Depressive 1.8
2.7
1gg2), a lay-administered interview that provides ICD-IO diagnoses (WHO, 1992). It can be seen from Table 1.4that despite the similar procedureof the CID1 and theDIS, the prevalencesin the ECA studies are closer to those in the British National Survey of Psychiatric Morbidity. These discrepancies require direct comparison of the i n s ~ m e n t in s a single sample before any idea of their clinical significance can be derived. Such study a is currently underway (TS. Brugha, H.M. Sharp, S.A. Cooper et aE., unpublished i~ormation), Depressive disorder has an ill-defined threshold and usually does not result in referral to specialist services, so there are particular difficulties in interpreting data based on referred series of patients (Bebbington, 1988). Nevertheless, provided we are cautious, such studies remain of value, especially if they corroborate and amplify community surveys.
le 1.5 Incidence of depressive disorder based on rates of first admission to hospital. Incidence perlo5
rydate
and
Place Study Depressive ~ s y c ~ o s etc. is, Shepherd (1957)
Norris (1959) Jaco (1960) Sadoun et al. (1978) Joenson & Wang (1983) DHSS (1985) Finlay Jones (pers.c o r n . , 1987) Wells (pers. comm., 1987)
52.3
-
Buck~ghamshire Non-nemo tic 1945-7 affective disorder London 1947-9 Manic depressive psychosis Texas 1951-2 Affective psychosis France 1975 Manic depressive reaction Denmark 1970-9 Manic depressive psychosis England Affective 1978-82 psychosis Depressive Western Australia psychosis 1977-83 New Zealand Depressive 1980-3 psychosis
All ~ ~ r e s s i o n s Joenson & Wang (1983) Denmark 1970-9 DHSS (1985) England 1978-82 Western Australia Finlay Jones (pers.c o r n . , 1977-83 1987) Wells (pers. comm.,27.9 1987) 1980-3 ZealandNew OHare & Walsh (1985) Ireland
6.5
12.6
14.1
25.9
13.0
22.0
11.0
17.0
-
-
6.0
10.0
11.6
15.5
9.8
14.6
-
-
32.0 28.9
54.0 39.2 47.2 87.9
Data on first admission rates from all over the developed world are presented in Table 1.5. It is, of course, dangerous to expect at the outset that very much weight can be placed on such data, to due the vagaries of di practice and admission policies. W i l e both of these influences have over the years, thisis particularly true of admission policies becauseof the recent acceleration in the drive to manage psychiatric disorders in the c o ~ uand~ the ~ resulting , reduction inbed a v a ~ a b i l iFor ~ . this reason I mited data to a 40-year period endingin the mid-1980s. During this it is like1 that most severe affective disorders led to admission to one would expect more variation in the practice regarding everi it^ appears tobe the main determinant of admission, overriding social influences (Hurry et al., 1980). The resultsin Table 1.5are of some interest. The admission rates for severe es of affective disorder (obviously loosely ~ e f i n e fall ~ ) into a relatively restricted range: from 6 to 14 per 1 0 5 for men and from 10 to 26 per 1 0 5 for l ~ rates for all women, The sex ratio ranges from 1.3 to 1.9. ~ u r p r i s i n gthe depressive disorders are also fairly consistent, from 5.2. 28 per to 1 0 5 for males
and from 39 to88 per 1 0 5 for females. The implicationis that the thresholds for admission to hospital were rather similar in the different locations quoted, otherwise the ratio of severe depressions to all depressions would have varied more. It is just possible, therefore, that the differences in the figures reflect real differences in susceptibility in the local populations, at least in some cases, for example the comparison between London and Buckinghamshire. It is also possible that the values for severe depression represent ballpark figures for the true incidence of the condition. However, before concluding this, we should see whether furtheri l l u ~ n a t i o n can be obtained from case register studies. These provide data on incidence based on all contacts with services (Table 1.6). They avoid duplicate counts and probably use categories that are more consistently applied than routinely collected national data (Wing, 1972). They are restricted to relatively small geographical areas, but at least the social characteristics of the areas are known in detail. It might be hoped that register data would reveala greater consistency especially for severe depressions, as they would virtually all be referred for specialist treatment. The inconsistencies shown in Table 1.6 probably do reflect considerable variation in the register definitions of disorder. It is suspected from other sources that the categories of depressive psychosis in Salfordofand manic depressive psychosis in Aarhus are broader than equivalent usages in the other sites. The differences between London and Aberdeen are likely to reflect a real difference in incidence between an inner city area anda predominantly rural one. Nevertheless, the values for severe depressions are around twice those obtained if first admission is used as a proxy for inci ence, suggesting that even in this period only around half of severe depressions resulted in admission. It may perhapsbe concluded that the incidence of severe depression is three or four times that of schi%op~enia. There are special difficulties in studying the epidemiology of bipolar affective disorder, centred around the fact that bipolarity can only be recognized at an unspecified time after the onset of first affective symptoms. I have discussed these at length elsewhere (Bebbin~tonSr Ramana, 1995). Moreover, surveys based on lay assessment are likely to over-recognize mania. For these reasons, community surveys provide poor data on prevalence; on balance the values for bipolar disorder in many studies are probably overestimates. Table 1.7 lists prevalence both for manic episode and for bipolar disorder from available studies.
The prevalenceof depressive disorders is not guaranteed be set to for all time. In par~cular,several authors have raised the possibility that the frequency is increasing as the century proceeds (Klerman, 1978,1988; Schwab et al., 1979).
le 1.7 Six-month prevalence of mania and bipolar disorder based on community psychiatric surveys. Prevalence (%) Study
Location
Faravelli et al. (1990)
Florence
Bipolar Manic episode
I and I
I1
1.7t
-
0.9 1.4 0.6
0.2
1.0 0.5 1.o 0.3 0.4
0.6
0.7
1.0
ECA studies (Robins & Regier, 1991; Myers et al., 1984; Bumham et. al., 1987)
New Haven1.2 Baltimore St Louis Durham NC Los Angeles0.7
Total
0.8 0.4 0.7
Other DIS studies Canino et al. (1987) Bland et al. (1988) Hwu et al. (1989) Oakley-Brown et al. (1989) Wittchen et al. (1992)
Puerto Rico Edmonton Taiwan Christchurch Munich
US National C o ~ o r b i d i ~ ~ Survey Kessler et al. (1994a,b)
United States 1.3' 1.3" 1.4"
0.3 0.1 0.19 0.1 0.2
*One-year prevalence. *
1- Estimatedfrompublisheddata.
A distinction must be made between age, ~ e r and ~ ocohort ~ effects, cohorts being groups defined by having birth dates occurring within a stipulated period. Subjects maybe exposed to risk of disorder because they are passing be apparent in all cohorts that h an age of risk( a n age effect). This will have passed that age. They may be exposed to disorder because they are passing througha timeofrisk. In this instance, different cohorts will suffer the disorder at different ages, corresponding to their age at the relevant date (a period effect). Finally, cohorts may exhibit differing lifetime risk because they are just different groupsof people-the differences canbe related neither to a cohort efect. It is very difficult given age nor to a given date, and there is a true to separate age, period and cohort effects statistically, because fixing any two of these factors constrains the third;it usually requiresassump~onsthat go beyond what the data really allow ( ~ i c ~ r a m a reta ~ al.,e 1989). An excellent account of the potentialpitfalls in studiesof this type is provided by Klerman ~ e i s s m a n(1989). There are a number of types of study that might reveal changing incidence and prevalenceof disorder. The crudest indicator wouldbe the finding that the mean rates from surveys within a geographical region were changing over time. However, little weight be can placed on such studies because of variation
show an increase in lifetime prevalence in older subjects, as they have had a longer lifein which to become depressed. In fact, there is a consistent ~ e c l i in ~ elifetime prevalence with age in a number of studies, including our own(Bebb~gtonet al., 1989). This su~gests a cohort effectof increasing, and possibly earlier, incidence of depression in of this possibility appears subjects with later birth dates. The direct examination erman (1988) and, more recently, Wolk and Weissman (1995) have reviewed the available databases to illustrate that the relationship between age and incidence of depression has become steeper in sequential birth cohorts.In other words,as the twentieth century proceeds, the population is becoming increasingly prone to depression, and this emerges at a progressively younger age. It is of interest that very similar patterns are & Regier, 1991). observed for bipolar disorder (Robins These illustrations are persuasive, but depend crucially on the subject’s ability to remember past episodes. These are likely to be more remote and thus easier to forget in older subjects, and this may account for the findings described by Klerman (1988). Estimatesof lifetime prevalencedo seem tobe of dou~tfulvalidity (Brometet al., 1986). Using a simulationof the effectsof defective recall on the creationof apparent cohort effects, Giuffra and Risch (1994) have recently shown that really quite small (but constant) rates of t h g produce curves strikingly similar to those that wouldbe obtained uine cohort effect existed without recall effects. m a The ~ i t u dof e recall effect required to mimic a true cohort effectinisline with whatis known of test-retest studies of lifetime recall of depressive disorders. 8) also raised thepossibility that older subjects areless able to recognize past disturbances in their f ~ c t i o n i n gas psychologicalin nature-this would also be a cohort effect, but of a different type. Respite this, the possibility of increasing proneness to depression in sequential birth cohorts clearly merits very serious attention. If the postulated cohort effects are genuine, they arevery interesting. It would be difficult to argue that they arise from changes in gene frequency over so short a time, and the explanation must therefore be environmental, and probably social.
ssi One of the key objectivesof epidemiology is the identification of sociodemographically defined high-risk groups. Although epidemiological study can, at best, only suggest the possible causal directionof observed associations, the results provide useful pointers for further research.
ocial class, usually expressed as occupational class, is a measure of access to and resources, and ought therefore to be inversely associated ty to depression. Twenty years ago, there were greathopes that the social originsof depression would be substantiated through studies of social class. There were certainly some grounds for these hopes (e.g. 973; ~ l e ~ uet tal.h Surtees et al., 1983). community surveys have reported a moderate predominance of minor affective disorder in the lower social classes, although the measurement of social class is far from consistent and the results are oftend i c h o t o ~ ~ in e da way to maximize the effect. Some large and well-conducted studies have to show much effectof social class. The ECA surveys virtually no relationship betweeen affective disorder and social class( egier, 1991),and the National Surveyof Psychiatric orbidity found on1 a reduced rate in Class I adults, with the prevalence of depression bei essentially ~ i f in the o r~e m a ~ classes g ~ e n ~etnal.,s 1997-b). The National om orb id it^ Survey, onthe other hand, did find significant associations with educational standing ( number of other studies have failed ession (e.g. Taylor Chave, 1964; Hare & Sha 1975; Brownet al., 1977; We 1981), nevertheless, there is still probably an overall consensus that the prevalence of disorder is indeed increased in those of lower social class, albeit much less so than some have claimed. This association seems likely to arise irect effect of the disadvantages faced by those in the lower sses, rather than any downward drift exhibited by people afflicted by depression. The associationwith social class is even less apparent for cases in contact ychiatric services( i r t c ~ e l l1971; , Bagley, 1973; Der& ngton, 1988). Thoseof lower social class may be lesstoable access treatme~tand services, and severe disorders may, in any case, be less influenced by socialadversity h the study of urry et al. (1980),severity was the major determinant of referral, not social class, and this would ten s u p ~ o rthe t secondintetation. Thus the belief that socialclassreflects influences that may be c lly related to the origins of depressive diso has never been convincingly demonstrated. een considerableargument over the social classdistribution of Ramana, 1995). Clinical stud r no associationwith class or a tendency to clusterin data, based on large community samples, suggest, on the contrary, that *
there is a consistent but small association with lower socialclass (Robins Regier, 1991).It is unclear how any overidentificatio~of bipolar cases would affect this finding.
The data already presented emphasize that sexofistheone majord e t e r ~ a n t s of frequency of depressive disorder. The female preponderance is extremely robust, and very few studies have failed to reflect it. Its causes are almost et al. 1997a,b), but are i~um~ated certainly multiple and complex (~ebbington by the identificationof high-risk groups definedin other ways.
The most straightforward influencein the domestic arena is thatof marital termination, for whatever reason. Early reports on marital status and depression relied on admission data ( degaard, 1953; Jaco, 1960), or case register information (~uel-Nielsene f al., 1961; Adelstein e f al., 1968; Der ebbington, ~ 9 8 7 although )~ these h e nowbeensupplementedby communitysurveys.The details vary ( bington, 1988), but the general finding of high rates in postmarital groups is robust. There are more women in postmarital groups because of an excess of widows and the longevity of divorced and separated women in relation to their ex-partners. Thus, postmarital status might account for some of the female excess of depression. Aseltine and Kessler (1993) have urged caution over findings from crosssectional studies, emphasizing the problem of selection effects in studyin the impact of marital breakdown on depression. They circumven following upa general population sample for 3 years. Maritaldisr responsible foran increase in scores on a standard depression screening scale of one-third of a standard deviation.This was worse in women, but was less apparent where there had been long-term problems in the marriage. However, in many studies women seem to cope better than men with the end of marriage (e.g.heshensel e f al., 1981). Moreover, age hasa bearing on this issue. Depression in young women who are separated or divorced is compounded by the fact that they are often involved in childcare in disadvantageous circumstances. The high rates of depressionin single mothers is confirmed in the recent British National Surveyof Psychiatric Morbidity (Meltzer et al., 1995). Gamer and his colleagues (1987) have pointed out that the increasing proportion of single mothersin western populations may have explosive consequences for mental health services (not to mention for the mothers themselves). Some authors (Weissman & Klerman, 1977; Paykel & Rowan, 1979; 1
~ e i s s m a net al., 1984) consider that being married connotes more stress for women than for men. Although not a universal finding (e.g. Ensel, 1982; Gebhardt & Nimitz, 1986), there is certainly good evidencethat it is much less protective for women than for men (Bebbingtonet al., 1981; Bebbington, ebbington, 1987; Bebbington & Tansella, 1989). Thus in one survey, single and divorced women had lower rates of minor affective disorder than their male counterparts, while wives had over five timesthe rate of husbands (Bebbington et al., 1981). Manystudies have fo married women looking after small children are particularly 1971; Richman, 1974,1977; Grad de Alarcon et al., 19 Brown & Harris, 1978; Bebbingtonet al., 1981,1984; However, maritalstatus has different associationswith affective disorder in different cultures. For instance, married women are at low riskof disorder diterranean countries (Mavreas et al.,1986; Vazquez- arquero et al.,~ $ 7 ) ~ in rural New Zealand(Romans- larkso on et al., 1988)and in British orthodox Jews (Lowenthalet al., 1995). These societies all accord a high value to the home-ma~n role. ~ This suggests not only that social variables are important in determinin~the sex ratio for depression, but that the association with relatively simple sociodemographic factors may itself be affected by more subtle sociocultural influences.
There seems little doubt that becoming unemployed increases the risk of affective disorder (e.g. Finlay-Jones & Eckhardt, 1981; Kessleret al., 1987; Eales, ( ~ aminor r r1984). , eltzer et al., 1995)~ although most resulting cases are A longitudinal study of young people demonstrated that unemployment has ill-effects on mental health which are reversed when work is obtained (Banks Jackson, 1982; Jackson et al., 1982). Jenkins and colleagues (1982) found similar effectswhen employees were threatened with redundancy, effectsthat were neutralizedif the threat was removed. The benefitsof employment are complex. Krause and Geyer-Pestello (1985) distinguishe~four aspectsof work: job satisfaction; whether work is full- or part-time; pay satisfaction; and commitment tothe work role. Surprisingly, they found that only the last two were related to depressive symptoms. any workers in the field of unemployment research have been influenced by Jahoda’s (1982) emphasis on the ’latent functions’ of employment. She lists six such functions: time structure; social contact; activity; status; purposefu~ess;and sense of control. These aspects of employment are plainly important, and the effects of unemployment must include consequences related to them. So, for instance, Bolton and Oatley (1987) and Eales (1988) argue that when social support continues to be available, the impact of
becoming unemployed is lessened. owever, unemployment itself usually causes changes in the individual's social network("Jackson, 1988). Kessler and colleagues (1987) studied three possible sources of strain: financial, marital,and that due to constrictionof 'affiliative interaction'. Their longitudinal study suggested that the mostimportant and direct consequence was financial, and that financial hardship increased the impact of any s u p e r v e n ~ g ~ s f o r t u nThe e s . importanceof money problems also emerged from thestudy of Frese and Mohr (1987). h most countries women are less likely to be in paid e men, particularly if they are involved in ho~e-making Moreover, if they do work, they may still retain arduous responsibilities at home. Overall, employment appears to have beneficial effects on psychological health, although this may be reduced or reversed if it gives rise to competing pressures and obligations. Leeflanget al. (1992) showed that the impact of short- and long-term unemploy~entdid not differ between men and women. For women, work has a particular set of m~aningsthat lie behind the general association with reduced levels of psychiatric disorder. Rosenfield 9) reviewed the value of employment in buffering women from the tendency to become depressed, and interpreted findings from c o ~ u n i ~ surveys in termsof power and personal control. The extentof any financial difficul~ and of other c o ~ i ~ e nand t sburdens, and the availabili~of social support are crucial, as are interestand involvement. Warr and Parry (198 concluded that most evidencepointed to a protectiveeffect of employ~ent in working class but not inmiddle class women.Parry (1986; Parry 1986) has suggested that the important factor is whether work provides extra social support. If it does not, it may actually increasethe burdens on women. The benefits of employment are weaker in married women (Roberts Csr erts et al., 1982; Warr &: Parry, 1982), still more so if they Gee et al., 1983; Parry, 1986). Full-time employment is particularly demanding (Cleary & Mechanic, 1983; ElliottSs Huppert, 1991). Haw (1995) reports results of a community survey sug~estingthat the beneficial effects of employment are reduced in women with preschool children, but not in those whose children are at school. The most likely ex~lanationfor these findings is role conflict and overload. Thus, part of the excess of depressive disorders in women may be related both to their reduced ~volvementin employment and to the particular strains they are exposed to if they do work. ~ n f ~ r t u n a tthe e l ~economic circumstances in most Western countries are inducin more women to work, and for longer hours.
A number of authors have obtained i n t e r e s ~results from popula~onsin which social differences between the sexes are artificially constrained. Thus er and Printz (1980) found sinxilar of rates depression in male and female ents, as did Jenkins(1985) in employed civil servants. Maffeo et al. (1990) found no sex differences in employed men and women once loyment attributes were controlled. Levav et al. (2991) argued that there more social equality for women living in a kibbutz, and thus if social factors were important deternxinants of depression the sex ratio should also 1. However, in their studF the sex discrepancy was maintained. 1m and Parker (1989) reported that men and women matched for the putative social causes of depression showedno sex differences, even at T-year follow-up. Neither did they show di~erences in attitudinal measures nor the experienceof life events, Despite the attractionof this type of study, their results are ambiguous, as they cannot take accountof the possibility of iased female representation due to selection effects.
difference seen in unpolar depression is not apparent in and this is one of the major ebbington & Ramana, 3~995)~ iscriminatin~features of the two disorders. However, it is of interest that among subjects with bipolar disorder, women tend to have more depressive episodes (see, e.g., Angst, 1986). It is also of interest that there is no sex-related varia~onin the associationof bipolar disorder with marital status: low rates are particular to married subjects of both sexes.
ortant effect on the sex ratio seen in depressive disorders. are methodological problems. Estimates of depressive by the relatively recent isorders in childhood and adolescence are hampered nt of standardized methods of identifying and assessing these n age groups where its expression differs from that in adulthood. S, there isa clear increase in rates as the form of depressive disorder moves in adolescence towards that recognizable in adults. In prepubertal childrenthere is, if anythi a male predominance in and disorders (Anderson . ,1987; Kashani & Carlson, ord, 1990)~while during adolescence the F:M ratio 2 : 1-value seen in adults (Cohen et al., 1993). Moreover, in a
large, clinical sample this ch conduct disorder or anxie( ~ rm that the adult sex ratio has fully develo cGee et al., 1992). idence of severe depressive disorder i both sexes, while the sex ratio ak in earlyadult life; as this stage at which the sex ratio with that for m ~ i (Ta a 1.8).
1987).
Age and the incidence of affective disorders (adapted from Der &~ebbi~~ton,
§ex Disorder type
25-34 15-24
Mania and h ~ p o m ~ i a
Severe affective disorder Male depressions Male All
35-44
45-54
5 5 4 4 75-165-74
Male Female
7.3 9.4
7.2 5.8
5.2 5.7
3.4 6.2
2.3 3.1
5.7 3.9
10.6 4.2
Female
15.3 44.6
31.2 70.0
37.2 72.3
46.8 70.6
39.2 52.6
59.8 71.5
34.2 46.6
Female
154.8 422.0
216.7 505.1
210.0 373.5
185.5 259.1
140.0 179.6
157.9 184.7
113.5 123.7
indings based on prevalence are more ambiguous. sman et al., ~9$$),the sex ratio for 1year -44,45-64 and 65 and over was3.0, National ~ o m o r b iSurve ~ i ~ ler etaE., 1993)~ the sex ratio was maintained sults from 2000 firstollaborati~eStudy on the Psychobiol of ~epressionconfirm th screpancy between therate of m de ressivedisorder in men en peaksbetweenadolescence and early
S,
both for depressive
dent of changes in maritalst by then women have passed through the menopause,and in the past this was seen as a time of increased vulnerability for women, because of both social and hormonal changes. Jorm (1987)carried out a meta-ana~ysisof studies of depression overthe
age range, analysing separately those based on diagnosed cases of depression and those using continuous measures of depressive symptoms. The scatter of results indicateda non-linear relationship, and was best accounted for by a quadratic curve which cut the zero sex difference line at just over 10 and just under 80 years. The equation accounted for 27% of the variance. Thus the sex ratio was much reduced at the ends of the lifespan, but was maintained for some time after the menopausal years. Both men and women peaked in their early twenties, but the female peak was higher and declined more quickly. While the timingof the developmentof a sex differential is clearly related to the timingof puberty, it is not clear whetherit is more strongly relatedto the social or the biological consequences of that change. Since puberty starts at various ages, one wayof assessing the impact of hormonal changes is to use age as a control variable. A number of authors -GUlXl ren, Angold & Rutter, 1992) ha d that ling 1989; Paikoff et al., 1991; for age eliminated the contribution of pubertal stage to the adolescent increase of rates of depre‘ssive disorders in girls, although one well-conducted longitudinal community survey found that time since menarche was the determinant of high levelsof depression in girls, rather than age itself (Patton et al., 1995). We conclude that most evidence favours the importance of bein in a pubertal cohort over the personal experience of pub er^; this is easier t explain in social terms. The adolescent rise in rates of depressive disorder in females cannot, however, be accounted for in unidimensional terms, and originates from the complex interplay of bodily cha es, individual and societal reactions to them, and the other things that te to be happening in irls’ lives at this time (Alsaker,1994). The inconsistent relationship between the sex ratio and the menopause make it hard to argue strongly that the ratio arises from of high depression rates during women’s reproductive lives. This, in turn, weakens the argument for explanations in terms of gonadal hormones, and in any case the hormone system most clearly (and probably causally) related to depression is the corticoid response to stress, which does not vary between (the ~ hsexes ec~ey, 1996). However, the strongest argument against a hormonal aetiology for the sex differenceis the lackof a genetic contribution(~erikangaset al., 1985). It is difficult to imagine a biological cause that does not show individual variability from genetic causes. This opens the door for non-biologica~ explanations, of which, in my view, the most plausible relates to the disadvantages almost universally experienced by women. These are reflected in in the burdens andconflicts of the roles open to them and, consequently, the sociodemographic groupings characterized by high riskof depression.
0
idemiology is endlessly capable of enerating hypotheses that can be tested in more refined and focused studies. This chapter attemptsto of the particular difficulties faced by epidemiologists when they study affective disorder, but at the same time to use epidemiological techniques to provide insights into the likely causes of individual variationsin susceptibility.
Adelstein, A.M, Downham, D.Y, Stein, Z. & Susser, Baldwin, J.A. (1971)Five year incidence of reported M.W. (1968) The epidemiology of mental illness psychiatric disorder. Aspects of the E p i d ~ i o lofo ~ in an English City.Social P s y c k i a t ~3,47-59. , Mental Illness: Studies in Record Linkage. Little, Alsaker, F.D. (1994) Timing of puberty and reacBrown & Co., Boston. tions to pubertal changes. In: Psychosocial Dis- Banks, M.H. & Jackson, P.R. (1982)Unemployrnent and riskof minor psychiatric disorders in young for Preventurbances in Young People: C~allenges people: cross-sectional and longitudinal evition (ed. M. Rutter), pp. 37-82. Cambridge dence. PsychologicalMedicine, 12,789-798. University Press, Cambridge. American Psychiatric Association (1980) Diagnos- Bebbington,P.E. (1987) Life events and schizophrenia: The WHO collaborative study. so cia^ Psytic and StatisticalManual of Mental Disorders, 3rd edn. American Psychiatric Association, Washchiatry, 22,179-180. ington, DC. Bebbington, P.E. (1988) The social epidemiology American Psychiatric Association (1987) Diagnosof clinical depression.In: andb boo^ of Studies in tic and Statistical Manual of Mental Disorders, reSocial P s y c h ~ a(eds t ~ A.S. Henderson & G. Burvised 3rd edn. American Psychiatric Association, rows), pp. 88-102. Elsevier, Amsterdam. Washington, DC. Bebbington, P.E. (1990) Population surveys of psyAmerican Psychiatric Association (1994) Diagnoschiatric disorder and the need for treatment. s, Social P s y c h i a t ~ and Psychia~ricEpidemiolo~, 25, tic and Statistical~ a n u aofl Mental ~ i s o r ~ e r4th edn. American Psychiatric Association, Wash33-40. ington, DC. Bebbington, P.E. & Ramana, R. (1995) The epidemiology of bipolar affective disorder. Social Anderson, J.C., Williams, S. & McGee, R. et al. , (1987) DSM-I11 disorders in preadolescent chil- Psychiatry and Psychiatric Ep id e ~io lo g y30, 279-292. dren. Archives of General P s y c ~ i a t44,69-76. ~, Aneshensel, CS., Frerichs, R.R. & Clark, V.A. Bebbington, P.E. & Tansella, M. (1989) Gender, (1981) Family roles and sex differences in marital status and treated affective disorders in depression.~ournalof ~ e a land South Verona: a case-register study. ~ournaZof t ~Social 3ekaviour, Afective Disorders, Iy, 83-91. 22,379-393. E. & Angold, A. & Rutter, M. (1992) Effects of age and Bebbington,P.E., Hurry,J., Tennant, C., Sturt, pubertal status on depression in a large clinical Wing, J.K. (1981) The epidemiology of mental sample. Developmentalpsycho pa tho lo^, 4,5-28. disorders in Camberwell. PsychologicalMedicine, Angst, J. (1986) The course of affective disorders. 11,561-580. Bebbington, P.E., Sturt, E., Tennant, C. & Hurry, J. P s y c ~ o p a t k o l o19 ~ ,(Suppl. 2),47-52. (1984) Misfortune and resilience: a community Aseltine, R.H. Jr & Kessler, R.C. (1993) Marital study of women. Psyc~ological disruption and depression in a community Medicine, 14,347sample. ~ournal of ~ e a l tand k Social 3ehaviour, 34, 364. Bebbington, P.E., Katz, R., McGuffin,P., Tennant, 23 7-251. Bagley C. (1973) Occupational classsymptom and C. & Hurry, J. (1989) The riskof minor depression before age 65: results from a community of depression.Social Science and Medici~e,y, 327survey. Psycholo~calMedicine, 19,393-340. 339. Bahar, E. (1989)Development and Mental ~ e a l t hA. Bebbington,P.E., Marsden, L., Brewin, &C. Lesage, psyckiatric Epidemiological study in Developing A. (1996a) Measuring the need for psychiatric Country: An Enquiry inPalembang,Indonesia. PhD treatment in the general population: the Comthesis. The Australian National Universitv. munitv Version of the MRC Needs for Care
Assessment.PsychologicalMedicine, 26,229-236. Carta, M.G., Carpiniello, B., Morosini, P.L. & Bebbington, P.E., Marsden, L. & Brewin, C. (1997) Rudas, N. (1991). Prevalence of mental disorders The needs for psychiatric treatment in the in Sardinia: a community study in an inland general population: the Camberwell Needs for m$ing district. Ps~chological Me~icine, 21,1061Care Survey.PsychologicalMedicine, 27,821-834 1071. Birley, J.L.T. (1990)DSM-III: From left to right or Checkley, S.A. (1996) The neuroendocrinology of from right to left? ~ r i t ~ s h Jofo Pus~yac lh ~ a t157, ~/ depression and of chronic stress. ~ r i ~ iMedical sh 116-118. Bulletin, 52,397-417. Birtchnell, J. (1971) Social class, parental social class Cleary,P.D. & Mechanic, D. (1983) Sex differences and social mobility in psychiatric patientsand in psychological distress among married people. general population controls. PsychoZog~cal Journal of Health and Social ~ehaviour/6,64-78. Me~icine,1,209-221, Cohen,,.'l Cohen, J.& Brook., J. (1993)An epideBland, R.C., Newman, S.C. & Om, H. (1988). miological study of disorders in late childhood Epidemiology of psychiatric disorders in and adolescence. 11. Persistence of disorders. Edmonton. Acta Psychiatrica Scandinavica, 77, Journal of Child psycho lo^ and P s y c h i a ~34, ~ ,869Supplementum 338. 877. Blazer, D., George, L.K. & Landerman R. et al. Cornstock, G.W.& Helsing, K.J. (1976) Symptoms (1985). Psychiatric disoders: a rural/urban of depression intwo communities.PsychoZogical comparison. Archives of General P s y c h i a t ~42, , Medicine, 6,551-563. 651456. Cross-National Collaborative Group (1992) The Blumenthal,M.D. & Diehan, T.E. (1975)Depreschanging rateof major depression.~ o u r nof a ~the sive symptomatology and role function in a Ame~icanMedical Association/ 268,3098-3105. general population. Archives of General P s y c h ~ t ~ ,Der, G. & Bebbington, P.E. (1987) Depression in inner London:a register study. Social Ps y c h ia t ~, 32, 985991. Bolton, W. & Oatley, K. (1987) A longitudinal study 22,73434. of social support and depression in unemployed DHSS (1985). In-patient Statistics from the Mental men. PsychologicalMedicine, 17, 453-460. Health Enquiry, 1982, London, HMSO. Bromet,E.J., hnn,L.O., Connell, M.O., Dew, M.A. Dohrenwend, B.P. & Dohrenwend, B.S. (1969) & Schulberg, H.C. (1986) Long-term reliability Social Status and Psychological Disorder: A Causal of diagnosing lifetime major depression in a ~ n ~ uJohn i ~Wiley . & Sons, New York. community sample. Archives of General Psychi- Dohrenwend, B.P. & Dohrenwend, B.S. (1982) Perspectives on the past and future of psychiatric at% 43,435-440. Brooks-Gm, J. & Warren, M.P. (1989) Biological epidemiology.American Journal of Public Health, and social contributions to negative affect in 72,1271-1279. young adolescent girls. Child D e v e l o p m ~60, ~ , Eales, M.J. (1988) Depression and anxietyin unemployedmm. P ~ c h o l o ~ Medicine, cal 18,935945. 4-55. Brown, G.W. & Harris, T.O. (1978) Social Origins of Elliott, J.& Huppert, EA. (199s)In sicknessand in health: associations between physical and menDepression: A Study of Psychiatric Disorder in tal well-being, employment and parental status London. ~ o ~ eTavistock, n . in a British Nationwide sample of married Brown, G.W. ,Davidson,S., Harris,T., Maclean, U., Pollock, S. & Prudo, R.(1977) Psychiatric disor- women. Psycho~ogicalMedicine, 21,515-524. Ensel, W.M. (1982) The role of age in the relationder in London and North Uist. Social Science and ship of gender and marital status to depression. Medicine, 11,367-377. Journal of ~ e r v o uand s Mental Disease, 170,536Burnham, M.A., Hough, R.L., Escobar, J.I. et al. (1987)Sixmonthprevalenceofspecific 543. Faravelli, C. Guerrini-Degl'hocenti, B., Aiazzi, psychiatric disorders among Mexican Americans and non-Hispanic whites in Los Angeles. L., Incerpi,G. & Pallanti, S. (1990). Epidemiology of mood disorders: a community survey in Archives ofGenera1 P s y c h ~ a44,687494. ~~, Florence. J o ~r n aofl A~ectiveDisorders, 20, 135& Regier, D.A. Burke, K.C., Burke, J.D., Rae, D.S. (1.991) Comparing age at onset of major depres- 41. Finlay-Jones, R.& Eckhardt, B. (1981) Psychiatric sion and other psychiatric disorders by birth disorder among the young unemployed. Auscohorts in five US community populations. tralia~and New Zealand Journal of Psychiatry, 15, Archives of General Psychiat~,48,789-795. Canino, G.J., Bird, H.R., Shrout, P.E. et al. (1987) 265-270. The prevalenceof specific psychiatric disorders Fleming, J.E.& Offord, D.R. (1990) Epidemiology of childhood depressive disorders: a critical in Puerto Rico.Archives of General P s y c h i a t ~44, , review. Journal of the American A c a d e ~ yof Child 727-735.
'
22
in a Dutch Health Area. Psyc~ological Medic~ne, and Adolescent Psychiatry,29,571-580. 16: 495-506. Frese, M.& Mohr, G.(1987) Prolonged unemployment and depression in older workers: a longi- Hurry, J., Tennant, C. & Bebbington, P.E. (1980) tudinal study of intervening variables. Social Selective factors leading to psychiatric referral. Acta Psychiatrica, Supplementurn, 285,31y”23. Science and Medicine, 25, 173-178. Gebhardt, R. & Klimitz, H. (1986) Depressive Hwu, H.-G., Yeh, E.-K. & Chang, L.-Y. (1989). Storungen, Geschlect und Zivilstand. Zeitschr~ Prevalence of psychiatric disorders in Taiwan defined by the Chinese Diagnostic Interview fir Klinische Psychologie, 15, 3-20. Schedule. Acta Psych~tricaScandinavica,yg, 136Gershon,E.,Hamovit,J.H.,Guroff,J.J. & Nurnberger, J.I. (1987) Birth cohort changes in 147. manic and depressive disorders in relatives of Jackson, P.R. (1988) Personal networks, support bipolar and schizoaffective patients. Archives of mobilization and employment. Psyc~ological General P s y c h i a t~44, , 314-319. Medicine, 18,397-404. Ghubash, R.(1992), Hamde, E.& Bebbington,P.E. Jackson, P.R., Stafford, E.M., Banks, M.H. & Warr, P.B. (1982) Work involve men^ and e m p lo y ~e n t The Dubai Community Psychiatric Survey: Prevelence and sociodemographic correlates. status as in~uenceson mental health: a test of an i n t e r - a c ~ i omodel. ~ l SAPU Memo404. University Social P s y c h i a t ~ and Psychiatric E p i d ~ i o l o27, ~, of Sheffield. 53-61. Giuffra, L.A.& Risch, N.(1994) Diminished recall Jaco, E.G. (1960) The Social € p id e m io lo ~of Mental Disorders. Russell Sage Foundation, New York. and the cohort effect of major depression: a Jahoda, M. (1982) EmpZoyment and ~ ~ p l ~ simulation study, Psycholo~cal Medicine,W ,375Cambridge University Press, Cambridge. 383. Grad de Alarcon, J., Sainsbury, P. & Costain, W.R. Jenkins, R.(1985) Sex differencesin minor psychiatric morbidity.PsychologicalMedicine (Suppl,7). (1975) Incidence of referred mental illness in J, & Strathdee, Chichesterand Salisbury.PsychologicaZMedicine, Jenkins, R., MacDonald, A., Murray, G. (1982) Minor psychiatric morbidity and the 5, 32-54. Hagnell, O., Lanke, J., Rorsman, 8. & Ojesjo, L, threat of redundancy in a professional group. Psyc~ologicalM e ~ i c ~12,799-807. ~e, (1982) Are we entering an age of melancholy? Depressive illness in a prospective epidemio- Jenkins, R., Bebbington,P., Brugha, T. et al. (1997a) The National Psychiatric Morbidity Surveys of logical study over25 years: the Lundby Study, Great Britain-an Overview.Psycholo~cal MediSweden. PsychologicalM e ~ ~ 13,279-289. c i ~ ~ , Hagnell, O., Ojesjo, L., Otterbeck, L.& R o ~ s m ~ n , cine, 27,765-774. Jenkins, R., Lewis, G., Bebbington, P. et al. (1997b) B. (1994) Prevalence of mental disorders, The National Psychiatric Morbidity Surveys of personality traits and mental complaints in the Great Britain-Initial findings from the HouseLundby Study.A point prevalence studyof the hold Survey. Psycholo~cal Medicine, 27, 775-791. 1957 Lundby cohort 2,612 of inhabitants ofa geographically defined area who were re-examined Joenson, S. & Wang, A.G.(1983). First admissions in 1972 regardless of domicile. Scandinavian for psychiatric disorders: a comparison between the Faroe Islands and Denmark. Acta PsychiaJournal of Social Medicine, 50, 1-77. trica Scandinavica,68: 66-71. Hare, E.H.& Shaw, G.K.(1965) Mental Health on a New Housing Estate: A Comparative Studyof Health Jorm, A.F. (1987) Sex and age differences in depression:a quantitative synthesis of published in two Districts in Croydon. Maudsley Monograph No. 12. Oxford University Press, Oxford. research. Australian and New Zealand Jo~r}zalof Ps y c h i a t ~22/46-53. , Hasin, D. & Link, B. (1988)Age and recognition of depression: Implications for a cohort effect in Joyce, P.R., Oakley-Browne, M.A., Wells, J.E., Bushnell, J.A.& Hornblow, A.R.(1990) Birth comajor depression. Psychologica~ Medicine,18,683hort trends in major depression: increasing rates 688. and earlier onset in New Zealand. J o u r ~ a olf Haw, C.E. (1995) The family life cycle: a forgotten Afective Disorders, 18/83-89. variable in the study of women’s employment Juel-Nielson, N., Bille, M., Flygenring, J. & and wellbeing. Psyc~ologicalMedicine, 25, 727Helgason, T. (1961) Frequency of depressive 738. states within geographically delimited popuHenderson, S., Duncan-Jones, P., Byrne, D.G., lation groups.3. Incidence (the Aarhus County Scott, R.& Adcock, S. (1979).Psychiatric disorders in Canberra:a standardised study of prevalence. Investigation). Acta Psychiatrica Scundinavica, Acta Psychiatrica Scandinavica,60,355-374. (Suppl.) 162,69-80. Hodiamant, P,, Peer, N. & Syben, N. (1986). Kashani, J.H. & Carlson, G.A. (1987) Seriously depressed preschoolers. American Journul of Epidemiological aspectsof psychiatric disorder
~
t
.
(1992)Health effects of unemployment. 11: Men P s y c h i a ~144, ~ , 348-350. Kessler, RC,, Turner, J.B. & House, J.S. (1987) and women. Social Scienceand medicine^ 341351Intervening processes in the relationship be363. tween unemployment and health. Psycho- Lehtinen, V., Lindholm, T., Veijola, J. & Vaisanen, E. (1990). The prevalence of PSECATEGO logical Medicine, 17,949-962. & Swartz, M. et al. Kessler, R.C., McGonagle, K.A. disorders in a Finnish adult population cohort. Social P s y c h i a t ~ and Psychia~ric~ p i d e m i o l o25 ~,, (1993) Sex and depression in the National 187-192. Comorbidity Survey. I: Lifetime prevalence, chronicity and recurrence. Journal of A ~ e c t i v e Leighton, D.C., Harding, J.S., Macklin, D.B., Disorders, 29,8596. MacMilla~,A.M. & Leighton, AH. (1963) The Character of Danger Stirling County Study. Vol. 3. Kessler, R.C.McGonagle, K.A. & Zhao, S, et al. Basic Books,New York. (1994a) Lifetime and 12-month prevalence of e, DSM-111-R p s y c h i a ~ cdisorders in the United Leon, AD., Herman, G.L. & W i c ~ a m a r a ~P. (1993) Continuing female predominance in deStates: results from the National Comorbidity pressive illness.American Journal of Public ~ e a l t h , Study. Archives ofGenera1 Psychiatry,51/8-19. Kessler, RC. & McGonagle, K.A.Nelson, C.B. et $3,754-757. Levav, I., Gilboa, S. & Ruiz, R. (1991)Demoralizaal. (199417) Sex and depression in the National tion and gender differences in a kibbutz. PsyComorbidity Survey. I1 Cohort effects. Journa~ chological Me~icine,21,1019-1028. ofA~ectiveDisorders, 30, s5-26. Klerman, G.L. (1978) Affective disorders. In: The Lewinsohn, P.M., Clarke, G.N., Seeley, J.R. & Rohde, P. (1994) Major depression in commu~ a r v a r dGuide to Modern Psychiatry (eds M. nity adolescents: age at onset, episode duration, Armand & M. D. Nicholi).Belknap Press,Camand time to recurrence. Journa~of the American bridge, M A . Klerman, G.L. (1988)The current age of youthful 33. Academy of Child and Adolescent Ps y c h ia t ~, melancholia:evidencefor increase in depression Lewis, G., Pelosi, A.J., Araya, R.C. & Dum, G. (1992) Measuring psychiatric disorder in the among adolescents and young adults. British J o u ~ aofl Psyc~ia~ry, 152,+"4. community: a standardized assessment for use Klerman, G.L. & Weissman, M.M. (1989) Increasby lay-interviewers. Psychological Medicine, 22, 465-486. ing rates of depression. Journa~of the American ~ e d i c aAssociation, l 261,22292235. tin, T.Y., Chu, H.M., Rin,H., Hsu, C.C., Yeh, E.K. & Chen, CC. (1989) Effectsof social change on Klerman, G.L., Lavori, P.W., Rice, J., Reich, T., mental disorders in Taiwan: observations based Endicott, J. & Andreason, N.C. (1985) Birthon a 15 year follow-up survey of general cohort trends in rates of major depressive disorder among relatives of patients with affective populations in 3 cornmunities.Acta Psych~atr~ca Scundinavica, (Suppl.) 348,1x-33. disorder. Archives of GeneralPsychiat~,421,689Lowenthal, K., Goldblatt, V.,Gorton, T.et al. (1995) 693. Gender and depression in Anglo-Jewry.PsychoKramer, M,, Brown, H., Skinner, A., Anthony, J. & logical ~ e ~ i c i n25, e , 1051-1064. German, P.(1987)Changing living arrangements in the population and their potential effect on Maffeo, P.A.,Ford, T. W.& Lavin, P.F. (1990)Gender differences in depression in an employment the prevalence of mental disorder: findings of l Assessment,551249s'iitthg.J o u ~ aofPersonality the East Baltimore Mental Health Survey. In: Psychia~~ E~ pc ~ ~ e(ed. ~ B. ~ Cooper), o ~ o pp. ~ 3262. 26. Croom Helm, London. Mavreas, V. & Bebbington, P.E. (1987).Psychiatric Krause, N. & Geyer-Pestello, H.F. (1985) Depresmorbidity in London's Greek Cypriot cornmusive symptoms among women employed outnity. I. Association with sociodemographic 22,150-159. side the home. American Jou~nulof C o m ~ u n i t y variables. Social Ps y c h ia t ~, Mavreas, V.G., Beis, A., Mouyias, A., Rigoni, F. & psycho log^, 13, 49-67. Lasch, K.,Weissman, M., Wickramaratne, P. & Lyketsos, G.C. (1986)Prevalence of psychiatric Bruce, M.L. (1990) Birth-cohort changes in the disorder in Athens: a Community study. Social Research, 33, 31-37. rates of.mania. P s y c h i a t ~ P s y c h ~ a tax, ~ , 172-181. Lavori, P.W., Klerman, G.L., Keller, M.B., Reich, McGee, R., Williams, S., Kashani, J. & Silva, P. T., Rice, J. & Endicott, J. (1987) Age-period-co(1983) Prevalence of self reported depressive hort analysis of secular trends in onset of major symptoms and associated factors in mothers in ~ y , 473depression: Findings in siblings of patients with Dunedin. British Journal of P s y c h ~ a ~143, major affective disorder. Journal of ~ s y c ~ i a t ~ i c479. Research, 21, 23-35. McGee, R., Feehan, M., Williams, S. & Anderson, J. (1992) DSM-111disorders from age 11 to age 15 Leeflang, R.L., Klein-Hesselink, D.J.& Spruit, I.P.
years. Journal of the American Academy of Child and Adolescent Psychiatryl31,50-59. McGuffin, P., Farmer, A. & Harvey, I. (1991) A polydiagnostic application criteria in studies of psychotic illness.Development and reliability of the OPCIUT system [news]. Archives of General Psychiatry,48,764-770. Meltzer, H., Gill, B., Petticrew, M. & Hinds, K. (1995) The prevalence of psychiatric morbidity
a ~ o nadults g living in private households.
OPCS Survey of Psychiatric Morbidity in Great Britain. Report I. HMSO, London. Merikangas, K.R., Weissman, M.M.& Pauls, D.L.
through the teens: a multiwave prospective study.
Poster presented at WPA Section of Epidemiology and Community Psychiatry Conference New York, 15-17 May. Paykel, E.S. & Rowan, P. (1979) Recent advances in research on affective disorders. In: Recent Advances in Clinical Psychiatry. (ed K. GrenvilleGrossman), pp. 37-90. Churchill Livingstone, Edinburgh. Pederson, A.M., Barry, D.J. & Babigian, H.M. (1972). Epidemiological considerations of psychotic depression. Archives of General Psychiatry,27,193-197. Richman, N. (1974) Theeffect of housing on preschool children and their mothers. Developm~tal Medicine and Child NeuroZo~,16,53-58. Richman, N. (1977) Behaviour problems in preschool children: family and social factors. British Journal of Psychiatry, 131,523-527. Roberts, C.R., Roberts, R.E. & Stevenson, J.M. (1982) Women,work, social support andpsychiatric morbidity. Social Psychiatry, 17,167-173. differences Roberts, R.E. & OKeefe, S.J. (1981) Sex in depression re-examined.Journal of HeaZth and Social Be~aviour,22,394-400. Robins, L.N. (1995) How to choose among the riches: Selecting a diagnostic instrument. In: Textbook in Psychiatric ~ p i d e m i o l o(eds ~ M. T. Tsuang, M. Tohen & G. E.P. Zahner), pp. 243252. Wiley-Liss,New York. Robins, L.N. & Regier, D.A. (1991) Psychiatric
(1985) Genetic factors in the sex ratio of major depression. PsychologicalMedicine, 15,6349. Moss, P. & Plewis, I. (1977) Mental distress in mothers of pre-school children in Inner London. PsychologicalMedicine, 7,641-652. Murphy, J.M. (1994) The Stirling County study: al of Psychiatry, then and now. I n t e ~ a t i o ~Review 6,329-348. Myers. J.K., Weissman, M.M. & Tischler, G.L. (1984). Six month prevalence of psychiatric disorders in three communities: 1980-1982. Archives of GeneralPsychiatry, 41,959-67. Norris, V. (1959). Mental Illness in London, Maudsley Monograph No. 6. London, OUP. Oakley-Browne, M.A., Joyce, P.R., Wells, J.E., Bushnell, J.A. & Hornblow, A.R. (1989). Christchurch Psychiatric Epidemiology Study, Part I: Six month and other period prevalences of specific psychiatric disorders. Australian and DisordersinAmerica:The ~pide~iological New ZealandJournal of Psychiatry,23,327-340. Catch~entArea Study. Free Press,New York. Odegaard, 0. (1953) New data on marriage and Robins, L.N., Helzer, J.E., Orvaschel,H. et al. (1985) mental disease: the incidence ofpsychosis in the TheDiagnosticInterviewSchedule. In: widowed and the divorced. Journal of Mental Epidemiologic Field Methods in Psychiatry: The Disease, gg, 778-785. NIMH ~ p i d e m i o l o Catchment ~c Area Program (eds Activities of Irish Psychiatric OHare & Walsh (1985). W.W. Eaton & L. G. Kessler), pp. 143-170. Hospitals and Units. Medico-Social Research Academic Press,Orlando. Board, Dublin. J.K., Wittchen, H.U.et al.(1988) Robins, L.N., Wing, Orley, J. Wing, J.K. (1979) Psychiatric disorders The Composite International Diagnostic Interin two African villages. Archives of General view. Archives of General Psychiatry, 45, 1069Psychiatryl 36,513-520. 1077. Paikoff, R.L., Brooks-Gunn, J. &X Worren, M.P. Romans-Clarkson, S.E., Walton, V.A., Herbison, (1991) Effects of girls’ hormonal status on G.P. & Mullen, P.E. (1988)Marriage, motherhood depressive and aggressive symptoms over the and psychiatric morbidity in New Zealand. course of one year. Journal of Youth and AdolesPsychologicalMedicine, 18,9834390. cence, 20,191-215. Rosanoff,A.J. (1916) Survey of mental disorders Parry, G. (1986) Paid employment, life events, soin Nassau County, NewYork. Psych~atricBulle~in, cial support and mental health in working class 2,109-23 1. mothers. Journal ofHealth andSocial Behavio~r,27, Rosenfield, S. (1989) The effectsof women’s em193-208. ployment: personal control and sex differences support and Parry, G.& Shapiro,D.A. (1986) Social in mental health. Journal of Health and Social life events in working class women. Archives of 30,77-91. Behavio~r, General Psychiatry,43,315-323. Rutter, M.(199%)Age changes in depressive disPatton, GC., Hibbert, M. & Bowes, G. (1995)Comorders: Some developmental considerations. In: mon psychiatrics y~ p t o mand s healthrisk b e ~ v i o u r s The evel lop men^ of motion Regulation and
syndromes in a Danish county. Acta Psyckiatrica Dysregu~ation(eds J. Garber & K. A. Dodge), pp. 273-300. Cambridge University Press, Cambridge. Scandinavica,51,28-41. Rutter, M., Tizard, J., Yule, W., Graham, I?. & Weissman, M.M.& Klerman, G.L.(1977) Sex differences and the epidemiology of depression. Whitmore, K. (1976) Research report: Isle of Wight studies, 1964-1974. PsyckologicalMedicine, Archives of General P s ~ c h i a t3,98-112. ~, Weissman, M.M.& Myers,J. (1978) Rates and risks 6,313-332. of depressive symptoms in a United States urban Sadoun, R., Quemada, N. & Chassague, M.M. (1978). Statis~i9uesMedicales des Eta~lissements community. Acta Psyckiatrica Scandinavica/ 57, 219-231. Psyckiatri9ues,Paris, EditionsLNSERM. Schwab, JJ., Bell, R.A., Warheit, G.J. & Schwab, Weissman, M.M., Bland ,R.C. & Canino, G.J. et al. R.B. (1979) Social Order and Mental Healtk: The (1996). Cross-national epidemiology of major Florida ~ e a ~Study. t k Brunner/Mazel, New York. depression and bipolar disorder. ~ournalof the American Medical Association, 276,293-299. Shepherd, M.(1957) A Study of the Major Psychoses in an E ~ g ~ i County, sk Maudsley Monograph Weissman, M.M., Leaf, P.J., Holzer, C.E., Myers, No. 3. London, OW. J.K. & Tischler, G.L.(1984) The epidemiologyof Srole, L. & Fischer, A.K. (1980) The Midtown depression: an update on sex differences in rates. M ~ a ~ longitudinal a n study vs. the ’Mental ~ o u ~of aA~ective l Disorders, 7,179-188. Paradise Lost’ doctrine: a controversy joined. Weissman, M.M., Leaf, P.J., Tischler, G.L. et al. Archives of General Psyckiatry/37, 209-221. (1988) Affective disorders in five United States Srole, L., Langner,T.S., Michael, S.T., Opler, M.K. communities.Psyckolo~calMedicine, 18,141-154. & Rennie, T.A.C. (1962) Mental Healtk in the Wickramaratne, P.J., Weissman, M.M., Leaf, P.J. & Metropolis. Vol. I:TkeMid-Town~ n h u t t a n Study. Holford, T.R. (1989) Age, period and cohort McGraw-Hill, New York, effects on the risk of major depression: results Stangler, R.S. & Printz, A.M. (1980) DSM-111: from five United States communities. ~ o u ~ofa l Clinical E p i d e ~ i o l o42,333-343. ~/ psychiatric diagnosis in a university population. A~erican ~ o u ~of aPsychiatry, l 137,937-940. Wilhelm, K. & Parker, G. (1989) Is sex necessarily a risk factor to depression? Psyckolo~calMediSurtees, P.G., Dean, C., Ingham, J.G., b i t m a n , cine, xg,401-413. N.B., Miller, P.McC. & Sashidharan, S.P. (1983) Psychiatric disorder in women from an EdinWing, J.K.(1972) Principles of evaluation. Zn: Evaluburgh community: associations with demoating a C o ~ m u n i t yPsychiatric Service: The C a ~ ~ eRegister, ~ e ~ 1967-71 l (eds J. K, Wing & graphic factors.British ~ o u r n of a ~Psychiatry/142, A. M.Hailey). Oxford University Press, Oxford. 238-246. Taylor, Lord& Chave, S. (1964) Mental HeaZth and Wing, J.K., Mann, S.A., Leff, J.P. & Nixon, J.M. (1978) The concept of a ‘case‘ in psychiatric €nviron~ent.Longman Green, London. population surveys. PsyckologicaZ Medicine, 8, Uhlenhuth, E.H., Lipmann, R.S., Balter, M.B. & Stern, M. (1974) Symptom intensity and life 203-217. stress in the city. Archives of General Psych~try, Wing, J.,Wmg, J.K., Babor,T., Brugha, T., Burke, J. &Cooper, J.E.(1990) SCAN Schedules for Clini3% 759-764. Vazquez-Barquero, J.L., Diez-Manrique, J.F., Pena,cal Assessment in Neuropsychiatry.Archives of General Psyckiatry/47,589-593. C., Aldana, J., Samaniego-Rodriguez, C. & Menendez-Arango,J.(1987) A community men- Wittchen H.U., Esau C.A., von Zerssen D., Kreig tal health survey in Cantabria: a general descrip- J.C. & Zaudig M. (1992). Lifetime and six-month prevalence of mental disorders: the Munich tion of morbidity. PsyckoZogical Medici~e,17, follow-up study.European Archivesof P s y c k i a t ~ 227-242. and Clinical ~euroscience, 241,247-258. Warheit,G. J., Holzer, C.E. & Schwab, J.J.(1973) An analysis of social class and racial differences in Wolk, S.1. & Weissman, MM. (1995) Women and depressive symptomatology: a community study.depression. In: Review of Psyckiatry Vol. 14, (eds J. M. Oldham & M. B. Riba), pp. 227-259. ~ o u r ~ofaHealtk l and Social Be~viour, 14,291-295. American Psychiatric Press, Washington, DC. Warr, P.B. (1984) Economic recession and mental health: a review of research. T ~ j d s cVOOY k~~ Sociale World Health Organization (1978) menta^ Disorders: Glossary and Guide to their Class~cationin Gesondkeidszorg,62,298-308. accordance with the ~ i n t Revision k of the InternaWarr, P. & Parry, G. (1982) Paid employment and tional Class~cationof Diseases. WHO, Geneva. women’s psychological well-being.Psycko1og~World Health Organization (1992) Tmtk Rev~sion cal ~ulletin,91,498-516. of the ~nte~ational Class~cation of Diseases. WHO, Weeke, A.,Bille, M., Videbeck, T., Dupont, A. & Geneva. Juel-Nielsen, N.(1975). Incidence of depressive
Coldberg and Huxley (1992)have suggested a framework for understanding the pathway by which individuals become defined as mentally ill and eventually reach mental illness services. This framework serves to organize e~idemiologicaldata about mental illness into groupings depend which upon how far along this pathway individuals have reached,and draws attention to the 'filters' through which patients must pass in order to receive specialized treatments. h many countries most patients are referredto the mental illness services by other professionals, so that one can postulate a filter between the community and the referring professional, as well as betweenthat professional and the mental illness service. The framework consists of five levelsat which survey data could be considered, each one correspondingstage to aon the pathway to psychiatric care. A set of four filters are postulated between these five levels, and these areshown in Table 2.1. Table 2.1 Five levels and four filters[from Goldberg &z Huxley, 19921. Level experience community 1who adults TheAll
an episode of mental disorder in the course of 1 year
1 Illness behaviour
2 Primary care attenders (total)
All adults who experience an episode of mental disorder and seek help from a primary care physician
2 Ability to detect disorder
3 Primary care
All adults who are considered mentally disordered their illness services by PCP: heth her or not they satis& research criteria
3 Referral to mental
4 Mental illness
All adults treated by the mental illness services during the courseof a year
4 Admission to psychiatric beds
attenders (PCP) (detected)
services (total)
5 Mental illness services (hospitalized)
Table 2.2
Figures for the annual period prevalence for allmental disorders andfor depression, England andWales. Level
250-315 community 1 The 2 GP attenders 3 Conspicuous psychiatric morbidity illness4 Mental services: all illness 5 Mental services: in-patients
All mental disorders
Depression only
230 101.5
sot
130"
30.6$
20.8
6.82s
3.4
1.2%
"Figures from OPCS 1995 inflated by 33% for annual inceptions. "Figures from Blacker & Clare, 1987 (confirmed Nielsen& Williams, 1980), decreased by 0.80, to account for consultation rate. $,g,¶ From Goldberg& Huxley, 1992 (p. 47); see Goldberg& Huxley, 1992 (p. 29) for all mental disorders, and for corrections to notes "and t.
Recent estimates for all mental disorders for England Wales and are shown in Table 2.2. The figure for thecornunity is obtained by averaging results from various surveys (which produce point prevalence data), and then i n ~ a them ~ g to take account of inceptions that occur for each disorder durin a calendar year. It can be seen that there are many more disorders in the cornunity than are treated by primary care services, and that more disorders by specialist care. However, the various filters are treatedin primary care than are selectively permeable to disorders of greater severity:so that severe mental disorders account only for about 4.5% of total disorders in the communit)r, yet they account for50% of those seen at level 5. The same effect is seen for severity within a particular disorder: so that many depressive illnesses seen in communi^ settings are self-limiting disorders that will remit w i t h o ~ t specialist treatment, whereas those referred through to the specialist services are morelikely to have a previous episode of illness, a family historyof that illness, and abnormal personalities. Depression is the commonest mental disorder seen c o ~inu n i t settings, y and estimates forits prevalence at all levels can alsobe found inTable 2.2.
Depressive disorders encountered in community surveys are most likely to be accompaniedby symptoms of anxiety, and least likely be seen to in subjects with abnormal personalities. Among women in Britain, the weekly period prevalence of xnixed anxie~/de~ression is gg/~oooat risk, and the weekly prevalence of pure depressive episodeis 25/1000at risk. Thecorrespond~g
res for menare 54 and 17/1000.There are interesting ethnic differences, with Asian and riental women having particularly high rates for mixed anxiety/depression, although there are no clear effects for males (see Table eatest in community settings,and becomes lessat level neral practitioner(GP) referral) is selectively permeable
Weekly period prevalence per1000 population at risk for mixed anxiety/ depression and depressive episode, by ethnic group (OPCS, 1995).
Mite Men
Women Disorder Men
Women
Mixed anxiety/ 96 depression Depressive episode
West Indies or Africa
Asian or Oriental
136
58
68
18
21
Men 54 17
24
51
160
19
Odds ratios for social and demographic characteristics with significant odds ratios for either mixed anxiety/depression or depressive episode (from OPCS, 1995).
Odds ratios
Mixed anxie~/depression
Depressive episode
Female gender Lone parents Living alone Unemployed Economically inactive Urban life
1.73"" 1.48"" 1.47"" 1.73"" 1.19 NS
1.47 NS 2.60"" 2.32"" 2.66"" 2.87**
1.30 NS
1.38""
**significant P
It can be seen from Table 2.4 that loneliness and unemployment areshared two conditions: there is a preponderance of females only factors between the for mixed anxiety/depression, and urban living, economic inactivity and ~ualifications(not shown, but odds ratio = 1.6'7"")for depressive n a study of 100 couples who had experienced at least one nic life event, it was shown (Nazroo et al., 1996) that women were at greater riskof depression than men. This greater risk was restricted to those events involving children, housing or reproductive problems, and was only seen in those couples with clear gender differences roles,inwith the woman viewing herself as being more responsible for these functions. In couples without these role differences, there was no gender difference in risk. Many of these individuals do not see themselves as ill, despite possessing sufficient symptoms fora diagnosis of a mental disorder. They may see their state as being inan understandable rela~onshipto some lifesituation, or they may seek help only for somatic symptoms which often accompany these states.
While nothing can reallybe said about the response tot r e a ~ e nin t people who have not sought medical care, there was an advertisement in local newspapers for people with "symptomsof depression (Lipman et al., 1986) who were not in psychiatric treatment but who were prepared to take part i a 10-week research project with drugs. The 'symptomatic volunteers' (SVs) needed to be two standard deviations above the mean for depressive s ~ p t o m sand , were given a range of depression scales( H a ~ l t o Depression n Scale, Raskin Depression Screen and Research Diagnostic Criteria for Depression). TheSVs were each given a 2-week placebo washout period, and it was necessary for them still to be depressed at the endof this period. They were then assigned randomly to imipramine, chlordiazepoxide or placebo, with regular clinic visits. Differences and treatment was continued8 weeks for at the end of the first week generally favoured chlordiazepoxide but these early advantages were soon lost, apart from subjects on this drug reporting generally better sleep. By the endof the trial most effects favoured i ~ p r a ~ e , and these effects were seen both for anxiety and depression. itHowever, should be noted that the effects of imipramine were only 13% better than placebo (65% vs. 52% for global improvement, patient self-rated).
Many patients attend theirGP with somatic symptoms that are accompanying states of depression and anxiety. They often do not see themselves as psychologically ill, but seek either an explanation for their symptoms or treatment which will relieve these Symptoms. The WHO study of mental disorders in general health care (Ustun st Sartorius, 1995) confirmed previous British studies (for example, Bridges st Goldberg, 1985), by demonstrating that mental disorders usually present to primary care physicians with phys symptoms. In Manchester 176% presented in this way; across the world 69% did so. Across all centres the most common maincomp~aintof patients with mental disorders was pain (29%), fatigue or poor sleep (17%)~ and other asso somatic complaints(33%); only 5% presented with psychological symptoms. Some of these illnesses (47% in Manchester), are accompanying physical illnesses known to their doctors, but others are presenting with unexplained somatic symptoms. Thus, the old 'either /or' dichotomy, whereby patients were to be thought of as having either a physical illnessor a mental one,is inappropriate: and this was true all in centres. Across the world, some 17.1% of male, and 12.5% of female, consecutive attenders at GP surgeries have major depression according to the primary care version of the Composite ~ t e ~ a t i o nDiagnostic al Interview (CIDI-PC).
~ h a ~ t2:e ~ r ~ ~ e rlevels e n tof care This gender difference is statistically significant (see Table 2.5). However, there is considerable divergence between centres, and even theodds ratios need to be interpreted with caution: for example, odds the ratios are very different for Rio and Manchester, but this is produced by a great difference in the male rates for depression; the female rates are nearly identical. In Nagasaki rates are low for each sex, while in Ibadan the male rate is (non-significantly) higher than the female rate. Table 2.5 Rates for current depression according to the CIDI-PC by gender for six
selected centresfrom the WHO Study of Mental Disorders in General Medical Settings, showing odds ratios (Goldberg& LeCrubier, 1995). Current depression
CID1 (primary care)
11.2 5.8 13.9 4.8 5.3 2.3 7.1
36.8 19.6 18.3 13.3 3.8 2.8 12.5
Centre OddsWomenMen Study in WHO Santiago de Chile Rio de Janeiro Manchester Bangalore Ibadan Nagasaki All 15 centres
4.73 3.95 1.40 NS 3.05 0.70 NS 1.28 NS 1.89
Detection of these depressive episodes byprimary the care physicians was comparable with previous reports. Across the world, about 54% were diagnosed, but there were great variations between centres, with physicians in Verona, Manchester and Santiago detecting about 7 0 % ~while their colleagues in Nagasaki, Athens and Shanghai detected only about 20%. Depression was associatedwith high rates of self-reported disability, The only demographic variableshown to be associatedwith depression in this study was pooreducational level, but it should be notedthat this was the only variable recorded in this particular study that reflected social disadvantage. Physical ill-health, as assessed by theprimary care physicians, was not significantly related to depression in this study, nor were parous women at greater risk (Usrun & Sartorius, 1995, p. 330).The depressive syndrome observed varied from centre to centre, despite the constraints imposed by the DSM-111-R definition. Thus fatigue, often accompanied by insomnia and depressive thoughts, was common in Athens, Berlin, Groningen, Ibadan, Manchester, Paris and Verona: all European centres, with the exception of Ibadan. In Nagasaki, fatigueand depressive thoughts are accompanied by worthlessness and weight loss.In Bangalore thereappeared to be a specific pattern, with thoughts of death and worthlessness being relatively common, while in the three American centres depressive thoughts and insomnia were common, often accompanied by weight gain (Goldberg & LeCrubier, 1995).
Ormel et al. (1993) have examined the long-term outcome of depressions and anxiety states encountered in primary care settings, using cases confirm by two-stage screening. They found that pure depressions had a better outco than either anxiety states or mixed anxiety/depressions at 3.5-year followup, but that partial remission, rather than full recovery, was the rule.
h centres with high rates for detection, it is doubtful whether detection of
these extra cases would bestow much more than marginal benefit on the patients; probably in the shape of a shorter duration of episode. Secondary analyses of data collected in surveys by Bridges and Goldberg, Kisely and Goldberg, and Creed and Ronalds indicateinthat Manchester the undetected depressions are milder than the detected ones, and have a similar good outcome at follow-up to those depressions detected and treated by the prim care physicians (unpublished data). Simon and VonKorff (1995) have produce of major depression detected at second similar data in Seattle: 64% of their cases stage of the WHO study were recognized as depressed by the family physician and 56% of these were prescribed and took antidepressant medication. The recognized cases were more severe than the u n r e c o ~ i ~ ecases d (GH scores: 15.3 vs. 11.1, P = 0.006), but the outcomes were similar. The authors argue thatin centres such as theirs t r e a ~ e nresources t shouldbe focused on more intensive follow-up and relapse prevention, rather than on better detection. Inother centres, with low rates of detection, more might be gained by increasing detection rates in terms of better patient outcomes.
sive illnessesat this level are necessarily a mixed bag, as many of these S are not foundto have illnesses at second-stage exam~ation, but to have other mental illnesses, or no illnesses at all. Freeling and colleagues ( studied 143 patients identified as depressed by their GPs and treated with antidepressants;of these only43% conformed with diagnostic criteria for either probable or definite depression.With modern aids to diagnosis, suchas the )~ primary care version of the ICD-10 (10-TO-PC,Ustun et al., ~ 9 9 5both diagnosis and m ~ a g e m e nof t depressive disorders are likely to improve in the future. Schulberg and colleagues (1995)have carriedout a feasibility study of pharmacological treatments for depression in primary care, following gested by a national body; they found that33% only of patients initially assigned to receive medication actually completed it, and that the guidelines were ’feasible but complex’,
A number of studies have randomized depressed patients to an active antidepressant drug or to a placebo, and failed to show a difference in outcome (see, for example Porter, 1970; Blashki et al., 1971; Hazel1et al., 1995). However, zell’s study relates to depressed adolescents, while earlier studies might ve contained heterogeneous populations. Tan et al. (1994) randomized depressed, elderly patients to lofepramine or placebo, and found that both improved by a similar amount during the trial. Lofepramine tended to be more effective than placebo in those patients who were more depressed. On the other hand, subjects who were less depressed improved more on placebo than on lofepramine. Brown and colleagues (1988) point outthat placebos are especially effectivein those who are less depressed (shorterduration or fewer symptoms), and may be as high as 70%~rather greater thanthe figures likely to respondto antidepressants among more depressed patients. Katon et al. (1995) randomized depressed patients to ’usual care’ or e~anced or ’multifaceted care’ (MC). The last consisted of greater frequency of visits, two with a psychiatrist. In those with major depression, MC produced greater compliance, greater satisfaction with treatment, and patients rated antidepressantsas more helpful and had greater symptom reduction. in minor depression, MC produced better compliance but not better or satisfaction. Banerji et al. (1989) found no differences in outcome between depressed primary care patients randomized to alprazolamor amitriptyline, although there was a more rapid responsein those receiving the antidepressant.This finding is consistent with that described by the Depression Guideline Panel ( ~ 9 9 3which )~ reports differences between the responses toactive drug and placebo which are as great for alprazolam and buspirone as those for conventional antidepressants (meta-analysisof eight studies for the former, 76 studies for the latter). In 1985 the National Institute of ( N I ~ HConsensus ) Development Panel ( ~ I ~ 1985) H , cou generalizing, from one field to another, but 8 years later this valuable lesson did not appear to have influenced the Depression Guideline Panel, which produced evidence for the efficacy of antidepressants almost entirely from samples treated by specialist mental health services (Depression Guideline Panel, 1993). Table 6 of that report claims to cover primary care studies, but only twoof the quoted studies are placebo controlled, and of these one (Rckels et al.,1987) should have been included with the mental health studies. Paykel et al. (1988) carried m t one of the best studies in British general practice, and this showed that confirmed depressives with Hamilton depression scores reater than15 (approximately comparable to ICD-10 major depression) did indeed benefit from amitriptyline rather than placebo medication.
33
Part of the problem isthat without quality controlof the patients being diagnosed, the placebo response is often as great as the antidepressant response. Quitkin et al. (1993) studied a group of atypical depressives with reactive moodand hyperphagia, hypersomnolence and rejection sensitivity. These were randomized to imipramine, phenelzine and placebo. Phenelzine was consistently found to besuperior to imipramine. ~~ectiveness of ~ s ~ c h o t h e r a ~ini egeneral s
Thomas (1987) showed the importance of 'being positive' in his management of emotional distress. Sixty-four per ofcent patients randomized toa 'positive' consultation improved,as compared with 39% ofthose receivinga 'negative' consultation.In the positive consultation the patient wasa firm given diagnosis and told confidentlythat he/she would be betterin a few days, while in the negative consultation the patient was told,'I cannot be certain what the matter is with you.' Prescriptionof a placebo made no differenceto outcome in this study. Brown (1994)argues that if the drug-placebo is given in a way which produces hope and an expectancy of improvement, then it is likelyto do so. Elkin et al. (1989) claim that the placebo response is often as high as5 0 % ~ and argue that no psychotherapeutic treatment for depression can better the effects of a properly administered drug-placebo. Robinsonet al. (1990)argue that the advantage of psychotherapies over pill-placebo are an effect size of only 0.28 (NS), to be comparedwith an advantage over waiting list control with an effect size of 0.84. Four studies are usually quoted in support of cognitive behaviour therapy in general practice (Blackburn et al., 1981; Teasdale et al., 1984; Ross & Scott, 1985; Miranda & Munoz, 1994)~but none of these assessed the psychotherapy againsta drug-placebo; the control conditions were activedrugs, treatment as usual and waiting list. Robinsonet al. (1995) examine the extent to which cognitive behavioural techniques are being used by primary care physiciansin their treatments of depression. Sixty-one per cent of patients in their survey say that physicians advise them to identify activities that they are already undertaking which help to make them feel better. These suggestions are associated with a greater useof such strategies, and with better compliancewith medication during treatment. Mynors-Wallace et al. (1995) have overcome Elkin and colleagues' objections, by showing that a method of psychotherapy (problem-solving) was as effective as an antidepressant in the treatment of major depression, and that both active treatments were more effectivea drug-placebo. than Sixty per cent of those patients given problem-solving improved, to be compared with only 27% given the placebo. In summary, before rushing into antidepressant treatment the wise doctor will carry out a careful diagnostic evaluation. Antidepressant medication
34
C ~ ~2: ~~ i~ ~e elevels r r e of~care ~
should be reserved for those with severe depressive illnesses, those with suicidal thoughts, and those with well-established, previously untreated illnesses. Others might benefit froman expectant treatment, with hypnotics used to secure sleep, and a return visit aftera short time. The importance of a positive attitude, restoring hope and the expectation of improvement cannot be overemphasi~ed.Whether the doctor uses a placebo is a matter of professional preference; what is important is that potentially dangerous drugs are not used for what is, in fact,a placebo response. Of the psychotherapeutic ~ o s s i b i l i ~only e § problem-solving seems better than a drug-placebo, but the use of cognitive t e c ~ i ~ umay e s enhance the quality of care and possibly reduce the riskof relapse.
It will be recalled fromTable 2.2 that we are now dealing witha small subset of depressive illnesses. In countries where GI's act as gatekeepers to the specialist services, the ready availability of psychotropicsin primary care and the delays often associated with secondary referral mean that placebo responders havebeen, to some degree, removed. Patients coming to mental health specialists are more likely to ahave positive family history, and to have had previous episodesof mental illness than those seen in primary care.On average, the illnesses seen by mental health specialists are more severe in the L% Huxley, sense of patients having higher mean symptom counts (Goldberg 1992),although it should be appreciated that owingto the greater sizeof the in absolute numbers there are more population in primary care,it is likely that severe illnesses in that location. Because assessments by mental health 20 minutes orso available professionals typically take much longer than the in primary care, it is likely that the diagnostic assessment will be more thorough, and antidepressants will only be offered to those satisfying diagnostic criteria for depressive illnesses. Piccinelli and Wilkinson (2994) have shown that depressive illnesses at this level have only a 25% chance of recovering from the episode and then x year, and75% staying well forIO years. Twenty-five per cent relapse within within IO years; 22% never recover.
It wouldbe tedious to rehearse the evidence that antidepressants do indeed have a significant effect on these illnesses; the interested reader is referred to the meta-analysesof treatment trials publishedby the Depression Guideline
35
Panel (1993, see pp. 47-54). It is worth recordin harmacological actions appear to be effective, an which is effective for all patients. table side-effects:amitriptyl~e, edating, amoxapine and the S )are alerting.SSRIs are free fr do not causeorthostati~hypotension, cardiaca so on. Tricyclics (TCAs) have therapeutic effe SSRIs for 54%, monoamine oxidase inhi heterocyclics for62%. The sizeof the active for the heterocyclics,20.1% for theSSRIs, 21.3% for the TCAs, and 30.9% for the MAOIS..As previously mentioned, amon 'anxiolytics' alprazolam and buspir~nehave effects comparable to tricyclics, but chlordia~epoxideand ) carried diazepam have little antidepressant effect. Katona e f al. ( 1 9 9 ~have out a placebo-controlled trial of lithium augmentation o therapy by either ~uoxetine(an SSRI) or lofepramine (a TCA); bothof these rugs were equally ctive, but the patients who had treatment supplemented by lithium hada reater responseto treatment, butit was necessaryto ensure a lithium levelof at least0 . 4 m o l L - to ~ achieve this advantage.
The Depression Guideline Panel (1993) carried out a meta-analysis of 1 2 randomised controlled trials (RCTs)of cognitive behaviour thera it to be effective in 46.6% of cases, witha therapy vs. drug-plac of only 8.3% (one study only), but an advantage over waitin of 15.3% (three studies). Interpersonal and over active drug therapy psychotherapy on the other hand, was effective with 52.3% of patients, with an advantage over drug-placebo of 22.6% (also only one study), and an advantage over active drug therapy of 12.3%. The advantage of behaviour therapy over drug-placebo is unknown;itbut is effective in55.3% of p with an advantage over active drugsof 23.9% (two studies only), a waiting listof 17%. Brief, dynamic therapy is effective in only of34.8% patients, with an advantage over active drugs of 8.4% (two studies only).
Patients with physical disease may become depressed a number for of reasons including: the disease process itself, treatment received or a psychological reaction to finding oneself with a serious illness. It is also possible for depressive states to antedate the onset of physical disease. depressive states will improve the quality of life r e m a i ~ n g to the patient, and
r survival times than controls, and Gruber et al. (1993)showed those receiving relaxation, guided imagery and biofeedback training had higher and lymphocyten6togen response compared with controls. reer et al. (1992)studied the effectof ’adjuvant psychological therapy’a brief, problem-focused, cog~itive-behaviouraltreatm specifically designed for the needs of individuals-on the with cancer. The new treatment produced higher scores on scales measuring spirit and psychosocial adjustment to illness, and lower scores on helplessness. At 4-month follow-up, differences between index and control patients were more marked for anxiety than for depression.
Patients who are admitted to in-patient beds are more severely depressed than those treated in ambulant settings; they have higher mean symptom counts, are morelikely to have psychotic phenomena and morelikely to be seriously suicidal. It follows that placebo responses are lessof a problem in this setting, and it becomes more reasonable to compare one drug with another, without including a placebo arm. Considering the greater severity of these illnesses, it is of interest that the results of treatment are comparable with those at level 4: the proportion responding to antidepressant§ is between 50 and 55% for all drug groups, and the drug-placebo differences are 18.4%, TCAs 25.1%, SSRIs 25.5% and heterocyclics39.3%. Raskin et al. (1978) carried out a multicentre study with360 depressed in-patients, randomized to imipramine, chlorpromazine or placebo. There were no significant differences between the effects of irnipramine and chlorpromazine, but both were more effective than the placebo. Patients receiving imipramine for more than 6 months had fewer relapses, while those responding positively to the placebo initially had the poorest outcome at Iyear: 57% had ‘ups and downs’ with marked fluctuationsin course. Elec~oconvulsi~e treatment (ECT) produces better short-term results for patients with retardation or depressive delusions, but the results at 6 months are no different from simulated ECT. For those without these features, real ECT confers no additional benefits over simulated EGT (Buchan et al., 1992). Bilateral, brief-pulse ECTjust is as efficacious as traditional ECT with constantvoltage modified sine-wave stimuli (Scott et al., 1992). As at level 4, lithium aug~entationhas a place in the treatmentof drugresistant depressions, and recurrent depression.
In sum mar^ the depressive illnesses that are seen in the community are unlik
those seenby mental health professionals, in that they are more likely to remit spontaneously so that the performance of a drug-placebo is very good. Improvement with a placebo is associated with aillness, short a precipitating event and a good response to previous antidepressant treatment (Brownet al., 1992).The enthusiastic claims for the efficacy of non-drug treatments for depressive illness seen in community settings be must considered with this in mind. However, severe depressive illnesses and those accompanied by ideas of suicide, should continue to attract pharmacological interventions. Depressions seen in association with serious physical illness merit energetic treatment, since not only will quality of life be improved,but the courseof the associated Dhvsical illness mav alsobe favourably influenced. I
J
J
Banerji, J.R., Brantingham, P., McEwan, G.D.et al. (1989) A comparison of alprazolam and amitriptyline in the treatment of patients with neurotic orreactive depression. Irish ~ o u ~ofa l Medical Science, 158,110-113. Blackburn, I., Bishop, S. & Glen, A. (1981) The efficacy of cognitive therapy in depression. British ~ournalof P s yc k i a t ~139, / 181-189. Blacker, C.V.R. & Clare, A.W. (1987) Depressive disorders in primary care. British ~ournalof P s y c k i a t~150, , 737751. Blashki, T.G., Mowbray, R. & Davies, B. (1971) Controlled trial of amitriptyline in general practice. British Medical ~ournai,i, 133-138. Bridges, K. & Goldberg, D.P. (1985) Somatic presentation of psychiatric diseases in primary care. ~ournalof Psyckosomatic Kesearck, 29,563569Brown, W.A. (1994) Placebo as a treatment for 10, 265depression. ~europsyckopkarmacolo~, 269. B.E. & Wernicke, J.F. (1988) Brown, W.A., Dornseif, Placebo response in depression: a search for ry 26,259-264. predictors. P s ~ c ~ i a tKesearck, Brown, W.A., Johnson, M.F. & Chen, M.G. (1992) Clinical features of depressed patients who do and do not improve with PBO. Psychiatry Kesearck, 41, 203-214. Buchan, H.,Johnstone,E., McPherson, K., Palmer, R.L., Crow, T.J. & Brandon, S. (1992) Who benefits from electroconvulsivetherapy? Combined results of the Leicester and Northwick Park trials. British ~ournalof Psychiatry, 160, 355-359. Cohen-Cole, S.A., Brown, F.W. & McDaniel, SJ. (1993) Diagnostic assessment of depression in In:Psychiatric Careof theMed~cal the medically ill. Patient (eds A. Stoudemire & B. Fogel), pp. 53-
70. Oxford University Press, New York. Costa, D., Mogos, I. & Toma, T. (1985) Efficacyand safety of Mianserin in the treatment of women with cancer. Acta Psyckiatrica Scandinavica,320, 85-92. Depression Guideline Panel (1993) D ~r e s s io nin PrimaryCare. Vol. 2: ~ r e a t m e n tof Major Depression. US Dept of Health and Human Services, Washington,DC. Elkin, I., Shea, M.T., Watkins, J.T. et al. (1989)NIMH treatmentof depressioncollaborative treatment programme: general effectivenessof treatment. Archives of General Psychiatry,46,971-982. & Haggerty, J.J. (1988) Evans, D.L., McCartney, C.F. Treatment of depression in cancer patients is associated with better life adaptation: a pilot study. Psyckoso~aticMedicine/ 50, 72-76. et al. (1990) Fawzy, F.I., Kemeny, M.L., Fawzy, N.W. A structured psychiatric interventionfor cancer patients. 11: Changes over time in immunological measures Arckives of General Psychiatry, 471 729-735. Fawzy, F.I., Fawzy, N.W., Hyun, C. et al. (1993) Malignant melanoma: effects of an early structured intervention coping, and affective state on recurrence and survival 6 years later. Archives of General P s y c ~ i a t50, ~ , 681-689. Freeling, l?.,Rao, B.M., Paykel, E.S., Sireling, L. & Burton, R. (1985) Unrecognised depression in general practice. British Medical ~ournal,190, 1880-1883. Goldberg, D.P. & Huxley, P. (1992)Common Mental Disorders: A Biosocia~ Model. Routledge, London. Goldberg, D.P. & LeCrubier, Y. (1995) Form and In: frequency of mental disordersacross centres. Mental Disordersin GeneralMedical Settings (eds V. Ustun & N. Sartorius),p. 398. John Wiley & Sons, Chichester.
39
Greer, S., Moorey, S., Baruch, J.D. et al. (1992) (NIMH) National Institutes of Mental Health Adjuvant psychological therapy for patients (1985) Consensus Development Panel-Mood with cancer: a prospective randomised trial. Disorders:Pharmacologicprevention of recurrences. A~erican ~ournal Bri~ish Med~cal ~ournal, 304, 675-680. of P s ~ c h i f f142, t~, Gruber, B.L., Hersh, S.P., Hall, N.R. et al. (1993) 469-476. (OPCS) Office of Populations Censuses and Immunological responses of breast cancer Surveys (1995)Mor~idityin GreatBritain, R ~ o r t patients to behavioural interventions. Biofeedbac~ Se~Regulation, 18/1-22. no. x. The prevalence of psychiatric ~ o r ~ i d among ity adults living in private housekolds. HMSO, Hazell, P., O’Connell, D., Heathcote, D. & London. Robertson, I. (1995) Efficacy of tricyclic dmgs T., Brilman, E. & in treating child and adolescent depression: a Ormel,J.,Oldehinkel, VandenBrink,W. (1993) Outcome of depression meta-analysis. British Medical ~ournal,310,897and anxiety in primary care. Archives ofGeneral 890. Katon, W., VonKorff, M., Lin, E. et al. (1995) Psychiatry/50,759-766. Collaborative management to achieve treatment Paykel, E.S., Hollyman, J. & Freeling, P. (1988) Predictorsoftherapeuticbenefit from guidelines: impact of depression in primary amitriptyline in mild depression: a general care. ~ournalof the American Medical Associa~ion, practice placebo controlled trial. ~ournalof 273, 1026-1031. Katona, C., Abou-Saleh, M.T., Harrison, D.A.et al. Afective Disorders, 14,83-95. (1995) Placebo controlled trial of lithium Piccinelli, M. & Wilkinson, G. (1994) Outcomes of depression in psychiatric settings. augmentation of fluoxetine and lofepramine. British ~ournal of Psychiatry/164, 297-304. British ~ o ~ r nofa Ps l y c k i a t ~166,80436. , Levy, S.M., Herbermam, R.B., Lippman, M. & Porter,A.M.W. (1970) Depressive illness in general controlled d’Angelo, T. (1987) Correlation of stress factors practice. A demographic study aand sk ~ o ~ ri, ~ a ~ , with sustained depression of natural killer cell trial of imipramine. B r ~ t ~Medical activity and predicted prognosis in patients with 773-778. Quitkin, EM., Stewart, J.W., McGrath, PJ. et al. breast cancer.~ournalof ClinicalOncology, 5, 34% (1993)Columbiaatypicaldepression: a 353. subgroup of depressives with better response Levy, S.M., Herbermann, R.B., Whiteside, T., Sanzo, K., Lee,J. & Kirkwood, J. (1990) Perceived to M O 1 than to tricyclic antidepressants or placebo. British ~ o u ~of aP ls ~ c h i a t21, ~ ,(Suppl.), socialsupportandtumouroestrogen/ progesterone receptor status as predictors of 30-34. N. & Schulterbrand, NK activity in breast cancer patients. Psyckoso- Raskin, A., Boothe, H., Reatig, J.R. (1978) Initial response to drugs in depressive ~ a t i c ~ e d i c52/73-85. ine, illness and psychiatric and community adjustLipman, R.S., Covi, L., Rickels, K.et al. (1986) Imipramine and chlordiazepoxide in depression ment a year later. Psychological Medicine, 8, and anxiety disorders. Archives of General 71-80. Richardson, J./Sheldon, D., Krailo,M, & Levine, Psychiat~,43, 68-86. A. (1990) The effect of compliance with treatMcDaniel, J.S., Musselman, D.L., Porter, M.R., ment on survival among patients with haematoReed, D.A.& Nemeroff, C.B. (1995) Depression logic malignancies.~ o u r n of a ~Clinical Oncology in Patients with cancer. Archives of General P s y c k i a t ~52’89-99. , 8,356364. Miranda, J. & Munoz, R. (1994) Intervention for Rickels, K., Chung, H.R., Csanalosi, etI.B. al. (1987) Alprazolam,diazepam,imipramineand minor depression in primary care patients. placebo in out-patients with major depression. PsychosomaticMedicine, 56,136-141. Archives of General Ps y c h ia t ~/ 44, 862-866. Mynors-Wallace, L.,Gath, D., Lloyd Thomas, A. & Tomlinson, D. (1995) RCT comparing problem Robinson, L.A., Berman, J.S. & Neirneyer, R.A. (1990) Psychotherapy for the treatment of solving treatment with amitriptyline and depression: a co~prehensivereview of conplacebo for major depression in primary care. trolled outcome research. PsychoZogic~l Bulle~~n/ Britisk Medical ~ournal,310, 441-445. Nazroo, J.Y.,Edwards, A.C.& Brown, G.W. (1996) 108~30-49. Gender differences in the onset of depression Robinson, P., Bush, T., VonKorff, M. et al. (1995) Primary care physician use of cognitive befollowinga shared lifeevent a study of couples. havioural techniques with depressed patients. PsychologicalMedicine 2y,9-19. Nielsen, A.C. & Williams, T. (1980) Depression in ~ a mPracfice, ~ l ~40, 352-357. ambulatory medical patients. Archives ofG ~ e r a l Ross, S. & Scott, M. (1985) An evaluation of the effectivenessof individual and group cognitive Psychiatry/3y, 999-1004.
therapy in the treatment of depressed patients E. (1989) Effectof psychosocial intervention on survival of patients with metastatic breast in an inner city health centre.~ournalof the Royal cancer. Lancet, ii, 888-891. College of General Practitioners,35,23cyq2. Sapolsky, R.M. & Donnelly, T.M. (1985) Tan, R.S., Barlow, R.J., Abel, C. et al. (1994) The effect of low dose lofepramine in depressed Vu~erabilityto stress-induced tumour growth elderly patients in general medical wards. increases with age: role of glucocorticoids. ~ n d o c r i n o l o117,662-666. ~, British ~ournalof Clinical P ~ a ~ a c o 37,321lo~, Schulberg, H.C., Block, M.R., Madonia, M.J., 324. Rodriguez, E., Scott, C.P. & Lave, J. (1995) Teasdale, J., Fennell, M. & Hibbert, G. (1984) Applicability of clinical pharmacology guideCognitive therapy for major depression in ~, lines for major depression in primary care primary care. British ~ournalof P s y c h i ~ t144, , 400-406. settings. Archives of Family ~ e d i c i n e4,106-112. Scott, A.I., Rodger, CR., Stocks, R.H. & Shering, Thomas, K.B. (1987) General practice consultaA.P. (1992) Is old-fashioned electroconvulsive tions: is there any point in being positive? Bri~ishMedical ~ournal,2g4,1200-1202. therapy more efficacious? A randomised comparative study of bilateral brief-pulse and Ustun, B. & Sartorius, N. (1995)Mental s is orders bilateral sine-wave treatments. British ~ournalof in General Medical Settings.John Wiley & Sons, Psyc~iatry,160,360-364. Chichester. Simon, G. & Von Korff, M. (1995) Recognition, Ustun, B., Goldberg, D., Cooper,J., Simon, G.E. & Sartorius, N. (1995) New classification for management and outcomes of depression in mental disorders with management guidelines ,primary care. Archives of Family Medicine,4,99105. Chapter 5. for usein primary care: ICD-10 PHC, Spiegel, D., Bloom, J.R., Kraemer, H.C.& Gottheil, British ~ournalof General Practice,45,211-215.
STUART G H E C K L E Y
The purpose of this chapter is to review those advances in biological psychiatry which are of theoretical or practical relevance to the clinicianwho is assessing and treating depressedpatients. The subject matter fallsnaturally into three sections. The first concerns aetiology: the biological causes of depression. The second concerns pathogenesis: the biological mediators of depression. The third concerns the biological mechanisms by which physical treatments exerta therapeutic effect.
at is genetic and whatis environ~ental?
Depression runs in families and so the first question address to is whether the depressed membersof a given familyshare their depression because of their shared genes or because of their shared environment. Monozygous (W)twins have 100%of their genes in common whereas dizygous (DZ) twins have 50% of their genesin common, and so by comparing the concordance rates of MZ with DZ twins it is possible to estimate the heritabilityof a disorder, or the proportion of the variation in phenotype that is accounted for by genetic factors. Nurnbergerand Gershon (1992) have reviewedeight twin studies of of MZ and DZ twins major depressionin which the overall concordance rates were 59.6%and 13.7%. They cited 59% as a typical estimateof the heritability of major depression on the basis of their review. In the case of bipolar affective illnessthe genetic influence is greater and concordance rates of 67% and 20% have been reportedin MZ and DZ twins, respectively (Bertelsenet al., 1977).On the basis of Bertelsen’s study, McGuffinand Katz (1989) have estimated the heritabilityof bipolar affectivedisorder to be-80%. There are many environmental factors, including pregnancy, psychotropicmedi~ation, sleep deprivation and expressed emotion in the social environment which inhence the onsetof illness in an individual who has a genetic predisposition for depressive illness. The ~ e r i t a b i l of i ~unipolar depression is lowerthan that of bipolar and,
depending on the assumptions made in the calculations, was between 48% and 75% for the unipolar depressives in the Maudsley Twin Register (McCuffin et al., 1996). When depression was identified ina community-based volunteer twin register then an even lower heritabilitywas found, but a genetic influence could still be detected (Kendler et al., 1992).
It is possible to determine from twin studies whether or not there is genetic covariation orshared heritability for different disorders. Several twin studies have reported genetic covariation for anxiety and depression including one which wasappropriately entitled ’Major depression and generalized anxiety disorder: same genes and (partly)different environments’ (Kendler et al., 1992). Environmental studies reviewed elsewhere in this book suggest that the e n v i r o ~ e n t a ldifference between the triggers for anxiety and depression are that the life events which trigger anxiety involve threat, whereas the life events which trigger depression involve loss, humiliationand entrapment. Twin studies have also shown that approximately half of the association between the personality trait of neuroticism and lifetime risk of depression is due to genetics (Kendler et al., 1993). The samegroup has also shown a common genetic liability to phobia, panic disorder and bulimia (Kendler et al., 1g95), which are genetically unrelated to the liability to depression, anxiety and neuroticism.
The method of linkage analysis has been outstandingly successful in the case of early onset familial Alzheimer’s disease for which linkage has been established with loci on chromosomes 21 and 14 on which chromosomal mutations have beende~onstratedsubsequently (Sham, 1996).omp parable genome-wide searches for linkage have been conductedin multiple affected bipolar families with sufficient rigour for it to be possible to exclude the possibility that a single locus explains the inheritance of bipolar affective illness (Berretini, 1996).Attempts to fit epidemiologicaland genetic data to mathematical models have also argued against the notion that bipolar affective illness isdue to a single geneof major effect (Craddock et al., 1995). At least three, and possibly more, genes are likely to be involved. Although the findings of many linkage studies in affective illness have never been replicated, two findings have been. As much as 25% of bipolar illness in multiply affected families may be linked toa locus near the centromere on chromosome 18 and as much as 20% to a locus on chromosome 216122.3
43
Low se/f-estee~
atic s u ~ ~ aofr the y sychosocial causes of ~ e ~ r e s §asi odescribed ~ in
ence of environmental stress on the development of several animal models of et al., 1993) and immobildepression including learned helplessness (Papolos ity in the forced swim test (de Kloet et al., 1988). In man, also, the central actions of cortisol can cause depression, as in Cushing's syndrome (Jeffcoate et al., 1979; Kelly et al., 1983). It may also be possible to treat depression by i ~ i b i t i n gthe sypthesisof cortisol (~urphy- ears on et al., 1991; O'Dwyer et al., 1995). Forall of these reasons, the corticosteroid response to e n v i r o ~ e n tal stressis a relevantbiological measure forstudies into the biological basis of the environmental causesof depression. Most of the environmental influences on depression canbe modelled in experimental animals and, in most cases, activation of the hypothalamic pituitary adrenal (HPA axis) results. Thus life at the bottom of the baboon social hierarchy involves limited access to food and mates and the intermittent receipt of displaced hostility from frustrated baboons higher up the hierarchy. This situation (which would be rated by social scientistsas a 'chronic difficulty') is associated with increased activation of the HPA axis (Sapolsky, 1989). So, too, is the loss of social rank (Sapolsky, 1990; Jones et al., 1995)~ which in man would be categorized as a life event with severe long-term consequences for the individual. In social psychiatry, vulnerability factors are psychosocial influences which increase the likelihood that an event or difficulty will be followed by depression. Childhood adversity is such a in zltero vulnerability factor, and in many animal studies environmental stress or in neonatal life results in enhanced responsiveness of the HPA axis throughout adult life (Plotsky& Meaney, 1993). Conversely, increased maternal attention (produced experimentallyby a procedure termed'handling') results in reduced corticosteroid responses to stress throughout life (Meaney et al.,1989). In humans, social supportis a protective factor which reduces the likelihood that a given event or difficulty willbe followed by depression. h the experimental animal, social support (providedby the presence of members of the social group) reduces the sizeof the corticosteroid response to stress (Lyons & Levine, 1994). Humanstudies of the effects upon cortisolof social adversity have been less detailed than the animal studies but the results are broadly consistent. The secretion of cortisol is increasedby life events (Willis et al., 1987), particularly in the presence of depression (Galloway & Dolan, 1989). Social support buffers the cortisol response to stress, at least in chil) ~ sexual abuse(a powerful vu~erabilityfactor dren ( ~ u ~et aal.,r 1 9 ~ 2and for subsequent depression) results in lasting neuroendocrine changes which are similar to those seen in depression (de Bellis et al., 1994). For all these reasons the corticosteroid response to envirorunentalisstress a convenient biological model of the enviromental causes of depression (Fig. 3.2). It is a simplified model, and even within the corticosteroid system there are other steroids, the secretion of which is influenced by stress such as
45
~ ~ o t j o n reactivity al
HPA axis reactivity
A schematic summaryof the effectsof psychosocial stress upon the secretion of corticosteroid hormones.
dihydroepiandosterone,which have an antiglucocorticoid action. There are also many neurotransmitter systems such as noradrenaline (NA) and 5-hydroxytryptamine (5-HT) which adapt to acute and chronic stress (Abercrombie Sr: Zigmond, 1995)but which are less accessible to human investigation at present. The model described in Fig. 3.2 should therefore be thought of as one aspectof the biological basis of the environmental causes of depression: the corticosteroid response to stress an excellent is 'marker'of the main environmental causes of depression, it may in some cases 'mediate' the e n v i r o ~ e n t ainfluence l (O'Dwyeret al., 1995)but it is not the whole story. ess c
hildbirth is a very powerful environmental trigger for the onset of bipolar 6 weeks afterchildbirth (Dean illness, with relapse rates of 30"50% in the first et al., 1989). Thewithdrawal of oestrogen and/or progesterone may be the cause, although so far it has not proved possible to prevent such relapses with oestrogen treatment. However, unipolar depression following childbirth can be treatedwith oestrogen, suggestingthat oestrogen withdrawal may be the cause (Gregoire et al., 1996). Gushing's syndrome is a rare endocrine disder which is commonly associated with depression (Kelly et al., 1983). yxoedema is much commoner and is also associated with depression (Cleare et al., 1996). Hypercalcaemia is another rare metabolic disorder which can be associated with depression. Although few depressive illnesses tare rig~ere~
by such endocrine disturbances they of aretheoretical importance and, in the case of resistant depression, can become practically important(see Chapter 20) *
si These fallinto three main groups: 1 changes in limbic forebrain function; 2 changes in cortical activity; 3 neurochemicalstudies. mere is, at present,no unitary hypothesis which satisfactorily links 1-3 and so in this chapter detail will only be given where this is of some clinical relevance.
These are reviewed elsewhere (Checkley, 1992; Holsboer, 1995) and so only the most general statements will be reviewed here. 1 Depression is associated with more widespread and more severe neuroendocrine disturbances than any other psychiatric disorder, implying a greater disturbanceof limbic and hypothalamic function in depression than in other cases. 2 Increased activation of the HPA axis leading to a sustained increase in cortisol secretion is the most common abnormality, being present in: 80% of the most severely depressed (i.e. psychotic) patients; 50% of in-patients with melancholia; and in less than 50% in mildly depressed out-patients. The causes of HPA axis activation in depression are multiple and probably axis mediated by: increased production include an increased central drive of the of the hypothalamic hormones arginine vasopressin and corticotropinreleasing hormone (Raadsheer et al., 1994); impaired negative feedback control of the axis by cortisol (Carroll et al., 1981; Young et al., 1991); and hypertrophy of the adrenal gland (Rubinet al., 1995) (Fig. 3.3). 3 Reduced secretionof t h y r o i d - s ~ u l a t i n hormone ~ (TSH) (Rubin et al., 1987) with consequent blunting ofEthe H response to thyrotropin releasing hormone (TRH) (Loosenet al., 1987) is only slightly less common. There are other changes in the thyroid axis, including subclinical h~othyroidism(Rupprechtet al., 1989) and an increased incidence of antithyroid antibodies emer er off et al., 1985). Growth hormone (GH) responses to most stimuli are impaired in depression, particularly those which involve the alpha, adrenoceptor influence on GH secretion. (for review, see Checkley, 1992).
(Primary sensory andmotor areas)
(Associationareas)
(Temporalpole, caudal, orbitofrontal," anterior insula, cingulate," parahippocampal")
(Septum, s.innominate, amygdala,* piriform cortex, hippocampus*)
Fig. 3.3 A schematic representation of behavioural specialization and neural connectivity in the cerebral cortex and related limbic structures (adapted from Mesulam, 1985). Regions implicated in the pathogenesis of depression by studies cited in this chapter are indicated by asterisks.
Prolactinresponses to indirectly acting 5- T agonistssuchasfenuramine and tryptophan are impaired in depression (Siever et. al., 1984; Cowen & Charig, 1987; O'Keane & Dinan, 199%). Most neuroendocrine changes in depression can be seen to be either the consequence of chronic stress (( )above) or of monoaminergic ~ysfunction )and (5) above) or possiblyof both. 7 Not only is depression uniquely capableof altering ne~roendocrinefunc-
tion but it is also uniquely sensitive to changes in that function. Depression is the only psychiatricdisorder which is causally relatedboth to Gushing’s syndrome (Jeffcoateet al., 1979; Kellyet al., 1983)and to myxoedema (Cleare et al., 1996). Also inthe postnatal period, it can be treated by oestrogen (Gregoire et al., 1996) and the therapeutic response to standard treatment can be enhanced by additional treatment with thyroid hormones (~oodwinet al., 1982). The neuroanato~callocation of the processes described i ypothalamic nuclei such as the ventromedial nuclei, where hormones ares ~ t h e s i ~ and e d the amygdala, which is involved in the activation of the stress response, and the hippocampus, which is involved in its termination. The hippocampus is uniquely endowed with corticosteroid receptors, which are known to mediate the influence of corticosteroid development of several animal models of depression (de Papolos et al., 1993). The hippocampus may therefore be involved in the pathogenesis of depression in Gushing’s syndrome. The hippocampus and related limbic structures are also richly endowed with thyroid and oestrogen receptors which may mediate the effect of oestrogen and thyroxine on depression. Sleep and depressi~n
There is alsoa unique biological relationship between sleep and depression. Sleep disturbance in depressionhas been characterized by Kupfer (1995) as: x a decrease in sleep continuityand duration in unipolar depression; an increased continuityand duration of sleep; 3 changes in rapid-eye-move~ent(REM) sleep in unipolar depression including reduced REM latency, increased first REM period and shift in distribution of REM sleep toward the firsthalf of the night; reduced slow-wave sleep, particularly during the first non-REM sleep increased electroenc~phalogram power in higher wavebands. Some of these changes are seendisorders in such as obsessive compulsive disorder, which is associatedwith depression, but all are more frequent in depression than in other psychiatricdisorders. The two-wayr e l a ~ o n s ~ bep tween sleepand depression isdemonstrated by the observation that among psychiatric disorders depression is uniquely amenable to treatment with sleep deprivation. Total sleep deprivation for one night results in a transient but significant reductionin the severity of depression in about 60% of all cases (Wu & Bumey; 1990). This effect develops progressively over the period of sleep deprivation and then is lost as soon as the patient falls asleep, even for a brief nap (Wu & B m e y , 1990). The involvement of the limbic systemin the
49
action of sleep deprivation in man has been suggestedby a positron ernission ET) study in which antidepressant response to total sleep deccompanied by a reduction in glucose metabolismin limbic regions ( ~ et ual., 1992).
n is a periodic disorder with tendencies both to relapse and remisn the periodicityis re ular then the question arises as to whether it is drivenby an endogenous biological rhythm or clock, or whether the timing is derived from external stimuli which themselves are periodic. The only way to resolve the matter is to study the periodic phenomena in isolation from external cues to time. ne patient with a 48-hour bipolar illness has been studied in such condiin complete sensory isolation and the periodicity of this illness was prehis sleep-wake served. Under the conditions of complete sensory deprivation cycle shortened to19.5 h so that every 19.5 h he woke upat what he thought his bipolar illness, however, ~aintaineda strict 48start of a new day: riodici~ showing that it was drivenby a truly endogenous biological clock (Dirlichet al., 1981). There is also some evidencethat diurnal variation of depressed mood maintains its timingin the absence of clues to real clock time. It remains controversial as to whether the circadian system is altered in ression (~irz-~ustice, 1995) and the authors, data suggest that it is not et al., 1988; Gheckley et al., 1993). However, there is no contro(~ompson versy aboutthe periodic natureof depression and the fact that in some cases this is driven by a truly endogenous biological clock (Dirlichet al., 1981). Such so suggests alar periodicity is not seen for other psychiatric disorders and a unique relationship between depression and biological clocks. Animal studie indicate that the suprachiasmatic nucleusin the hypothalamusis the site of the 'master clock' (Moore, 1995; Ralph &: Hurd, 1995)~and so the evidence that depression is uniquely sensitive to biological rhythms is further evidence that theh ~ o t h a l a m uis s involvedin the pathogenesisof depression.
e neuroendocrine and other physiological changes which I have reviewed thus farpoint to adysfunctio~ in h ~ o t h a l a and ~ c related subcortical limbic system regionsin depressive illness. The following section will give an overview of recent advancesin the functional neuroimaging of depression. It will show that changes in cerebral blood flow best are described in those cortical regions (orbitofrontal and cingulate) with reciprocal c o ~ e c t i o nto s the limbic system. Before changes in brain function can be studied by neuroimaging it 0
must firstbe determined whether or not there are abnormalities in brain struclateral ventricals in elderly de ressed patients were first red Levy(1980) using computed tomograph^. tic resonance imaging (MRI) have confirme that they are associated with widenin 82; Rabins et al., 1991) particularly temporal lobes (Dolan et al., 1986). These chan~eshave elderly depressives but in younger depressives abnormalities in structure &: §chlaepfer, 1 ). have not been consistently detected .thus far (Pearlson Elderly depressives have also been found to have areas of increase ity (termed hyperintensities) in subcortical a and thalamus: thes i ~ i f i c a n c of e these r have ins et al., 1991). § i ~ i l a rsubcortical white ~ a t t e lesions ( also been reportedyoun~er in patients with bipolar affective disorder et al., 1990; §wayze et al., 1990). ~unctionaln e u ~ o i ~ a gini ~~ge ~ ~ e s s i o n
Since neuronalactivity is dependent on the supply o whenever there is an increase in the neuronal activi a circumscribed brain region there willbe an accompanying increasein the regional cerebral blood flow (rGBF) to that part of the brain.This can be measured in in theoxygenation of functional MRI (whichdetectscha (SPET), in both of whicha or by PETorsinglephotonemissiography radioactive tracer is used (Goodwin, 1996). §tate-dependent changesin cortical activity have attracte terest because they control for abnormal any it^ in structure of reports to date have described state-dependent changes in depression in thoqe sysparts of the cerebral cortex with the strongest connections to the limbic tem. The limbic association cortex includes the orbitofrontal cortex, t cingulate gyrus and theparahi~~ocampal cortex, and changes have consi ently been reported in orbitofrontal cortex and anterior cingulate gyrus as et al., 1995) and patients have recovered from depression. In one PET (Rench one SPET (Goodwin et al., 1993) study dru status was well cont tween the depressed and recoveredstate, a in both experiments reduced in orbitofrontal cortex and anterior cingulate gyrus in the depressed state and increased with clinical recovery. Changes in rCBF in the same areas but in the opposite direction have been reported following electro~onvulsive et. al., ~ 9 9 4total ) ~ sleep deprivation (Ebert et al., 1994; VVu therapy (ECT; Lerer et. al., 1992) and following drug treatment (Drevets et al., y p ) , but in none of these studies were the treatment conditionsof the depressed and recovered patients well matched. Further studies are needed, but at present it would
appear thatactivity in some limbic association areas is altered in depression and changedby a variety of antidepressantt r e a ~ e n t s . ~omparisonsbetween depressed and control subjects, although not controlling for any structural change in depression, do reveal broadly similar findings as has been reviewed recentlyby Goodwin (1996).Reduced cerebral blood flow in anterior brain structures, particularly thei~eromedialfrontal cortex and anterior cingulate gyrus, have been reported in many studies, and reductions in parietal and temporal brain regions have also been frequently reported. The induction of sad mood by the recollection of sad life events has been shown to activate the medial prefrontal cortex in studies using PET (Pa et al., 1993;George et al., 1995) andmagneticresonancei (S. Baker, unpublished data). The generation of depressing pla to attend one’s mother’s funeral) similarly activates the medial prefrontal cortex, whereas the generation of relatively emotionally neutral plans (preparing for one’s mother coming to d i ~ e rdoes ) not (Partiot et al., 1995). One c o ~ o n feature of these two activation procedures is the activation of depressogenic schemata as described in Chapter5. For this reason it has been argued that depressogenic schemata are ‘stored’ in the medial prefrontal cortex. Since the operation of stereotactic subcaudate tractotomy (SST), which is used for resistet al., 1995)~ isolates the medial prefrontal cortex and ant depression (Poynton reduces oxygen metabolism therein (Biver et al., 1995) it is reasonable tosuggest that SST disrupts the storage or processing of depressogenic schemata. The neuroendocrine effects of depression mostly point to a disturbance of brain function at hypothalamic or higher levels. The effects of hormones (corticosteroid, oestrogen and thyroxine) on depression be must mediate their receptors which are mostly localized in subcortical limbic forebrain s ~ c t u r e ssuch , as the hippocampus, amygdala and h~othalamus.h animal studies, hippocampal glucocorticoid receptors are clearly involved in animal models ofdepression,The first gention of functional neuroimaging studies have shown that the brain r ns which show state-depen~ent changes in depression are mostly within the limbic association cortex. The functional relationship between these different areas has been described by Mesu1a.m (1985) as described in Fig. 3.3.
It is not possible to review in a short chapter the very large numberof studies which have been conductedinto the neurochemistryof depression over the last 30 years. Detailed reviews are available elsewhere of the many changes in 5-hydroxytr~tamine(5-HT; Maes & Meltzer, 1995) and the fewer but still important changesin noradrenaline (NA; Schatzberg & Schild~raut,1995) as
well as the good reasons for implicating dopamine (Willner, 1995)~ acetyl choline (Janowsky & Overstreet, 1995) and neuropeptides (Holsboer, 1995; Plotsky et al., 1995)in the pathogenesisof depression. For none of these neurotransmitters is there a single hypothesis that explains satisfactorily the pathogenesis of depression. In the case of the monoamines NA and 5there is now good evidence that depletion of NA and yHT, respectively, does not result in depression (Delgado et al., 1991; Miller et al., 1996a). However, as will be reviewed in the next section, both proceduresdo reverse the therapeutic action of antidepressant drugs which are thought to act by increasing NA and 5-HT neurotransmission.
S
This section will give a simplified account of current understanding of the mechanisms of action of antidepressant drugs such as is needed by a clinician who prescribes such drugs, and needs to know the scientific basis for their actions and interactions. Thirty years ago when tricyclic antidepressants (TCAs)and monoamine oxidase ~ i b i t o r s ( ~were 0 1 s )introduced into clinical practice, it was thought that they alleviated depression by increasing monoamine availability as a result of uptake inhibition and monoamine oxidase inhibition, respectively. However, it was soon appreciatedthat there is a discrepancy betweenthe time courseof drug action upon monoamines (which is immediate) and upon mood (which is delayed by 3-6 weeks). Fifteen years later it was discovered that adaptive changes at monoamine receptors were seen following chronic antidepressantt r e a ~ e n tand , so it was proposed that these effects might explain the delayed onsetof antidepressant action. The down-regulation of beta adrenoceptors (together with uncoupl~g from its second messenger) was the first adaptive change to bereported following treatmentwith MAOIs, TCAs and ECT, and so this was proposed as the common mechanismof action of antidepressantdrugs (Sulser, 1984). Over the ensuingyears many more receptors have been described, many drug more effects upon them have been reported and attempts have been made,both in preclinical and clinical studies, to describe the net effect upon monoaminergic neurotransmission of the multiple effects of antidepressant drugs upon monoaminergic receptors. The wheel has now come full circle and once again it is thought that antidepressant drugs work by increasing monoaminergic neurotransmission. Adaptive chan esat several receptors playan important part in this effect, and so an understanding of how at least some of these receptors function is needed by any clinician who is trying to understand how antidep~essantmedication works.For everyday purposes a working understanding of the serotonergic and noradrenergic systems suffices, and so
3
this chapter will focus almost exclusively on the tidep effects press of ant drugs upon transmitter systems for 5-HT and NA.
5 - H y d r o x y t ~ ~ t(5-HT) ~ ~ i ~and e 5-HT r e c e ~ ~ o ~ s
The cell bodies of 5-HT neurones are located in the raphe nuclei of the hind brain from where projections arise toall brain regions. There areat least 14 receptor subtypes (Watson & Girdlestone, 1996) but so far a role in human psychopharmacology has only been established in four cases. These receptors fall within three families of receptors: 5-HTl, 5-HT2 and5-HT3.Members of of receptors havein comrnon couplingof the receptor to the the 5-HT1 family enzyme adenylate cyclase; the consequence of this is that the stimulation of 5HTl receptors results in inhibition of adenylate cyclase activity, with a subsequent fallin intracellular concentrationsof the second messenger cyclic AMP. The 5-HT2 family of receptors are coupled to the phosphoinositide cycle. The consequence is that stimulation of the 5-HT2 receptors results in increased intracellular concentrations of the second messengers diacylglycerol and inositol triphosphate. 5-HT3 receptors are linked to the fast cation channel in the cell membrane. Stimulation of 5-HT3 results in a rapid change in intracellular calcium and potassium concentrations. Of the 5-HTl family of receptors 5-HTl, receptors are of particular relevance to human psychopharmacology. They bind anxiolytic drugs such as buspirone, gepirone and ipsapironeas well as the beta antagonist pindolol, all of which can be given to human subjects. 5-HTlA receptors are located postsynaptically and are abundant in the hippocampus. Receptors are also located in the cell bodies and dendrites of the 5-HT neurones in the raphe nucleus as sornatodendritic receptors. The stimulation of somatodendritic 5HTlA receptors results in inhibitionof 5-HT cell firing and consequent 5-HT a negative release. Inthis way the 5-HTl,somatodendritic receptor maintains feedback control over 5-HT release. The presynaptic 5-ml-like receptor which is found on 5-HT axon terminals is the 5-HTlB receptor in the rat and the 5HTl,receptor in humans; in human psychopharmacology the 5-HTl, agonist sumatriptan is effective in the relief of migraine throughits action at vascular 5-HTl, receptors. The 5-HT behavioural syndrome refers toa pattern of motor behaviour which is seenin experimental animals following maximal stimulation of the 5-HT system with drug combinations such as the 5-HT precursor tryptophan and a MAOI. The syndrome, which includes hyperactivity, tremor and stereotyped motor behaviours (Straub tail, forepaw treading, hind limb abduction, lateral head shakes), involves activation of 5-HT1, receptors since it
is reversed by 5- TIAantagonists ( Green, 1 8). The term '5-HT syndrome' is applied in clinical practice to somewhat c rable behaviours which are seen follow excessive stimulation of th system in man includin~confusion, tation, myoclonus, hyper-reflexia, ~iaphoresis, mor, diarrhoea, incoordination and fever (Sternbach, 1991).
to be associated with geneticvulnerabili~towards depression.
*
affect 5-HT cause an immediate in Artigas, 1991; Sharp et al., 199 ts in the stimulation of the 5a negative feedback controlby 5Chaput et al., 1986; of increased 5-HTa
5
However, with chronic antidepressant treatment an adaptive downregulation of overof the somatodendritic5-HTlAreceptor takes place and has the effect coming this initial negative feedback. Asa result, 5-HT cell firing and 5-HT release return to pretreatment values; the effect of uptake blockadeby SSMs is then to increase 5-HT neurotransmission (Blier et al., 1988).After chronic TAD treatment an increased responsiveness of postsynaptic 5-HTlAreceptors is seen (Blieret al., 1986,1987, 1990)together with downregulationof 5-HT2, receptors (Goodwin et al., 1984). Since these changes occur in the absence of change in 5-HT cell firing, the net effect is an increase in 5-HT neurotransmission through postsynaptic5-HTlAreceptors. An enhanced reelecsponsiveness of postsynaptic 5-HT2, receptors is also seen after chronic troconvulsive shock (EGS) (Keller et al., 1981, Biegon Bt Israel, 1987; Burnet et 5-HT neurotransmission since5-HT al., 1995)~which also results in increased cell firing and 5-HT release is unchanged following chronic EGS (Blier Bt Bouchard, 1992). A net increase in 5-HT neurotransmission is also seen following treatment with MAOIS as a result of downregulation of 5-HT1, somatodendritic receptors, and also of the alpha2 adrenoceptor which is located on 5-HT terminals and which inhibits5-HT release (Mongeau et al., 1994). Although these different treatments have different acute and chronic effects upon 5-HT, in all cases, the net effect of chronic treatment is an increase in 5-HT neurotransmission in experimental animals. Efects upon 5-HT of l i t h i u given ~ inc o ~ ~ i n a t iwith o n a~tidepressantdrugs.
Lithium inhibits the functioningof the second messenger systems adenylate cyclase and phosphatidyl inositol turnover andas a result has many effects upon neuroreceptor function, including increased responsiveness of postsynaptic 5-HTlAreceptors (Goodwin et al., 1986; Blier et al., 1987). Lithium et al.,1991).The also increases5-HT release following acute treatment (Sharp combination of lithium with antidepressant drugs results in greater increase in 5-HT release than is seen with either treatment alone and this may explain the basis of augmentation by lithium of antidepressant drug treatment(see below). ~nteractions~etweenr-HTl, antagonists anda n t i d ~ r e s s adrug ~t treat~en~
The abilityof SSM treatment to reducecell firing in the raphe nucleus can be reduced by the selective blockadeof 5-HT1, receptors (Arboreliuset al.,1995; Gartside et al.,1995)~which suggests that the effect of SSRIs on 5-HT function can be increased by concurrent autoreceptor blockade(see below).
Chapter 3: ~ i o l o g i c a l ~ o of d ede~yessi~n ls Eflects ofs-HT, agonists on 5-HT
5-HTl, agonists such as gepirone and buspirone are effectiveas anxiolytics (Robinson et al., 1990) and possibly as antidepressants (Jenkins et al., 1990). They have the same effect on 5-HT cellasfiring SSRIs and MAOIS.With acute treatment, cell firing is inhibited and with chronic treatment cell firing returns to baseline value as a result of downregulation of the 5-HT1, somatodendritic autoreceptors (Blier & de Montigny 1990). The net effect is an increase in 5-HT neurotransmission. Effects ofantide~ressant trea~ment on S-H T~nction in man
Models have been developed for the clinical investigation of the above pharmacological mechanisms and many havebeen demonstrated to occur in man. Neuronal 5-HT uptake is conveniently modelledin platelets and platelet 5HT content is reduced following the chronic administration of the SSRI fluvoxamine (Celadaet al.,1992). The hypothermic response to 5-HTl, agonists, which is thought to be a measureof 5-HT1, autoreceptor function, is reduced et following chronic treatment with a number of antidepressant drugs (Lesch al., 1990,1991). Neuroendocrine responses to 5-HT1, receptors are thought to involve postsynaptic 5-HT1, receptors and but some not all of these responses have been seen tobe reduced in depression (Lesch, 1992; Moelleret al., 1994) and following antidepressant drug treatment (Lesch et al., 1991). Effects ofantide~ressanttreatment on neuroendocrine measuresof $-HT neurotransmission in man
The prolactin response to the 5-HT precursor tryptophanbeen has generally found tobe reduced in depression and increased by chronic depressant treatment (Charney et al., 1984; Price et al., 1985, 1989; Gowen et al., 1990). The prolactin response to the indirect 5-HT agonists fenfluramine and d-fenfluramine have also been reported tobe reduced in many though not all, studies of depressed patients. Both responses are increased following chronic treatment with TCAs and ECT (Upadhyayaet al., 3991; Charney & Delgado, 1992; O'Keane et al., 1992; Shapira et al., qcpa,b). On balance the evidence suggests that, at least in neuroendocrine systems, 5-HT neurotransmission is increased in humans treated clinically with TCAs, SSRIs and probably also MAOIs. elations skip between drug eflects upon 5-HT and antidepressant response
There is good evidence that interfering with antidepressant effects upon
terferes with antidepressant response. This was first demonstrated when the antidepressant effect of imipramine was rever d by ~ e a ~ ewith n t depleting agentparachlorophenylal~nine(§h in et al., 1975, 1976). ore recently, a more acceptable formof 5- T depletion has been produced by the administration of drinks rich in amino acids with and without tr~tophan-the 5-HTprecursor. The amino acid load stimulates the synthesis of protein which requires the essential amino acid tryptophan. When tryptophan is suppliedinthedrinkthetryptanisavailable, but inthe tryptophan-free condition the tryptophan is hdrawnfrom the body’s t r ~ t o p h a nstores and since these are needed for the synthesis of 5synthesis is reduced. Animal studies have demonstrated falls in brain concentrationsoftryptophanand5-HT,andthe5T metabolite5hydroxyindoleacetic acidin rats treated withtr~tophan-freediets (Young et al., 1989). Gomparable fallsin plasma tryptophan have been reported in human subjects given the tryptophan-free amino acid drinks (Delgado et al.,
.
epressed patients who have responded to antidepressant treatment retemporarily withina few hoursof ingesting tryptophan-free amino acidrich drinks (Delgado et al., 1990). This effect has been demonstrate^. for all classes of antidepressant drugs with selective action upon 5-HT uptake (i.e. TGAs, SSRIs and MAOIs), but interestingly not for selective inhibitors of noradrenaline uptake such as desipramine and lofepramine (Heninger, 1995). Furthermore, when~yptophan-freedrinks are given to untreated, depressed patients, no deterioration of their depression results (Delgado et al., 1991); this meansthat 5-HT depletionas produced by ~ptophan-freedrinks is not a sufficient causeof depression. However, the studies that have been reviewed e good evidence that the antidepressant effects of TGAs, SS S involve an increase in 5-HT neurotransmission. ence for a causal link between antidepressant response and neurotransmission comes from reports that the antidepresu r o ~ a n s ~ s s i can o n be increased sant response toa drug that enhances5by e n h a n c ~ gthedrugeffectupon 5 hiswasfirstshown by enhanc01, tranylcypromine, with the ing the antidepressant response to tryptophan (Pare, 1963; Glassman & ,1969). The tryptophan-~AOI combination,whichinanimals is a potentstitothe5-HTsyndrome, is a more potent antidepressant treatm I alone. More recent indirect evidence suggeststhat the augmentation of the antidepressant effectof antidepressant drugs by lithium may depend 0n enhanced 5-HT neurothat pindolol transmission (Gowenet al., 1991). Finally, there are suggestions may accelerate the antidepressant response to SSMs by removing 5-HT1, autoreceptor-mediated negative feedback inhibition of the 5-HT neurone (Artigas et al., 1994). The presentstate of knowledge suggeststhat enhance*
5
ment of 5-HT release is a common action of TCAs, SSRIs and MAOIs and a sufficient cause of antidepressant response. However,it is not a necessary component of the response because antidepressant effects can be obtained also by selective in~bitionof uptake.
~ o r a d r e ~ and ~ l i its ~ ereceptors
The noradrenergic system is similar in many respects to the Serotonergic and SO will be described in less detail. Noradrenergic projectionsto the cerebral cortex and limbic forebrain structures arise from cell bodies in the locus coeruleus in the hind brain. The system that is activated in response to the arousal is thought tobe involved in psychological processes suchas the mediation of anxiety, selectiveattention, learning and reinforcement (Robbins & Everitt, 2995). The receptors at which NA (and adrenaline) exerts its physiological effects are termed adrenoreceptors. Their subdivision into alpha and beta adrenoceptors was proposed on the basis of the opposing actionsof the chemically related compounds adrenaline and NA on the same physiological processes (Ahlquist, 2948). Subsequently, alpha, adrenoceptors were distinguished from alphal adrenoceptors on the basis of the presynaptic location of alpha2 adrenoceptors, although it was later recognized that alpha2 adrenoceptors are located both pre- and postsynaptically (Fillenz, 1990). Clonidine is a clinically useful drug which can be used to probe the function of the alpha, adrenoceptors in depressed patients (see below). The subdivision of beta adrenoceptors was made on the basis of the preferential activity of different beta adrenoceptor antagonists for cardiac and vascular of these adrenoceptors tissue (Landset al., 1967). Now that the genes for each have been identified, it is possible to study their structure and molecular biology. More recentlya new classof receptor-the imidazoline receptor-has been described, which also regulates the release of NA. Imidazoline receptors have high affinity for imidazoline, clonidine and also for the alpha2antagonist yohimbine, and a recently described endogenous ligand agmatine(Li et al., 1994). Imidazoline receptors have been subcategorized into I1m d I, receptors on the basis of their preferential affinity for clonidine and imidazoline, II and , I receptors inhibit NA release respectively (Michel Eisberg, 1992). at N A terminals. The noradrenergic system has receptors located on glia; both are present on human platelets and so are available for clinical research.
59
E ~ e c tof s antide~ressantdrugs u ~ o the n NA system inex~e~imental animals
The net effectsof antidepressant drugs upon NA and 5-HT neurones are surprisingly similar. For both neurotrans~tters(see Table 3.1) the immediate consequence of uptake inhibition is increased transmitter availability, and in both cases this results in stimulation of somatodendritic autoreceptors (alpha, adrenoceptors at NAneuronesand 5-HTl, autoreceptors on 5-HT neurones), which resultsin reduced rates of cell firing and reduced transmitter release. With both types of somatodendritic receptordownre~lationoccurs following long-term treatment with selective re-uptake inhibitors, as a result of which cell firing returns towards pretreatment values. The main differences between the adaptive responsesof the NA and 5-HT systems concern changes in postsynaptic receptor function. Concerning the noradrenergic sys tem, dow~egulationof postsynapticbetal adrenoceptorsand/or uncoupling of betal adrenoceptors from the second messenger system is seen following chronic treatment with almostall antidepressant treatments includingECT (Sulser, 1984). Although studied in less detail, do~nregulationof both pre- and postsynaptic alphaz adrenoceptors has been found following chronic antidepressant treatments, as has upregulation of postsynaptic alphal adrenoceptors (Heninger & Charney 1987). Alpha, and imidazoline I, receptors on NA terminals are both downregulated by chronic anti~epressanttreatment (Garcia-Sevillaet al., 1996). E ~ e c tof s a n t i ~ ~ r e s s adrugs n t u ~ o NA n in man
h general, the effects of antidepressant drugs upon NA in man are equivalent
to those that are seen in animals. Various measures of the metabolismof NA can be made in depressed’patients treated with antidepressant drugs and these indicate a reduction in the metabolic turnoverof NA in man (Golden et al., 1988; Sharma et al., 1990). The alpha2 and imidazoline Il receptor agonist clonidine has measurable effects upon blood pressure, sedation, te~perature and the secretion of CH. The effects of antidepressant drugs upon these alpha, adrenoceptorand/or imidazoline Il receptor-mediated responses are similar in animals and man, suggesting that chronic treatment with antidepressant drugs downregulates alpha, adrenoceptors in experimental animals, healthy human volunteers and depressed patients (Checkley et al., 1986). ~ow~egulatio ofnalpha2 adrenoceptors and imidazoline Il receptors has been et al., 1986, demonstrated in the platelets of depressed patients (Garcia-Sevilla et al., 1996), 1996) andin brain regions such as the frontal cortex (Garcia-Sevilla et al., 1994). hippocampus (Gonzalei et al., 1994) and locus coeruleus (Ordway From a clinical perspective the most important question is the determination of the net effect of antidepressant drug treatment on noradrenergic
60
of ~ ~ r e s s i o n le 3.1 This table shows the acute( ~ i n u t e s and ) chronic (7-14 days) effects of selective serotonin re-uptake inhibitors (SSRIs) and selective noradrenaline re-uptake inhibitors (SNRIs) on 5-HT and NA cell bodies in the raphe nuclei and locus coeruleus, respectively. (For the effects of SSRIS and TADs on pre- and post-synaptic receptors, please see text.) Acuteeffectsof SSRIs andSNRIs, respectively, on5-HT and NA neurones
Chronic effects of SSRIs and ti) SNRIs, respectively, on5-HT and NA neurones
5-HT
NA
(i) increased 5-HT availability (ii) stimulation of 5-HT1, somatodendritic receptors (iii) reduced 5-HT cell firing and reduced 5-HT release 5-HT availability returns to baseline value
(9 increased NA availability
do~regulationof 5-HT1, somatodendritic receptors firing and (ii) cell 5-HT release return to pretreatment values (iii) 5-HT uptake is still inhibited (iv> is (iv) 5-HT availability therefore increased
(ii) stimulation of alpha, somatodendritic receptors (iii) reduced NA cell firing and reducedNA release ( W NA availability returns to baseline value (i) downregulationof alpha, somatodendritic receptors (ii) cell firing and NA release return to pretreatment value (iii) NA uptake isstill inhibited NA availabilityis therefore increased
neurotransmission. Someof the adaptive changes seen in animals will result in increased noradrenergic neurotransmission (downregulation of alpha2 autoreceptors and upregulation of postsynaptic alphal adrenoceptors), but others will have the effect of reduced noradrenergic neurotransmission (downregulation of postsynaptic betal and alphaz adrenoceptors).How can the net effect of antidepressant drug treatment on noradrenergic neurotr~smission be determined in man? The secretionof the ineal hormone melatoninis appropriate for this purpose, since the pineal glandis innervated by noradrenergic fibres which are endowed with prejunctional alpha2 adrenoceptors and pos~u~ctional alphal and betal adrenoceptors. Eachof these receptorshas been shown to influence the secretionof melatonin in animals andin man (CheckleySs Palazidou,1988). The secretion of melatonin is therefore controlledby the output of a typical noradrenergic system. The critical question is whether this is increased or reduced by chronic antidepressanttreatxnent in man. Most studies have reported an increase (Checkley,1991).One particularly informative experiment involved healthy volunteers receiving two stereoisomers of the highly selective noradrenaline uptake inhibitor oxaprotiline. The isomers differ in mainly their ability to inhibit noradrenaline uptake. The isomer that inhibited NA
62
increased the secretion of melatonin after acute and chronic treatment, whereas the isomer with no effect on NA uptake had no effect on melatonin. Similar effects upon melatonin (and so presumably upon NA) have been found in healthy volunteers (Thompson et al., 1985) and depressed patients (Sack C4s Lewy, 1986; Palazidou et al., 1992) treated with the selective NA uptake inet al., 1986). hibitor desipramineor with other antidepressant drugs (Murphy Is the e~ect o~antide~ressant drugs on NA causally rela~edto antid~ressantresponse?
It will be recalled that whereas tryptophan depletion reverses the antidepressant effectof SSNs, tryptophan depletion has no effect on the antidepressant effect of relatively selectiveNA uptake inhibitors (Delgadoet al., 1990). There is one report that the antidepressant effect of desipramine canbe inhibited by inhibiting the synthesis of NA with alpha-methyl-para-tyrosine (AMPT; Delgadoet al.,1993)~ which suggests that the antidepressant response to desipramine depends onits ability to increase noradrenergic neurotransmission. AMPT given alone had no effect on drug-free depressed patients (Miller et al., 1996a, 1996b). To date there is no evidence that increasing the effect of antidepressant drugs on noradrenergic transmission increases antidepressant response. At present it can be concluded that chronic treatment with selective NA uptake inhibitors results in a sustained increase in noradrenergic neuro~ansmissionin man, and that this effect be may causally related to the mechanism of antidepressant action. As is the case for 5-HT, there is no evidence that an NA deficiency causes depression, even though there is evidence from neuroendocrine and other studiesof NA changes in depression.
For the purposes of the everyday treatment of depressed patients it is not necessary to review the many other neurochemical effectsof antide~ressant drugs. However, one whichis of interest andhas relevance to otherparts of this chapteris the recent finding that antidepressant drug treatment increases the number of corticosteroid receptors in brain regions, particularly the hippocampus (Seck1 &: Fink, 3992). As a result of corticosteroid receptors in the hippocampus being involved several in animal modelsof depression (de Uoet et al., 1988; Papoloset al., 1993)~the effects of antidepressants on these receptors may be relevant to the mode of action of antidepressant drugs. Corticosteroid receptors modulate 5-HT and NA neurotrans~ssionthrough the hippocampus, and antidepressant drug effects on corticosteroid receptors require the presence of intact 5-HT pathwaysto the hippocampus. The view
62
orders: new and recurrent findings. Molecular Abercrombie, E.D. & Zigmond, M.J (1995) . Modification of central catecholaminergic systems by Psychiat~, I, 172-173. stress and injury. In: P s y c h o p h a r ~ a c o l oThe ~: Bertelsen, A., Hawald, B. & Gauge, M. (1977) A Danish twin study of manic-depressive dis~ o u r Generat~on ~h of Progress (eds F. E. Bloom & D. J. Kupfer), pp. 351-361. Raven Press: New orders. ~ritish~ournal of Psychiatry, 130,330-351. York. Biegon, A. & Israel, M. (1987) Quantitative Adel1,A. & Artigas, E (1991) Differential effects of autoradiographic analysis of the effects of electroconvulsive shock on serotonin-;? receptors clomipramine given locally or systemically on in male and female rats. ~ o u r ~ofa lNeuroextracellular 5-hydroxytryptamine in raphe nuclei and frontal cortex. An in vivo brain che~zistry, 48,1386-1391. microdialysis study. Naunyn Sch~iedeberg's Biver, F., Goldman, S., Francois, A. et al. (1995) Changes in metabolism of cerebral glucose Archives of P k a r ~ a c o l o3~ , Ahlquist, R.P. (1948) A study of the adrenotropic after stereotactic leucotomy for refractory receptor. A ~ e r i c a n~ournalof ~hysiology/153, obsessive-compulsive disorder: a case report. 586-600. ~ournalof N e u r o l o ~ , ~ e u r o s and u r gPsychiat~, e~ Arango, V.,Underwood, M.D. & Mann, J.J. (1992) 58,502-505. Alterationsin monoamine receptors in the brainBlier, P. & Bouchard, C. (1992) Effect of repeated of suicide victims. ~ournalof Clinical Psychoelectroconvulsive shocks on serotonergic neup k a r ~ a c o l o12,8-12. ~/ rons. European ~ournalof P h a r ~ a c o l o g 211, ~, Arborelius, L., Nomikos, G.G., Gulhner,P. et al. 365-373. (1995)~ a u n y n S c h ~ i e d e bArchives e r g ~ s of P h a r ~ a - Blier, P. & de Montigny, C. (1990) Differential effect of gepirone on presynaptic and postsynaptic Arora, R.C. & Meltzer,H.Y. (1989) Increased serotonin receptors: single cell recording studserotonin 2 (5-HT2) receptor binding as measies. ~ournalof Clinical Psychophar~acology/10 ured by 3H-lysergic acid d i e ~ y l a m i(3H-LSD) ~e (Suppl. 3), 135-205. S of depressed patients.Life Blier, P., de Montigny, C. & Azzaro, A.J. (1986) ~odification of serotonergic and noradrenergic Artigas, F., Perez, V.& Alvarez, E. (1994) Pindolol neurotransmission by repeatedadminis~ation induces a rapid improvement of depressed of monoamine oxidase inhibitors: electropatients treated with serotonin reuptake inphysiological studies in threat.~ o u r ~ofa Pharl hibitors. Archives of General Psychiatry/51,248~ a c o l o g yand E x p e r i ~ e ~ t a l ~ h e r a p e u t237, ics, 251. 987-994. de Bellis, M.D., Chrousos, GP., Dorn, L.D. et al. Blier, P., de Montigny, C. & Tardif, D. (1987) Short(1994) Hypothalamic pituitary adrenal axis term lithium treatment enhances responsivedysregulation in sexually abused girls with his-ness of postsynaptic 5-HTI, receptors without tories of abuse. Journal of Clinical E n ~ o c r i ~ o l o ~altering 5-HT autoreceptor sensitivity: an and Metabolis~'78,249-250. electrophysiological study in the rat brain. Bench, C.H., Friston, K.J. &Brown, R.G. et al. (1993) Synapse, I, 225-232. Regional cerebral blood flow in depression Blier, P., Chaput,Y. & de Montigny, C. (1988) Longmeasured by positron emission tomography: theterm5-HTreuptakeblockade,butnot relationship with clinical d~ensions. Psychologimonoamine oxidase inhibition, decreases the cal Medicine/ 2 3 , 5 7 ~ 5 9 o . function of terminal 5-HT autoreceptors: an Berretini,W. (1996) Genetic studies of bipolar dis- electrophysiological study in the rat brain.
~~~~e~ 3: ~ i o l o ~ i c ~ of ~ ~ e ~~ o~~e es lssi o ~ Naunyn Schmiedebers’sArchives of Ph a r ~ a c u l u a , Checkley SA (1991) Neuroendocrine effects of psychotropic drugs. Clinical Enducrinulu~and 3371 246-254. Blier, P., de Montigny, C. & Chaput, U. (1990) A Metabulis~,5, 15-33. role for the serotonin system in the mechanismCheckley, S A . (1992) Neuroendocrinology. In: of action of antidepressant treatment preclinical ~ a ~ ~ bufuAffective uk Disorders (ed. E.S. Paykel), evidence.~uurnal of Clinical Psychiat~, 51,(Suppl. pp. 255-266. Churchill Livingstone, London. Checkley, S.A.(1998) The managementof resist4), 14-20. Bridges, P. (1992) Resistant depression and ant depression.In:The Ma n a s e ~e noff D ~ r e s s ~ u n psychosurgery. In: ~andbuukof Affective Disur(ed. S.A. Checkley), pp. 430-457. Blackwell ders (ed. E. S. Paykel), pp. 437-452. Churchill Science, Oxford. Livingstone, Edinburgh. Checkley, S.A. & Palazidou, E. (1988) Melatonin Brugha, T.S. & Bebbington P., McCarthy, P. et al. and antidepressant drugs: clinical pharma(1987) Social networks social support and the cology. In: Melatunin: Clinical Perspectives (eds type of depressive illness. Acta Psychiatrica A. Miles, D. R. S. Philbrick, C. Thompson), Scandinavica,76,664-673. pp. 190-204. Oxford University Press, Oxford. Burnet P.W.J., Mead, A. & Eastwood, S.L. et al. Checkley, S.A., Corn, T.H., Glass, LB. et al. (1986) (1995) Repeated ECS differentially effects rat Neuroendocrine and other studies of the mechabrain 5HT1, and 5HT2, receptor expression. nism of action of desipramine. In:An~~depressants NeuruRepurt, 6,901904. and their Receptor ~ ~ n c t(eds i u ~R. Porter, C. Bock, Calloway,P. & D o h , R. (1989) Endocrine changes S. Clark), CIBA ~uundatiun Sympusiu~, 123, and clinical profiles in depression. In:Life Events 126-147. CIBA Foundation, London. and Illness (eds G.W. Brown & T. 0.Harris), pp. Cleare, A.J., McGregor,A., Chambers, S.M. et al. 139-160. Unwin Hyman, London. (1996) Thyroxine replacement increases central Carroll, B.J., Feinberg, M., Greden, J.F. et al. (1981) 5-hydroxytryptamine activity and reduces A specific laboratory test for the diagnosis of depressive symptoms in hypothyroidism. melancholia. Archives of General Psychiut~,38, ~ e u r o e ~ d u c r i n u64,6549. ~u~, 15-22. Collier, D.A., Stober, G., Li,T. et al. (1997) SuscepCelada, P., Dolan, M., Alvarez, E. et al. (1992) tibility to bipolar affective disorder and unipolar depression is influenced by allelic variation of Effects of acute and chronic treatment with functional serotonin transporter expression. fluvoxamine on extracellular and platelet serotonin in the blood of major depressive Mulecular Psychiu~ry(in press). patients. Relationships to clinical improvement. Cowen, P.J.& Charig, E.M. (1987) Neuroendocrine ~uurnalufAffective Disorders, 25,243-250. responses to intravenous tryptophan in major Chaput, Y.& de M o n t i ~ yC. , (1991)Presynaptic depression. Archives of General Psychiatry, 44, and postsynaptic modifications of the serotonin 95&-966. system by long-term administration of antide- Cowen, P.J., McCance, S.L., Gelder, M.G. et al. (1990) The effect of amitriptyline on endocrine pressanttreatments:an in vivo electroresponses to intravenous L-tryptophan. Psychiphysiologic study in the rat. ~ e u r u p s y c h o atry Research, 31, 201-208. 5, 219-229. pharmaculu~, McCance, S.L., Ware, C.J.et al. (1991) Chaput, Y.,de Montigny, C. & Blier, P. (1986) Cowen, P.J., Lithium in tricyclic resistant depression: correEffects of a selective 5-HT reuptake blocker, lation of increased brain 5-HT function with citalopram, on the sensitivity of 5-HT autoclinical outcome.~ r i ~ i s ~ of ~psych^^^, u ~ r n a159, l receptors: electrophysiological studies in the rat. ~ a u n Sch~eide~erg’s ~ n Archives of P h a r ~ u 341-346. Craddock, N., Khodel, V.,Van Eerdewegh, P. & coloar 3331 342-345Reich, T. (1995) Mathematical limits of multiCharney, D.S.& Delgado, P.L. (1992) Current concepts of the role of serotonin function in depres- locus models: the genetic transmission of bipolar disorder. American ~ournuluf an sion and anxiety. In: Serutunin ReceptorSubtypes: Genetics, 57, 690-702. P ~ r ~ a c u l o g i cSa il ~ ~ ~ aand n cClinical e Implication (eds S. Z.Langer, N. Bmello, G. Racagni, J. Deakin, J.F.W. & Green, A.R. (1978) The effects of putative 5-hydroxytryptamine antagonists Mendlewicz), pp. 89-104. Karger, Basel. on the behaviour produced by administration Charney, D.S., Heninger, G.R. & Sternberg, D.E. of tranylcypromineandL-tryptophan or (1984) Serotonin function and the mechanism tranylcypromine andL-dopa to rats.~ r ~ Juurt i s ~ of action of antidepressant treatment: effects nal uf P h a r ~ a c u l u64,201-209. ~, of amitriptyline and desipramine. Archives of Dean, C,, Williams, R.J.& Brockington, I.F. (1989) General Psychiatry,41, 359-365.
Is puerperal psychosis the same asbipolar manic Garcia-Sevilla, J.A., Escriba,P.V., Sastre, M. et al. (1996) Immunodetection and quantitation of depressive disorder? A familystudy. Psychologiimidazoline receptor proteins in platelets of cal Med~cine,19,437447. patients with major depression and in brains of Delgado, P.L., Charney, D.S. & Price, L.H. et al. suicide victims. Archives of General Psychiatry, (1990)Serotoninfunction and the mechanism of 53,803-810. antidepressant action:reversal of antidepressant induced remission byrapid depletion of plasma Gartside, S.E., Umbers, V., Hajos, M. & Sharp, T. (1995) Interaction between a selective 5-HTrA tryptophan. Archives of General Psychiatry, 47, receptor antagonist and an SSRI in vivo: effects 411-418. on 5-HT cell firing and extracellular 5-HT. Delgado,P.L.,Price,L.H. &Miller,H.L.et al. (1991) British Journalof P h a r ~ a c o l o115,1064-1070. ~, Rapid serotonin depletion as a provocation challenge test for patients with major depression: George, M.S., Ketter, T. & Parekh, P. et al. (1995) Brain activityduring transient sadness and haprelevance to antidepressant action and the neurobiologyof depression.P s y c h o p h a r ~ a c o l o ~ piness in healthy women. A ~ e r i c a n~ournalof Psychiatry, 152,341-351. Bulletin, 27,321-330. Delgado, P.L., Miller, H.L. & Salomon, R.M. et al. Glassman, A.H. & Platman, S.R. (1969) Potentiation of monoamine oxidase inhibitor by (1993) Monoamines and the mechanism of tryptophan. Journal of Psychiatric Research, 7, antidepressant action: effectsof catecholamine 83-88. depletion on mood of patients treated with Golden, R.N., Markey, S.P. & Risbey, E.D. et al. o~ 3, antidepressant. P s y c h o p h a r ~ a c o l Bulletin, (1988) Antidepressants reduce whole blood 389-395. norepinephrine turnover while enhancing 6Dirlich, G., Kammerloher, A. & Schulz, H. et al. hydroxymelatonin output. Archives of General (1981) Temporal co-ordination of rest-activity Psychiatry, 45,150-154. cycle. Body temperature, urinary free cortisol and mood in a patient with 48-hour unipolar de- Gonzalez,A.M., Pascual,J., Meana, J.J. et al. (1994) Autoradiographic demonstration of increased pressive cycles in clinical and time-cue-free a-2 adrenoreceptor agonist binding sites in the environments. Biological Psychiatry,16,163-179. hippocampus and the frontal cortexof depressed Dolan, R.J., Calloway, S.P., Thacker, P.F. & Mann, suicide victims. Journal of Neuroche~istry,63, A.H. (1986) The cerebral cortical appearance in 256-265. depressed subjects, Psychological Medicine, 16, Goodwin, F.K., Prange, A.J. & Post, R.M. et al. 775-779. (1982) Potentiation of antidepressant effects by Drevets, W.C., Videen,T.O. & Price, J.L.et al. (1992) L-triiodothyronine in tricyclic non-responders. A functional anatomical study of unipolar deA ~ e r i c f fJournal n of Psychiatr~,~39~34-38. pression. Journal of Ne~roscience,12,3628-3441. Dupont, R.M., Jernigan, T.L. & Butter, N. et al. Goodwin, G.M. (1996)Functional imaging, affective disorder and dementia. British Medical (1990) Subcorticalabnormalities detected in biBulletin, 52,495-512. polar affective disorder using magnetic reso47, Goodwin, G.M., Green, A.R.& Johnson, P. (1984) nance imaging.Archives of General Psychiatry, 5-HT2 receptor characteristics in frontal cortex 55-59. and 5-HT2receptor-mediated head-twitch beEbert, D,, Feistel, H. & Kaschka, W. et al. (1994) haviour following antidepressant treatment in Single photon emission computerized tomo~, mice. British Journfflof P h a r ~ a c o l o83,235-242. graphy assessment of cerebra1 dopamine D2 receptor blockade in depression before and Goodwin, G.M., de Souza, R.J. & Wood, A.J.et al. (1986) The enhancementby lithium of the 5-HTIA after sleep deprivation: preliminary results. Biomediated serotonin syndrome in the rat: evilogical Psychiatry,35,880-885. dence for a post-synaptic mechanism. PsychoFillenz, M. (1990) Noradrenergic Neurones. Camp h a r ~ a c o l o90,488-493. ~, bridge University Press, Cambridge. Goodwin, G.M., Austin, M.-P. & Dougall, N.et al. et al. (1995)AsimFlint, J., Corley, R., DeFries, J.C. (1993) State changes in brain activity shown by ple genetic basis for a complex psychological the uptake of 99mTc-exametazine with single photrait in laboratory mice. Science, 269,1432-1435. ton emission tomography in major depression Garcia-Sevilla, J.A.,Guimon, J., Garcia-Vallejo, P. before and after treatment. ~ournalof Afective & Fuster, M.J. (1986) Biochemical and functional ~isorders,29,243-253. evidence of supersensitive platelet a2adrenoreceptors in major affective disorder: effect Gregoire,A.J.P., Kumar,R. & Everitt, B. et al. (1996) Transdermal oestrogen for severe postnatal deof long-term lithium carbonate treatment. Arpression. Lancet, 347,930-933. chives of General Psychiatry,43,51-57.
Gunnar, M.R., Larson, M.C. & Hertsgaard, L.e fal. Major depression and generalized anxiety (1992)The stressfulness of separation among disorder:samegenes(partly)different nine-month-old infants: effects of social context enviro~ents?Archjves of genera^ P s ~ c h ~ a t ~ ~ variables and infant temperament. Child Devel716-722. o p ~ e ~63,290-303. t, Kendler,K.S.,Neale, M C ,Kessler, R.C.et al. (1993) Guze, B.H. & Baxter, L.R.,Jr (1986) The serotonin A longitudinal twin study of personality and syndrome: case responsive to propranolol (letmajor depression in women. Archi~esof G e ~ e r a ~ ter). ~ournalof Clinical Ps~chophar~acology, 6, 119-120. Kendler, K.S., Walters, E.E.,Neale, M.C.el al. (1995) Heils, A., Teufel, A., Petri, S. et al. (1996) Allelic The structureof the genetic ande n v i r o ~ e n t a l variation of human serotonin transporter gene risk factors for six major psychiatric disord expression. ~ o ~ r nofa~l e u r o c h e ~ i s t(in r y press). in women. Arc}~ivesof General Psyc~zi~try, Heninger, G.R. (1995) Indoleamines: the role of 374-383. serotonin in clinical disorders. In: Psycho- Kenny, G.N., Oates, J.L., Lesser, J.et al. (1992) Effip h a r ~ a c o l o The ~ : ~ o u r t hGeneration of Progress cacy of orally adm~istered o~~danse~on in the (eds F. E. Bloom & D. J. Kupfer), pp. 471-482. prevention of postoperative nausea and vomitRaven Press, NewYork. ing: a dose ranging study. ~ r j t i s k~ o ~ of~ n a ~ Heninger, G.R.& Charney,DS. (1987)M e c h ~ i s m Anaesthetics, 68,46&470. of action of antidepressant treatments: implica-de Kloet, E.R., de Kloet,S., Schild, V. et al. (1988) tions for the etiology and treatmentof depres~ ~ g l u ~ o c ~ r t iRU c o38486 i d attenuates retensive disorders. In: P s ~ c h o p h a r ~ a c oThe l o ~ ;Third tion of a behaviour and d i s i ~ i b i t sthe hypoof Progress (ed.H. Y. Meltzer), pp. 535Ge~eration 544. Raven Press, New York. Holsboer, F. (1995) Neuroendocrinology of mood disorders. In: Psychophar~aco~ogy: The ~ o u r t h illness: clinical implications: a tasting menu. of Progress (eds F. E. Bloom & D. J. Genera~~on Kupfer), pp. 957-970. Raven Press, New York. Lands, A.M., Arnold, A., M~Auliff, JP., Luduena, Jacoby, R.J. & Levy, R. (1980) Computed tomoF.P.& Brown,T.G. (1967) Differentiation of graphy in the elderly. 3: Affective disorder. ~ r i t i s~ournal h of Psychiat~/~36,270-275. Janowsky, D.S. & Overstreet, D.H.(1995)The role of acetylcholine mechanisms in mood disorders. neurobiological consequences of ECT implicaIn: Psychaphar~acolo~: The Four~hGeneration of tions for the p a t h o ~ e n e ~and i s treatment of Progress (eds F. E. Bloom & D. J. Kupfer), pp. affective disorders.In: Psycho~harmacolo~: The 945956. Raven Press,New York. Third Geffer~tion of P ~ o g r e ~ (ed. s H. U.Meltzer), Jeffcoate, W.J., Silverstone, J.T., Edwards, C.R.W. pp. 577-588. Raven Press, New York. et al. (1979) Psychiatric manifestations of Lerer, B., Krausz, Y., Bonne, 0. et al. (1994) ECT, Cushing’s Syndrome: response to lowering of an~depressantresponse and cerebral blood flow plasma cortisol.~ u a ~ ~ ~ l y ~ oof ~ u er dn iacli ~4 e , in patients with major depression. europ s y c h ~ h a r ~ a c o10,568s. ~o~, 465-472. Jenkins, S.W. e fal. (1990) Gepironeinthe treatment Lesch, K.P. (1992) 5 - ~ ~ , r e c e p t ores~onsivity r in of major depression.~ournalof Clinical Psychoanxiety disorders and depression. Progress in p h a r ~ a c o ~10~,~(Suppl. y, 3), 77S85S. Neuropsychopharmacology and Bio~ogica~ PsychiJones, I.H., Stoddart,D.M. & Mallick, J.(1995)Toatry, 15723-733. wards a sociobiological model of depression. X , Disselkamp-Tietze,J., Hoh, A. el al. ~ r i t i~ournul s~ of P s ~ c k i a166,475-479. ~~, (1990) 5-HT1, receptor function in unipolar deKeller, K.J., Cascio,CS. & Butler, J.A.et al. (1981) pression: effect of chronic amitriptyline treatDifferential effects of electroconvulsive shock ment. ~ourna2of N e ~ ~ r~a rl ~ n s ~ l a f f t u txi o f f , and antidepressant drugs on serotonin-2 157-161. receptors in rat brain. ~ u ~ o p e ~a n a u r n aof~ Lesch, K.P., Hoh, A.H., Schulte,H.M. et al. (1991) P h a r ~ ~ a c o 69,515-518. lo~, Long-term fluoxetine treatment decreases 5Kelly, W.F., Checkley,S.A., Bender, D.A.et al. (1983) HTI, receptor responsivity in obsessiveGushing’s Syndrome and depression: a prospec-compulsive disorder. P s y c h o p h a r ~ a c o l o105, ~/ tive study of 26 patients.~ r i t i s ~ h o u r ~ofa Psyl 410-420. chiatry, 142,16-19. Li, G., R e g ~ a t h a nS., , Barrow, C.J., Eshraghi, J., Kendler,K.S., Neale, M.C., Kessler, R.C. et dl. (1992) Cooper, R. & Reis, D.J. (1994) Agmatine: an
66
C ~ ~3: ~ i ~o l eo ~r i ~ ~ofl ~ e ~~r oe §~s ieo lns (1986) Human plasma melatonin is elevated endogenous clonidine-displacing substance in during treatment with monoamine oxidase the brain.Science, 263,966-969. inhibitors clorgyline and tranylcypromine but Loosen, P.T., Garbutt, J.C. & Prange, A. (1987) not deprenyl. Psychiatry Research, 17, 119Evaluation of the diagnostic utility of the TRH127. induced response in psychiatric disorders. Murphy-Pearson, B., E., Dhar, V.,Ghadirian, A.M. Pharmacopsychiat~,20,90-95. et al. (1991) Response to steroid hypertension in Lyons, D.M. & Levine, S. (1994) Socioregulatory major depression resistant to antidepressant effects on squirrel monkey pituitary-adrenal therapy. Jo~rnaZof Clinical Psychopharmacolo~, activity: a longitudinal analysis of cortisol XI, 121-126. and ACTH. Psychoneuroendocrinology, 19, 283-291. Nemeroff, C.B., Simon, J.S., Haggerty,J.,J.Jr et al. (1985) Antithyroid antibodies in depressed Maes, M. & Meltzer, H.Y. (1995) The serotonin patients. American Journal of Psychiatry,142,840hypothesis of major depression. In: Psychopharmacolo~:The Fourth Generation of Progress 843. Nurnberger,J.I., Gershon, E.S.(1992) Genetics. In: (eds F. E. Bloom & D. J. Kupfer), pp. 933-944. ~ a n d ~ofo Afective o~ Disorders (ed E.S. Paykel), Raven Press, New York. pp. 131-148. Churchill Livingstone, Edinburgh. McGuffin, P. & Katz, R. (1989) The geneticsof depression and manic depressive illness. British ODwyer, A.M., Lightman, S.A., Marks, M.N. et al. (1995) Treatment of major depression with Journal of Psychiat~,155,294-304. metyrapone and hydrocortisone. Journal of McGuffin, P., Katz, R. & Watkins, S. (1996) A hosAfective Disorders, 33,123-128. pital based twin register of the heritability of DSM-TV unipolar depression. Archives of General Ogilvie, A.D., Battersby,S., Bubb, V.J. et al. (1996) Polymorphism in serotonin transporter gene asPsychiatry, 53,129-136. sociated with susceptibility to major depression. Meaney, A.J., Aitkin, D.H., Viau, V. et al. (1989) Neonatal handling alters adrenocortical negaLancet, 347,731-733. tive feedback sensitivity and hippocampal typeO'Keane, V. & Dinan, T.G. (1991) Prolactin and cortisol responses to d-fenfluramine in major deI1 glucocorticoid receptor binding in the rat. pression: evidence for diminished responsivity Neuroendocrinolo~, 50,597404. of central serotonergic function.American JourMesulam, M.-M. (1985) Patterns in behavioral nal of P s y ch ia t ~, 148,1009-1015. neuroanatomy: associative areas, the limbic OKeane, V., McLoughlin, D.& Dinan, T.G. (1992) systema and hemispheric specialisation. In: Principles of Behavioral NeuroZogy (ed. M,". D-fenfluramine-induced prolactin and cortisol Mesulam), pp.1-70. Davis, Philadelphia. release in major depression: response to treatment. Journal of Afective Disorders, 26, 143-150. Michel, M.C.& Eisberg, P. (1992) Keeping an eye on the I site: imidazoline-preferring receptors. Ordway, G.A., Widdowson, PS., Smith, K.S. & Halaris, A.E. (1994) Agonist binding to a2Trends in PharmaceuticalSciences, 113,369370. adrenoreceptors iselevated inthelocus Miller, H.L., Delgado, P.L., Salomon, R.M. et al. coeruleus from suicide victims. Journal of (1996a) Clinical and biochemical effects of ~eurochemistry,63,617-624. catecholamine depletion on antidepressantPalazidou, E., Papadopoulos, A., Ratcliff, etH.al. induced remission of depression. Archives of (1992) Noradrenaline uptake inhibition inGeneral Psychiatry,53,117-128. creases melatonin secretion, a measure of Miller, H.L., Delgado, P.L., Salomon, R.M. et al. noradrenergic neurotransmission in depressed (199610) Effects of a a-methyl-para-tyrosine (AMPT) in drug free depressed patients. Neuropatients. PsychologicalMedicine, 22,309-315. Papolos, D.F., Edwards, E., Marmur,etR. al. (1993) psychopharmacolo~, 14,15~--157. Moeller,F.G., Steinberg, J.L., Fulton, et M.al. (1994) Effects of the antiglucocorticoidRU 38486 on the A preliminary neuroendocrine study with induction of learned helplessness in Spraguebuspirone in major depression. Ne~ropsychoDawley rats.Brain Research, 615,304-309. Pardo, J.V., Pardo, P. & Raichle, M.(1993) Neural pharmacolo~,10,75-83. Mongeau, R., de Montigny, C. & Blier, P. (1994) correlates of self-induced dysphoria.American Electrophysiologic evidence for desensitization Journal of Ps y c k ia t ~, 150,713-719. of alpha 2-adrenoceptors on serotonin terminals Pare, C.M.B. (1963) Potentiation of monoamine oxidase inhibitors by tryptophan. Lancet, ii: following long-term treatment with drugs increasing norepinephrine synaptic concentration. 527-528. N e u r o ~ s y c h o p h a ~ a c o10,41-51. lo~, Partiot, A., Grafman, J., Sadato, N. et al. (1995) Brain Murphy, D.L., Tamarkin, L., Sunderland,T. et al. activation during the generation of non-emo-
tional and emotional plans. ~ e u r o R e p o r t6, , 1269-1272. Pearlson, G.D. & Schlaepfer, T.E. (1995) Brain imaging in mood disorders: In: Psychophar~ a c o l the o ~Fourth Generation of Progress (eds F.E. Bloom & D.J. Kupfer), pp. 1019-1028. Raven Press, New York. Plotsky, P.M. & Meaney, M.J. (1993) Early postnatal experience alters hypothalamic cortico~opin-releasingfactor (Cm), mRNA, median eminence CRF content and stress-induced release in adult rats. Brain Research. Molec~lar Brain Research, 18,195-200. Plotsky,P.M., Owens, M.J.& NemeroEf, C.B. (1995) Neuropeptide alterations mood in disorders.In: P s y c h o p k a r ~ a c o l oThe ~ : F o ~ r t kGeneration of Progress (eds F. E. Bloom & D. J. Kupfer).Raven
Press, New York. Poynton, A.M., Kartsowis, L.D. & Bridges, P.K. (1995)Aprospectiveclinical study of stereotactic subcaudate tractotomy. Psychological ~ e d i c i n e , 25, 763-770* Price, L.H., Chamey, D.S. & Heninger, G.R. (1985) Effects of tranylcypromine treatment on neuroendocrine, behavioural and autonomic in depressed patients. responses to tryptophan Life Science, 37,809-818. Price, L.H., Chamey, D.S., Delgado, et P.L. al.(1989) Effects of desipramine and fluvoxamine treatment on theprolactin response to tryptophan. Arckives of General Psychiatyy, 46,625-631. Raadsheer, F.C., Hoogendijk, W.J.C., Stam, F.C. et al. (1994) Increased numbers of corticotropinreleasing hormone expressing neurons in the hypothalamic paraventricular nucleus of deo ~436, pressed patients. ~ e u r o e ~ d o c r i n o l60, 444. Rabins, RV.,Pearlson,G.D., Aylward,E. et al.(1991) Cortical magnetic resonance imaging changes in elderly inpatients with major depression. A#erican ~ournalof Ps y c h i a t ~148,617-620. , Ralph, M.R. & Hurd, M.W. (1995)Circadianpacemakers in vertebrates. In: Circadian Clocks and Their A d ~ u s t ~ e n(eds t D. J. Chadwick & K. Ackrill) CIBA Foundation Sy~posiu#,183,67437. CIBA Foundation, London. Robbins, T. & Everitt, B. (1995) Central norepinephrine neurones and behaviour. In: Psychopkar~acology: TheFourth gene ratio^ of Progress (eds F. E. Bloom & D. J. Kupfer), pp.
363-372. Raven Press, NewYork. Robinson, D.S., Richels, K., Feighner, J.et.al. (1990) Clinical effectsof the 5-HTI, partial agonists in depression: a composite analysis of buspirone in the treatmentof depression. ~ournalof Clini3), 67s-768. cal P ~ c ~ o p k a r # a c o l10 o ~(Suppl. ,
6
Rubin, R.T., Poland, R.E., Lesser, I.M.et al. (1987) Neuroendocrine aspectsof primary endogenous depression.IV. Pituitary-thyroid axis activityin patients and matched control subjects. Psycho~euroendocrinolo~, 12,333-347. Rubin, R.T., Phillips, J.J., Sadow, T.F. et al. (1995) Adrenal gland volume in major depression. Arc~ivesof General Ps y c h ia t ~, 52,213-218. Rupprecht, R., Rupprecht, C., Rupprecht, M. et al. (1989)Triio~oth~onine, thyroxine, and TSH rein depressed patients sponse to dexamethasone Psych~atry,25, and normal controls. B~ologica~ 22-32. Sack, R.L. & Lewy, A.J. (1986) Desmethylimipramine treatment increases melatonin B~o~ogical P s ~ c ~ 21, i a ~ ~ , production in humans. 406-409. Sandyk, R. (1986) L-Dopa induced 'serotonin syndrome' in a Parkinsonian patient on bromocriptine (letter). ~ o u ~ nofa lClinical Psychophar~ a c o l o6,194-194. ~, Sapolsky, R.M. (1989) Hypercortisolism among socially subordinate wildbaboons originates at CNS level. Ayckives of General Psyckiatry, 46, 1047-1051. Sapolsky, R.M. (1990) Adrenocortical function, social rank and personality among wild ba~, boons. ~iologicalP s ~ c h i a t28,862-872. Schatzberg, A.F. & Schildkraut,J.J. (1995)Recent studies on norepinephrine systems in mood diso ~~: o u r t hGenorders. In: P s y c k o ~ k a ~ a c o lThe eration of Progress (edsF. E. Bloom&D.J. Kupfer), pp. 911920.Raven Press,New York. Seckl, J.R. & Fink, G. (1992) Antidepressants increase glucocorticoid and, minera~ocorticoid receptors inRNA expression in the rat hippolo~, campus in vivo, ~ e u r o e n d o c y i ~ o 55,621-626. Sham, P. (1996) Genetic epidemiology.Britisk Medical Bulletin, 52,408-433. Shapira, B.,Yagmur, M.J. & Gropp, C. et al. (1992a) Effect of chlomipramine and lithium on fenfluramine induced hormone release in t~, major depression. Biological ~ s ~ c k i a15,975983. Shapira, B,, Lerer, B. & Kindler, S. ef al. (199213) enhanced serotonergic responsivity following electroconvulsive therapy in patients with ~ o ~ r ~ a l major depression. ~ r ~ t ~ s k of Psychiatry/ 160,223-229, Shapira, B., Cohen, J., Newman, M.E. et al. (1993) Prolactin response to fenfluramine andplacebo challenge following maintenance pharmacotherapy withdrawal in remitted depressed 33,531-535. patients. ~iological Psychiat~, Sharma, R.P., Janicak, P.G., Javaid,J, I. et al. (1990) Platelet MAO inhibition, urinary MHPG, and
leukocyte beta-adrenergic receptors in deat~, jects. British ~ournalof P s ~ c ~ i i 147,389-393. pressed patients treated with phenalzine.Ameri- Thompson, C., Franey, C., Arendt, J. & Checkley, 147,1318-1321. can ~ournalof Psychia~ry, S.A. (1988)A comparison of melatonin secretion Sharp, T., Branwell, S.R., Lambert, P. & Grahamein depressed patients and normal subject. British ~oz~rnal of Psyc~iatry,152,260-265. Smith, D.G. (1991)Effects of short- and longterm administration of l i ~ u m on the release Upadhyaya, A.K., Pennell, I., Cowen, P.J. et al. (1991)Blunted growth hormone and prolactin of endogenous 5HT in the hippocampus of the gy, rat in vivo and in vitro. ~ e u r o ~ h a r ~ a c o l o30, responses to L-tryptophanin depression,a statedependent abnormality. ~ o ~ r nofu lAfiecbive 977-984. ~ i s o r ~ e 21,213-218. rs, Shopsin, B., Gershon, S., Goldstein,M.et al. (1975) Use of synthesis inhibitors in defining a role of Watson, S.P.& Girdlestone, D. (1996)TiPS receptor biogenic amines during imipramine treatment and ion channel nomenclature. Trends in PharI, in depressed patients. Psycho~har~acology, macological Sciences,17,1-44. Willis, L.,Thomas, P., Garry, P.J. & Goodwin, J. 239-249. Shopsin, B., Friedman, E.& Gershon, S. (1976) (1989)A prospective study of response to stressParachlorophenylalanine reversal of tranylful life events in initially health elders. ourn nu^ of cypromineeffects indepressed patients.Archives Geronbolo~, 42,627430. of General Psychiatry,33,811-819. Willner, P. (1995)Dopaminergic mechanisms in Siever, L.J., Murphy, D.L., Slater,S., de la Vega, E. depression and mania. In: ~ s y c k o ~ k a r ~ ~ c o l o ~ : The Fourth Generation & Lipper, S. (1984)Plasma prolactin changes of Progress (eds F. E.Bloom following fenfluramine in depressed patients & D. J. Kupfer), pp. 921-932.Raven Press,New compared with controls: a n evaluation of cenYork. tral serotonergicresponsivity in depression. Life Wirz-Justice, A.(1995)Biological rhythms in mood Science, 34, 102cp-1039. l o ~~: o ~ rGenth disorder.h P s y c h o ~ ~ r m a c oThe Sternbach, H. (1991)The serotonin syndrome. eration of Progress (eds F.E.Bloom & D. J. Kupfer), A~erican ~ournal of P s y c k i a t ~148,705-713. , pp. 999-1018.Raven Press,New York. Sulser, F. (1984)Regulation and function of Wu, J.C. & Bunney, W.E. (1990)The biological noradrenaline receptor systemsin brain. Psychobasis of a n antidepressant response to sleep pharmacological aspects. ~ e u r o ~ h a r ~ a c o l o g y , deprivation and relapse: Review and hypo23,255-261. of Psychiatry, 147~14-21. thesis. A~erican ~ournal Swayze, V.W., Andreasen, N., Alliger, R.J. et al. Wu, J.C., Gillin, J.C., Buchsbaum, MS.et al. (1992) (1990)Structural brain abnormalities in bipolar Effect of sleep deprivation on brain metabolisms affective disorder. Archives of General P s ~ c h i a t r ~ , in depressed patients. A ~ e r i c ~ournal a~ of Psychiatry, 149,53&543. 47,1054-1059. Tanaka, Y., Hazama, H., Fukuhara, T. & Tsutui, T. Young, S.N., Ervin, ER., Pihl, R.O. et al. (1989) (1982)Computerized tomography of the brain Biochemical aspects of tryptophan depletion in in manic-depressive patients: a controlled study. lo~, primates. P s y c h o ~ k a r ~ a c o 98,508-511. Folia Ps~chiatrica ~eurologica ~ ~ a ~36,137-144. a n ~ , Young, E.A., Haskett, R.F., ~urphy-Weinberg, Thompson, C., Mezey, G., Corn, T. et al. (1985)The V. & Watson, S.J.(1991)Loss of glucorcorticoid effect of desipramine upon melatonin and fast feedback in depression. Archives of General cortisol secretion in depressed and normal subPsychiatry, 48,693-699.
* *
In the last decadeit has become increasingly acceptedthat the aetiology of depression is multifactorial. Its psychosocial origins are now recognized by those for whom formerlybiological factors were the sole consideration. There is, however, a considerable gap between acknowledgement that something contributes to onset or perpetuation of a condition and translating that knowledge into practical management strategies. While gap this may merely indicate a scarcity of the resources necessary to effect this translation, it can also sometimes reflect the fact that, despite repeated reference to the importance of a particular factor, understanding of the process at stake is incomplete, and the implications thus not followed through. Much of social psychiatry deals with gross demographic variables such asand age, marital sex status, the relevanceof which, when planning for the individual patient, is often not immediately apparent. The rest consists largely of narrowly focused subtopics such as negative cognitions, social support or negative life events, and these are often defined differently so that integrating them into a mean~gful whole is not easy. Furthermore, much work uses rather mechanical definitions of these factors; for example, frequency of contact, rather than anything more qualitative about the relationship, is still often used an index as of good social support, so that the emotional meaningof any findings may not emerge with full strength. This chapter will s u m a r i z e recent work on the role of psychosocial factors in depression ina way which highlightsthe emotional meaningof the various experiences identified as predictive across the lifespan. It will also draw out the implications of this psychosocial perspective forthe management of depression.
Studies on life eventsperhaps exemplify morethan those in other areas the way in which researchreports can often fail to convey the full meaning of their findings. A life event is a substantial change in life style or prospects,
particular event or difficulty,but they reduce bias by ignoring the person’s subjective account. In fact the latter can be deliberately compared with the contextual rating for the same event to gauge whether the person is respondi in a manner that exaggerates or minimizesits stressfulness, compared with A set of rules the reactions expected from others with similar backgrounds. laid out in training manuals embodies decades of these consensus decisions and ensures high rates of interrater reliability for the contextual ratings (Temant et al., 1979; Parryet al., 1981).
Events
There are two ways in which this approach has thrown light on the links between life stress and depression. First, it has revealed crucial items that are missing from existing checklists, particularly revelations and disclosuresof sensitive informa~onabout theself and other close ties. These include finding out that a child has been involved in criminal activities, although as yet u n ~ o w to n the police, or declaring one’shomosexual it^ to parents. Second, it has shown that, in a sense, these lists contain too many items, as only a small proportionof the life events listed have relevance for depressive onset. Progress here has been through ever-increasing specificity. From theinitial definition of an event in termsof the change entailed, the focus moved to the et al., 19178)~ especially its tolong-term undesirability of that change (Sarason threat or unpleasantness above a certain severity threshold. This has distinguished ’severe events’ from other, more minor, undesirable ones; the negative implicationsof which onlylast a few days, since the latter were no more common among the recently depressed than in the general population (Brown et al., 1973). Thus events such as routine house moves, which were not rated severe, were not associated with depression. Nor were events whic were severe in the short term, such as finding a breast lump and learning from biopsy within 2 weeks thatit was benign. Whereas19% of women in a representative British innercity population experienced at least one of these severe eventsin a nine-month period (Brown & Harris, 1978), figures for those in similar communities with recent onset of depression ranged from 91% (Brown et al., 1986) to 62% (Bebbington et al., 1984). i om parable figures for psychiatric patients with neurotic unipolar depression ranged from 73% (Brown et al.,1994a) to 39% (~ebbington et al.,1988), while those for psychiatric 58% patients with melancholic or psychotic unipolar depression ranged from (Brown & Harris, 1978) to37% (Bebbington et al., 1988). As willbe discussed later the different rates among patients may well be due to different proportions, with first episodes between the different samples reported.
Later research focused even more specifically whaton was depressogenic. The severe loss events which precede depression were distinguished from the severedangers where as yet there has been no loss, for example a child’s acute asthma attack. The latter tend to precede anxiety disorders without comorbid depression (Finlay-Jones&: Brown, 1981; Brown et al., 1987). An important aspect of this definition of a severe loss wasthat it included not only severe lossesof people and material possessions but also of cherished ideas, for example of the trustworthy nature of close others. Examples of such ’losses of cherished idea’ include discovering a partner’s extramarital affair or one’s close friend’s continued failure to repay a substantial loan from another. The insights gained into the importance for depression of the ‘revelation events’ omitted from standard stress checklists were thus amplified by the parallel findings on the importance of loss of cherished ideasabout the fidelity and honesty of the people we rely on. More recently ithas become apparent that as a predictorof depression, the conceptof loss, needsfurther refinement. It is really only severe loss events with an element of humiliation or entrapment, or the finality of a death that play the crucial depressogenic role (Brownet al., 1995). Thirty-one per cent of women in the general population with at least one such event became depressed. Other losses, if substantial, even such asjob losses involving considerable financial hardship or adult children emigra~ngto the Antipodes were followed by a much oflower depressive rate onset-only 5%. However, it is important to note herethat if such ajob loss resulted from dismissal,if or such an emigration involved some estrangement from the parent, then a ratingof humiliation would be very likely. Moreover, if there had already been a financial difficulty, the job loss might also prove to be ’entrapping’. It mightunderline the improbabilityof climbing out of the debt trap, thus embody in^ the loss of a cherished idea, namelythe hope of paying off all the creditors. In those instances, therefore, the rate of depressive onset might be expected to be closer 30%tothan to 5%.
As mentioned earlier, difficulties are ongoing problemsassuch overcrowded housing or violent interpersonal relationships, lasting at least 4 weeks, which may or may not give rise to severe events. When they do so, this is often experienced asan entrapment. Another set of insights has arisen in connection with difficulties through the concentrated empathic listening required of the investigator by theLEDs processof data collection. It seems that even in the absence of an event, such difficulties,if involving morethan physical health problems and if rated on the top threepoints of severity (’major difficulties’), were more common before depressive onset than in the general community (termed ’major difficulties’, see Brown &:Harris, 1978). The reason why these
73
Previously it was thought that such unprovoked episodes hada distinctive symptom morphology, which was more severe and included melancholic or psychotic features (Carneyet al., 1965), but careful studies with clearer, less confounded definitions of events and depressive subtypes did not confirm this perspective (Leffet al., 1970; Paykel et al., 1971, 1984; Brown et al., 1979; Roy et al., 1985; Bebbington et al., 1988). A more recent study suggests that while first episodes of melancholic-psychotic depression resemble the neurotic subtype in following after'provoking agents' as defined above, subsequent episodes of this depressive subtype are more often than not unprovoked. Moreover, there is also a subgroup of neurotic depressions without provoking agents (Brownet al., 1994a). Amechanism involving kindling (Post et al., 1986) has been proposed to account for this, whereby once sensitized by an environmentally provoked first episode depression can be more easily internally triggered.
ic
About one in five of those who experience a severe event and one in three experiencing a humiliation/entrapment event become depressed. Characteristics that render some women more vulnerable to these events have become known as 'vulnerability factors'. In principle these could involve biological as well as psychological andsocial factors, since hyper-reactivity to stress may also stem from neuroendocrine factors. In practice, however, the term vulnerability factor in the literature on depression has referred to psychosocial factors which increase the likelihood that severe events and major 8;.Harris, difficuties willbe followed by depression. Early explorations (Brown 1978) pinpointed, among other characteristics, lack of employment outside the home and the presence of three or more children within it as vulnerability factors, but went on to try to give meaning to these links with work an motherhood by invoking the conceptof the 'role identities' available to each woman. These would vary in importance according to her involvement in the different domainsof motherhood, sexual partnership, work, housework, sport and other social and recreational activities. This concept, as it were, forms a bridge between gross demographic groupings such as employment status or parenthood and theindividual's sense of him/herself.It was argued that the number of such role identities available could prove crucial in protecting someone against the feelings of total hopelessness entailedby depression. A sense of 'hopelessness-specific-to-the-loss' would arise toa minor extent, at least temporarily, even a in person of robust character when faced with a severe a tendency to event of this kind,but more vulnerable individuals would have
5
ck's well- own cognitive pression (theself seems worthless, the world seems pointless eless) produces all further the interferences with vegetative tions which accompany depressio~( eck, 1967; see also ne woman when faced with her tee herself 'isn't that just typical of ross-I can never succeedat an ralization with thought ood secretaryan without me'. (The first,who generalized her negative role performance to became depressed, whilethe second, who,though very d her Self-esteem byr e ~ n d i n g herself of her positive roles, did not reach criteria for clinical depression.) ~ l t h o u g later h research has only rarely been able toatlo women between the life eventand the onset of depression inorder to CO m these anecdotal eculations, it has been ableto show the importance of this typeof negative evaluation of self in role performance (see below) and in predisposing women ressive reaction. Thus these early speculationsabout the protective of multiple role identities can still proveformative; although over the last 30 years changes in society have altered the meani a~ticularfactors from which the ideas originally arose,and wi l to render women vulnerable to depression formation). Thus changes concerning the propor in the workforce, the availability of nursery school placesand low-cost, easytion haveso altered the m e a ~ n gof work outside the home, ree or more children within it, that the longer figure as rs in the clearwaythey did in 1970. most important S are thus the 'othercharacteris~cs' briefly referred to above. .I gives as u ~ a r of y various vulnerability factors for depression ed at different times. They fall mainly into the two b ve self-evaluationand poor emotionalsupport. The of course, the low self-esteem which was hypothesized to promote the eneralization of hopelessness, and later found to predict a threefold higher te of depressive onset after a provoking event or difficulty in a 12-month eriod-33% compared with 11% (Brown et al., 1986). Sixty per of such onset cases exhibitedsuch ne ative self-evaluation, compared 30%of the rest of those with compa le stressful experience. In the sample the second broad type of ~lnerabilityfactor, lack of social rt, made a similar-sized con~ibution:41% without and 4% with an ortive context became depressed after a provoking with agent, 90% of such onset cases lacking adequate support, compared with 34% of the rest of those with at least one comparable severe event ormajor difficulty. The
~ ~ a4: A p p~ s y ce~ o~s o c i ~ l of ~ oaae epl~ e s s i o ~ Table 4.1 Vulnerability factors that potentiate the effects of severe events and difficulties among women by promoting the generalization of the hopeless feelings induced in anyone by such provoking agents. Early i d e n t ~factors: e~ de~ographiccategories/socialroles 1 Lack of employment outside the home’ 2 Presence of 3 or more children under 15 at home’ 3 Loss of mother by death or long-term separation before age 111
Recent changes in women’s rolehave revealed these to have been only ‘indicators’ of the more crucial psychic factors listed below Core vulnera~ility factors 1 Variants of psychological vulnerability (i) Negative evaluation of self (ii) Chronic anxiety and subclinical depression3
2 Variants of lack of social support (i) Intimate confiding with partner1 (ii) Lack of crisis support from core tie (iii) Negative elements in core relationships3 (iv) Lack of adequate parental care in childhood4 References from lBrown and Harris (1978); *Brown et al. (1986); 3Brownet al. (1990); 4Bifulcoet al. (1987).
four variables listed under this second heading either involve aof concern lack with her emotional well-being on the part of members of the individual’s core network, suchthat she cannot really speak freely to anyoneabout her worries, or else involve distinctly negative features such as coldness, or frequent criticism or arguments (Brown & Harris, 1978; Brown et al., 1986, 1990). All theseamount to a lackof support with stressful experiences, but they are also examplesof unsuccessful roles in particular relationships. It is therefore not surprising that such poor support is associated with negative self-evaluation (Brownet al., 1986,1990). However,despite this association, logistic regression shows that both are required to model depressive onset (Brown et al., 1986,1990). Mention should also be made of further psychological vulnerability indicated by states of chronic anxietyand subclinical (or minor) depression (Brownet al., 1990). The evolutionof ideas about vulnerability can be seen to have followed a similar course to those concerning life stress and depression, starting from work with straightfo~ard social survey items, such as employmentstatus or life change, and moving to the consideration of the emotional meaning of these items in different historical contexts. hindsight With it is easy to that see these two broad types of vulnerability echo the qualities identified as depressogenic among life events, with poor support mirroring exactlywhat is ’revealed’ by revelation events-the absence of a trustworthy close otherand negative self-evaluation resonatingwith humiliation events. However, this echoing phenomenon only becomes apparent once the critical thresholds
77
for support and self-evaluation have been located, andin the process their As already mentioned, support can be of many mutual influence appreciated. kinds, from the diffuse variety afforded by neighbours in the pub to the intimate type afforded by free and open confiding of private, possibly shameful, thoughts and feelings. It is the latter not the former which has pro protective against severe events. In a similar way this measure of selfevaluation includes, not only the general view of the self and its acceptance or rejection found in quick questionnaires (e.g. Rosenberg, 1965), but also opinions on performance in key roles.It spans a broad dimension (Brownet al., 1990) from the exuberant recounting of successes among those with marked of those Positive Evaluationof Self (PES) through the modest self-descriptions with neither negative nor positive self-evaluation, to the perjorative comments of those with definite Negative Evaluationof Self (NES). It is the latter, not the absenceof PES, which has proved a predictor of depressive onset during a 12-month followup (Brownet al., 1986,1990). This insight has been confirmed in work contra§ting the predictive power of the negatively and positively worded items of self-esteem questionnaires (Miller et al., 1989). Once these depressogenic thresholds wereid en ti fie^, the links reported between social support and self-evaluation (op. cit.) became intuitively easier to understand: pleasant chatter in a pub might add to a person’s already good opinion of him/herselfbut wouldbe unlikely to affect any change in a negative viewof the self built around particular shameful feelings too private to voice in such a setting. Talking one-to-one with a trusted friend, who is accepting, despite knowledge of the shaming items, may mitigate the impact of any new humiliation^ and so prevent the generalization of hopelessness, and, in the long run, may decrease, or even eliminate, the NE§, particularly if the confiding occurs regularly.
The vulnerabilities discussedso far have all involved risk factors pertaining before the life event occurs, although it is their role in the immediate aftermath which is proposed ascrucial, In addition the important role of both cognitive and pragmatic coping styles should also be registered, since these may be amenable to influence during this critical, mediating period. Both self-blame and ‘denial’ (a tendency to miss seeing the negative aspects of a situation until it is toolate) have been associated with increased risk of depression, as has helplessness in dealing with the situation (Lazarus & F o l h a n , 1984; Bifulco & Brown, 1996). By contrast ’downplaying’, or what is sometimes This type of cognitive coping termed ’suppression’ (Vaillant, 1976), augurs well. consists of fully confronting the meaning of the provoking stressor-again
note the protective function of developing a narrative-but also involves on some as ect which serves to mitigateits impact. This can occur e negative implications(’My son with cerebral palsy may not be able to move but about at least he is intelligent and sweet-na~red. I am so lucky I do not have MrsP’s Down’sdaughter’) or in discovering some new positive aspect to the situation (’One goodthing my cancerhas done is to bringthe whole family back together again’). Because the coping literature has tended to conflate this response with denial, sometimes calling both ’minimization’, it is often more difficult for professionals to appreciate the of the formerand distinguish it from the depressogenic role of ey difference between the two captures the essence of the psychotherapeutic concept of ’working through an issue’, again via the development of a coherent narrative:in denial this has been avoidedand in downplaying largely achieved.
So far, the modelof depression has been presentedin terms of onset, but very similar considerationsapply to recovery or improvement, wherethe role of hope, and its loss or regain, emerges once again.appears It that the processes involved may broadly mirror those leading to onset; in this case by changing circumstances b r i n ~ n gnew hope to a situation. Recovery or improvement from depressive conditionsthat have lasted for20 weeks or more is highly related tothe occurrence of events which, on common sense grounds, would be expected to bring hope, such as being rehoused, receiving a proposal of marriage from a new boyfriend or findinga job after months without one. oreover, in addition to such ’fresh starts’, reduction inthe severity of the difficulty and receiving support with it also playeda part ( 1992; Brown, 1993). findings concerning social support, already suggested by other interventions (van der Eyken, 1982;Pound Br Mills, 1985;Corney 1987; Elliott et al., 1988)’ have now been confirmed experimentally a trialinof a befriendin tervention with chronically depressed women(TO.Harris, G.W. Brown Robinson, unpublished information). In these experiments a trained volunteer was pairedwith a depressed womanwith whom she met weekly for at least an hour and encouraged her to treat her a confidante as in the way described above. In other words, she provided additional social support. Recovery/improvement (measured bythe Present-State Examination (Wing et al., 1974)related both to experience of a fresh start (or difficulty reduction) and to membership of the befriending rather than control group. Logistic regression suggested there were two other predictors, one involving very poor coping and another involving attachment style (see page 81). As predicted, *
79
depressed befriendees paired with volunteers who offered marked, active emotional support did better than those whose volunteer offered less suppo Another two reports focusingon chronicity gavea similar picture, but raise an additional issue concerning the relevance of distant past experience for meaning: both ina psychiatric patient sample (Brown et al.,199413) and in the general population (Brown& Moran, 1994) women with episodes lasting1 year or more had more often had severe interpersonal difficulties at onset or had suffered severe neglect and/or abuse in childhood. Previous findings had pointed to the former-for example, poor marriages had often figured as prolonging depressive episodes (~ounsavilleet al., 1979; Waring & Patton, 1984; Goeringet al., 1992; Hickie& Parker, 1992). Other reports had identified different forms of lack of social support as delaying recovery (Mann et al., 1981; Brughaet al., 1990; Georgeet al., 1989), as well as low income and other 'chronic strains' (Huxleyet al., 1979). However, the association with childhood adverse experience had not previously been documented and fitted neatly with the emerging perspective.
So far, the model of depression describedhas involved psychosocial factors which are all current (see the right of Fig. 4.1). In considering the practical implications of this perspective it is bound to be useful to consider the origins of these factors in the present and whether these too can be influenced a way in that can somehow reduce depression. What has emerged is conveyed diagrammaticallyby the series of arrows, with the continuous line representin the more sociostructural external pathways (strand l), and the dotted line the more internal psychic ones (strand2). However, this is not justsome mechanistic system whereby one thing leads to another and thence to depression, the various experiences along the lifespan seem to echo each otherofin terms meaning. The losses, shame and of lack emotional support found in the period immediately preceding depression often seem to have begun years before. A pointer to this emerged in early work where of loss the mother either by death or by separation of at least 12 months before the age of 11 years was found to render women more vulnerable to severe events and difficulties (Brown & Harris, 1978). Another study in Waltharnstow (North London), expressly designed to explore experiences intervening between such loss of mother and current mental state, suggested it was not the childhood loss itself but the poor qualityof relationships with the substitute care-givers that was the critical depressogenic feature. Once again it was the supportiveness of relationships which was so crucial, and this was highlighted by reports fcom other research teams (Parker & Hadzli-Pavlovic, 1984;Birtchell et al., 1988). Premarital pregnancy, which seemed to trap women in unsupportive
80
Strand 1 External sociostructural Recent
Childhood
--~~-----------------~
Childhood helplessness/ mastery , '
, , , II
1)
Fig. 4.3. Key psychosocial factorsin the aetiology of depression-a lifespan model.
partnerships that they might otherwise not have chosen, turned out to be a more frequent sequel to such lack of adequate replacement care in childhood (Harris et al., 1986, 1987). The Walthamstow study further explored other aspects of the modelof depression outlined above by including a measureof the woman's situational helplessness/mastery. This was found to relate both to poor substitute parental care after the maternal loss and to poor support (lack of intimate confiding) in current relationships, as well as to depression (Harris et aE., lggoa, 199ob). It seemed that, from early in life, some women had moved from one unsupportive situation to another and seemed unable to be masterful enough to break free of this cycle. This picture was confirmed in an inner city sample (Islington),which looked at sexual abuse and violence from a member of the familyas well as at marked parental rejection or neglect; it found such 'childhood adversity' in approximately one-thirdof the women before the ageof 17 years. These women went on to have roughly double the risk of experiencing depression at a clinical level in any one year during adulthood (Bifulco et al., 1987, 1994). Here the adversity and lackof support in childhood was clearly linked, not only with vulnerability such as lack of support and low self-esteem in adulthood (Harris, but 1993), also with severe life events in the current period, particularly those involving humiliation/ entrapment (T.O. Harris 82:G.W. Brown, unpublished information). Although the apparent inevitability of moving from early to later adversity emerges powerfully in every sample it is also important to note other pathways
(Rutter, 1987, 1989); some women did seem to have survived potentially damaging childhoods without a psychiatric episodeof any kind. Usually they were found to have succeeded in developinga close, confiding relationship without negative elements soon after leaving the childhood home, most often & Rutter, 1984; Brown et al., 1990). Sometimes, with a sexual partner( ~ u i n t o n however, the fresh start came much later: rs A, with three in-~atient o miscarriages, a stillbirth depressiveepisodes,twomaritalseparations brought on through domestic violence and four children taken into care for adoption, for the first time,a had lo-year period withno depression just before participating in our population survey. At the suggestion of her family doctor in Yorkshire she had, 10 years previously, moved downto London with the only child not removed from her and had met a different sort of boyfriend. She decribed the ama~ementshe felt at how much they had to say to each other: 'Sometimes we talked so much we almost forgot to go ~ometimes, however, the fresh start can come much earlier, as when a girl placed for adoption five times previously andback sentto thechildren's home because she was 'difficult', finally founda more tenacious foster mother when she was1 2 years old: 'I think I calmed down when I realized shewasn't going to send me back for not polishing thestairs, or not for anythin said. Formally adopted after 15 months she made an excellent relationship with her new mother, married, without haste, a most sup and, by the time of the survey, had reached age 48 years with no adult U psychiatric episodes despite the multiple repeated abuse which she hadbeen removed from hern a ~ r aparents. l
wit~out These two last, brief life-histories exemplify, not only how survival depression can often be a matter of good fortune,but also how both inner and outer in~uencesmay combine to produce a vicious spiral dragging someone down intopsycho pa tho lo^. The parallelcont~uousand dottedlines across t complex interplay between internal and the centre of Fig. 4.1 ~ g h l i g h the external factorsin determining the qualityof social support available to any individual. On the one hand, externally the range of ersons in the close network may include only critical or thick-ski~edpersonalities, unable to empathize with the predi~amentsin which women typically find themselves. On the other hand, a woman's intrapsychic characteristics may be contributing to her lack of support, either through her fearfulness and hesitation in finding anyotherfriendswith a moresympatheticutlook,orthroughher quarrelsomeness ord e m a n d ~ behaviour g drivin potential support figures away. In the latter case her style of relating to others (or her attachment style) 2
is asimportant as her external resources. Until departure her for London Mrs A was trapped in a victim role, not only bythe conveyor belt from childhood poverty and a premarital pregnancy(strand I) but also by her own helpless style (strand z), which had partly contributed tothat pregnancy and kept her in both marriages: the first eroding her self-esteem through the partner's constant infidelity; the second, literally, beating her down still further. The fannily doctor's successful intervention in advising her to move south somehow fired her with sufficient confidence not only to move but also tobuil relationship. Unfortunately the focus of that particular study was not upon management, so there is no information about the length of time she had been visiting the family doctor and the treatmentshe received. It is likely that he must have talked to her at some length to give her the confidence to make such a change.Be that as it may, what is documented isthat Mrs A, despite her helplessness, did not seem to have exhibitedan overall avoidanceof, or ambivalence about, relationships with other people. She had had the potential to enjoy friendships with women throughout her life, as also briefly with work colleagues, but circumstances had meant that, more oftenthan not, she was unable to meet up with these peoplewhen she most needed their support. In other words, although externally she had received poor social support, her attachment stylehad been fairlystandard for her classand generation. This contrast between 'standard' and 'non-standard' adult attachment styles (Harris & Bifulco, 1991; A.T. Bifulco et al., unpublished information) closely parallels the concept of internal working models of relationships developed by'John Bowlby (1988),whose notionof 'secure' and 'insecure' attachment in infancy as thefoundation of development and psychopathology has stimulated much recent research. Work in this tradition with adults has mainly rated their attachment security on the of basis how they have cognitively and emotionally processed their childhood relationships (in the Adult Attachment Interview, rather than focusing on theircurrent adult ties and Main & Goldwyn, 1988)~ whether they show avoidance, ambivalence or enmeshment therein. Studies of the latter type relating non-standard adult attachment style to depression have found it prospectively to predict both onset (A.T. Bifulco et al., unpu~lishedinformation) and failure to recover(TO. Harris, G.W. Brown & R. Robinson, unpublished information), even controlling for whether the environment is actuallyproviding good enough support (strand I). A nonstandard attachment stylehas also beenfound more commonly among those with childhood adversity (Harris & Bifulco, 1991; Bifulco et al., 1996b). The reason for labouring thispoint is that, both in the case of Mrs A and in the more general case of any managementof depression, it is useful to distin~uish the inner and outer origins of poor support in order better to target the way forward. If Mrs Ahad had a morefearful/avoidant attachment style, she might not have felt able to trust the family doctor enough to take his advice,and
3
might therefore never have had her fresh start. She might have been reached only with a more psychotherapeutic approach. It is perhaps finally worth noting that during herfirst 6 years Mrs A had a good relationship with her father beforehis death, only being sent into a home when her mother failed to 13 to help this same cope with the bereavement, and was summoned out aged mother care for the younger siblings born in the interim. Without that early supportive experience with her father, her attachment style might have developed differently to become non-standard. In that case, without intensiv psychotherapy, she might never have been to able build a successful relationship with her third partner. By contrast, Mrs B’s adult experience raises none of these issues. On the other hand there is some hint that, until her placement with her adoptive mother, her attachmentstyle in childhood (unfortunately not offically part of that study’s measured variables) may not have been standard but rather dismissive, verging on externalizing behaviour or even conduct disorder. At any rateit seems that something about her behaviour caused a succession of in psychiatric foster mothers to return her to the children’s home. Certainly patient samples many women recount life-histories where the combination of dependency and acting-out in adulthood hasleft them witha trail of failed partnerships and unsupportive friendships. In the light of Mrs B’s remark about how she ’calmed down’, it is difficult not to interpret her adoptive mother’s role as changing her early insecure attachment style to standard or I may secure: in other words, a positive experience along the external strand have altered the depressogenic potential of a strand-;? factor, her disposition towards disruptive behaviour. This process is, of course, just what emergesas the essential ingredient in many theoretical accounts of psychotherapy. What is important, however, is the insight that the process may not require a in the c o n ~ ~ i social ng professional therapist. In fact, empathic support figures network maybe better equipped than expertsin the clinic to help co~teract the long historyof humiliation or neglect suffered by many with depression, and canbe enlisted as key alliesin the therapeutic process, along the lines of a successful trial of treatment for agoraphobia using husbands and female friends (Oatley& Hodgson, 1987). These two life-histories provide interesting parallels with anotherissue, the differing origins of the severe events provoking depression. Some are more clearly contingent upon the subject’s own impulsive or argumentati~e behaviour, whereas others, such as deaths, are clearly independent. The forme (contingent) events have been found more frequently among depressed patients with personality disorderof the cluster B variety-that is, dramatic or histrionic (TO. Harris & G.W. Brown, 1996a,b, unpublished information). In between contingent and independent events are grouped others where th personality of a close relative or friend, usually a partner or child, is crucial in
~ h ~4: A p psychosocial t ~ ~ ode^ of d ~ p ~ ~ s s ~ o n bringing about the event,but some suspicion remainsthat the subject could have played a more active role preventing in it-in other words,that she was somehow passively involved in its origin. Before managing to escape her conveyor belt, MrsA seemed to have suffered events of this last type, while as a child Mrs B seemed to have been contributingto her own frequent events in being sent back to the children’s home once again-in other words, experiencing contingent events. It is tempting to speculate that without adoption Mrs B might have later left the children’s with home something like the histrionic personalityfound associated with contingent events,and that if she had gone on to have a premarital pregnancy this personality might have become more and more entrenched. This speculation is rendered more plausible when one remembers how borderline personality disorders, despite producing contingent events, have often stemmed from independent environmental abuseand neglect of the kind MrsB underwent in her first3 years (Herman et al., 1989). Moreover, borderline personality has been linked with insecure attachment style (of the enmeshed subtype; Patrick et al., 1994).
Before turning to the implicationsof this psychosocial model for management of depression, it important is to emphasize one of its assumptions that has not so far beenspelt out: that physiological changesare assumed to parallel the psychosocial responses, and that pharmacological treatments arein no way seen as unimportant just becausethe model has not so far incorporated them. It is already pre-eminently a multifactorial model with room to include new factors as they emerge. In fact work is already beginning to illuminate such connections; for example, severe events and difficulties predict a poor response to antidepressant treatment (Lam et al., 1994). Furthermore, experiencing severe events before onset has been linked with higher urinary-free cortisol (Calloway & Dolan, 1989). It is asif a third physiological strand was also inextricably involved in the interplay between strands 1 and 2. A corollary of this is the issue of comorbidity which can be crucial for management decisions. ~nfortunatelythis chapter precludes detailed consideration of comorbidity issues, but it is noteworthy that the model for depression has also beenfound of relevance in the study of anxiety (Finlay-Jones& Brown, 1981)and alcoholdependence (Gorman & Brown, xcyp), albeit with key variations in meaning. For example,events before anxiety onset tend to involve of threat future loss, rather than humiliations which have already occurred. In the same way, the central implication of this multifactorial psychosocial model isthat even psychosocial treatmentshould be multimodal. This may not have emerged clearly so far, as the resum6 of this modelhas concentrated
on the similarity of meaning emerging from various studies despite their being focused on separate factors within it. Thus the sense of entrapment in a family where there is abuse from a parent figure (childhood adversity) is paralleled by entrapment in a violent marriage (adult adversity); the sense of isolation likely after a humiliation or death event (recent adversity) is paralleled by the isolation likely among those in a crisis without any close confidante (poor social support). The implication is that the sufferer’s sense of hopelessness and powerlessnessis not purely the resultof the depressive episodebut has arisen from his or her life context and that confrontingall aspects of thisevents, difficulties and the tendency to produce them; support available, copin and attachment style-will, in some way, alleviate the intensity of the depressive feelings. Such a thought is hardly new; most treatment programm in practice now subscribe to such avision, even if their handbooks seem more focused on a single issue, such as problem-solving or negative c o ~ i t i o n s . There are, however, four ways in which this lifespan model can sharpen appreciation of what such a perspective really implies.
First, the complexity of the model suggests that confronting the sufferer’s hopelessness should occur from whatever direction is possible, that treatmen should be multimodal. The fresh start events, engineered by a treatment figure, which change the external world and thereby end the entrapment-for example, rehousing-may, in the short term, be no less effective therapeutically than fresh starts brought about by the patient herself after changing her coping style in response to solution-focused or cognitivetherapy. In the long term, however, further severe events may occur in another domain than housing and development of a more masterful coping style may then prove more effective in postponing relapse. Yet in the short term the patient may be incapable of absorbing the cognitive therapy required to increase her mastery until shecan see some real shift in her environmental burdens. Taking a simila multi~odalperspective, the supportive listening, which may help a patient make sense of her feelings by setting them in context and softening selfderogatory tendencies, shouldbe widely aimed for among all clinical staff, not just the therapist or key worker, whose sickness and holidays can often prove to be setbacks in the recovery process. Relatives can alsobe enlisted, although this may require special induction (see below). h some settings with scarce resourcesof professionals befriending may prove a useful alternative, and even where once-a-week therapy is available it may be a helpful adjunct. To set this plea for multimodality more clearly in the contextof this chapter, 3 and 2 means thatboth should be the continuing interplay between strands targeted for change lest by dealing only with one, the door is left open forthe
86
other to undo the therapeutic work with the first. Thus, even where the humiliation caused by a poor marital relations~ipseems more clearly the precipitant of the depressive episode than the family’s financial troubles, a treatment package focusing exclusively upon the former mi possibili~that the next electricity billwould reopen the marital discord,as each memberof the couple blamedthe other forthe new debt.
In contrast to the generality of the multimodal treatment perspective, the second c o n ~ i b ~ t i of o nthis aetiological model is to up point certain specificities; using it at the assessment stage reveals that not everyone res severe eventwith depression has arrived there by the same route. On the one hand it is possible to ask both of severe eventsand of poor emotionalsupport how much is due to internal strand-2 factors and how much simply to the environment. In this way the relative distribution of resources between psychotherapy and social inter~entioncan be more clearly planned. Whether the provoking event was independent or contingentupon the patient’s agency will be relevant when deciding how best ensure to something similar will not occur againthus preventing s ~ p t o m a t i improvement. c Is the poor emotional support the result of the patient’s attachment style which endows her with a reluctance to confide in anyoneabout her difficulties? Or are there external reasons, maybe because her best friend died 2 years previously, and her husband became intolerant after he suffered a strokeand became bedridden 3 years ago,so that she has been unable to out getto meet new friends? Further specificities revealed inusing this modelin initial assessments would be the different types of non-independent events-whether depen person’s activeimpul§iv~nessor her passive compliance-and different types of non-standard attachment styles-whether enmeshed in ambivalent relationships with frequent negative interactions, or whether fearful or dismissive of others who could give emotional support, with consequent absence of social ties. The parallels between these contrasting types of event production (or stress generation)and the contrasting typesof non-standard attachment styles have already been mentioned when discussing Mrs A and Mrs B, but hig~ightingthese similiarities in talking to the patient about herself can often provean extra therapeutic tool. Insteadof providing treatment for generally dysfunctional attitudes, a more specific focus can be given for impulsive hostile tendencies as opposed to passive self-derogatory ones.
The third important feature of this model is its lifespan nature. Again in terms
of theory this may seem nothing novel and yet, in practice, many therapies for depression seem to emphasize oneofend a person’s history more than the other. Some psychoanalytic-based therapies miss opportunities to help the patient deal with the meaning of their life history because they highlight past relationships and largely ignore all current onesbar the transference (thatof patient with doctor). By contrast, the focus of interpersonal therapy (Klerman et al., 1984) on current relationships, of and cognitive therapy (Beck et al., 1979) on current dysfunctional attitudes, might often benefit from linking these w childhood adverse experience. Asthe patient comes to understand that the origins of her current snags and dilemmas in relationships-as formulated in cognitive analytic therapy (Ryle, 1ggo)”often lie in the past when she was too young to be held toblame and stem from the environment outside herself, she may feel it less demeaning to find herself in her current situation, and her self-esteem may improve. She may also find it less threatening to work on changing herself. This aim of linking past with here and now as a way of understanding hypotheses in psychotherapy has much in common with the ‘conversational’ therapy model (Coldberg et al., 1984; Hobson, 1985), even though this appears to discourage too great a focus on the past.If these links are made during discussions of concrete memories which revive earlier emotions, it is easier for the patient to understand how she is relivingpast the in the present and to feel motivated to change. This applies particularly to parenting style where, despite best intentions, those who experienced childhood adversity often find themselves inflicting similar neglect or abuse upon their own children, who then produce further severe events which, in turn, keep them depressed. One further implication of the model’s long time perspective is, however, less encouraging. It would be surprising if a mere 20 therapy sessions, however supportive, could on their own undo the of work 20 or more years. Unfortunately the health services’ current cost-reductiondriven plans leavelittle room to carry out proper tests of the valueof longer therapies with longer-term follow-ups. The implications of the model are that the cognitions in strand 2 are highly dependent upon external factors in strand I; thus if someone is to get off the conveyorbelt they need to change more than their automatic negative thoughts; they will need to ensure their curre roles will enhance not diminish their identity, and most particularly the role in a core relationship which will reflect back to thema favourable self-image. Again this is often what is done in practice by some therapists, who see the establishment of a supportive marital partnershipas a crucial outcome fora patient, but withoutit being speltout this maybe missed by others.
The fourth set of implications involves the supportive process itself.
88
Naturalistic studies in the general population comparing relationships which did and did not prove protective against depression suggest a range of nonsupportive responses (Harris, 1992). Their nature may be informative for professionals, both in therapy sessions or encounters on the ward, and in advising relatives. Such failures of support may either involve negative responses to the depressed person’s confiding, or anticipated negative responses which have prevented confiding (Table 4.2). The most common are critical responses:’Well I did warn you; you have only yourself to blame’; or insensitive, thick-skinned reactions: ’It’s really not as bad as all that.’ Paradoxically, in some cases these responses may embody exactly what the model implies is true. In the first response, that the severe event provoking depression was in some way the result of the individual’s own cluster B personality; or in the second response that the individual’s low self-esteem has caused her to over-react. The model also implies that to emphasize this part of the truth would be counterproductive at this stage. Building a supportive therapeutic relationship involves developing trust, which takes Table 4.2 Some key considerationsin supportive interactions. -
~~
Three questions must be answer~dpositively: 1 Does the patient feel able to confide? 2 Is there a moderately or markedly active emotional response to the confiding by the other person? 3 Is there an absenceof definite negative response (such as ’I told you didn’tI’ or ’Why don’t youdo the decent thing and agree to their demands’)? Features that might cause one of above answers to be negative External: Nature of other person who may: Internal: Patient expects other person to:
Criticize patient for feeling upset about event Criticize patient’s other behaviour/attitudes Fail to understand implications of event: unempathic Deny implicationsof event (because covers up guilt) is a co-victim) Avoid implications of event (because Deflect discussion of event to focus on own problems Side with a third person over event against patient React falsely: act contradictorily behind patient’s back Fail to maintain confidentiality: gossip Adopt interfering (even if not critical) stance over event Be unavailable (for reasons of geography, timetable) Become too emotional about patient’s confidences Hurry patient into action before emotionally prepared Try to distract patient misguided kindness
-
time and requires the patient to have confidence that disclosure of private shortcomings will not later work against her. Thus opening up to other members of a ward, whether to fellow patients in a group or to key staff members, may feel more difficult than a therapist to or research worker outside
the ward. The acceptance of the latter would seem less crucial for the patient since he/she will notbe involved in the daily routine of the next few weeks. Furthermore, one of the ways of helping people to open up is to sympathetically ask key questions about what they have experienced, it may and be easier for such an outsider to do this without being experienced as intrusiv Nevertheless, by taking things at the patient’s own pace, bearing rnindinher attachment style and needs for self validation, ward staff can almost certainly facilitate the assimilationof problematic experiences, which is such a central process in successful psychotherapy (Stiles et al., 1990) and which was referred to above in terms of the capacity to construct a coherent narrative as an essential prerequisite to change.Table 4.2 sumarizes possible sources of failures in support. Although these have been identified in the context of everyday relationships, their relevance for management of depression is equally apparent. Concern about other patients’ scornful opinions be as can powerful as anxiety about private information reaching rival relatives or neighbours, and worries about professionals passing on shameful secrets to their more critical colleagues, or even siding with visiting relatives, can sometimes imped the therapeutic process unless directly confronted and assuaged, Before concludingit is important to remember that research has identified subgroups of those with depression where the psychosocial model does not appear to apply (see above), and among whom antidepressant medication has been found effective. Even among those whose history seems toa exhibit classic resemblance to Fig:4.1, the roleof pharmacological inte~entionsmay often prove crucial. In managing the depressive outcome of the complex and lengthy psychosocial processes described here, it is important to take no risks and to consider every potential pathway to recovery.
The research, on the basis of which this model was developed, was conceived by George W. Brown and supported by the Medical Research Council, the Social Science Research Council (as it then was known) and, in part, by N e ~ o r Ik of the John D. and Catherine T. MacArthur Foundation. I am indebted to the many colleagues who participated in the data collection and analysis overthe last 28years, and to the many women who recounted their us in order to contribute to our understanding. experiences in detail for
Bebbington, P.E., Stult, E.,Tennant, C.& Hurry, J. Bebbington, P.E., Brugha, T., Maccarthy, B. et al. (1984) Misfortune and resilience: a community (1988) The Camberwell Collaborative Depression study of women. P s y c ~ o ~ ~~g ~e c~u ~ i c14, ~ ~ e Study. , I: Depressed probands: adversity and form the 347-363of depression. British ~ o ~ r of ~ aPsyl
90
chiatry, 152,754-765. Beck, A.T. (1967)Depression: Clinical, Experimental and heo ore tical Aspects. Staples Press, London. Beck,A.T., Rush, A., Shaw,B.F. & Emery, G. (1979) Cognitive Therapy of Depression. Guilford, New York. Bifulco, A.T. & Brown, G.W. (1996) Coping and onset of clinical depression. 2: An aetiological and Psychiatric model. Journal of Social P s y c h i a t ~ 31,163-172. Epidemiolo~, Bifulco, A., Brown, G.W. & Harris, T.O. (1987) Childhood loss of parent, lack of adequate parental care and adult depression: a replication. Journal of Afective Disorders, 12,115-128. Bifulco, A., Brown, G.W. & Harris, T.O. (1994) Childhood experience of care and abuse (CECA): a retrospective interview measure. Child PsychoZo~and Psychiatry,35, 14q-1435. Bifulco, A.T., Moran, P.,Brown, G.W. & Ball, C. (1996a)Cognitive coping and onset of depression in vulnerable London women. Bifulco, A.T., Brown, G.W., Lillie, A.& Moran, P. (1996b)Adult AttachmentStyle and Depression: a new measure of ability to make supportive relationships. Birtchnell, J., Evans, C. & Kennard, J. (1988) Life history factors associated with neurotic symptomatology in a rural community sample of 4049-year-old women. Journal of Afective Disorders, 14,271-285. Bowlby, J. (1988) A Secure Base: Parenf-Child Attachmenf and ~ e a l~ fu ~ m a~Development. n Basic Books, New York. Brown, G.W. (1993) Life events and affective disorder:replications and limitations.Psychosomatic Medicine, 55, 248-259. Brown, G.W. & Birley, J.L.T. (1968)Crises and life changes and the onset of schizophrenia.Journal o f ~ e a l t ti h Social Behavior, g, 203-214. Brown, G.W. & Harris, T. (1978) Social Origins of
Depression: A study of Psychiatric Disorder in Women. Tavistock Press,London and Free Press,
New York. Brown, G.W. & Moran, P. (1994) Clinical and psychosocial origins of chronic depressive episodes. 1: A community survey. British Journal of Psychiatry, 165,447-456. Brown, G.W.,Sklair,F., Harris, T.O. & Birley, J.L.T. (1973) Lifeevents and psychiatric disorders. I: Some methodologicalissues. Psycholo~caZ Medicine, 3’74-78. Brown, G.W., Ni Bhrolchain, M. & Harris, T.O. (1979) Psychotic and neurotic depression. 3: Aetiological and background factors. JournaZ of Afective Disorders, I, 195-211. Brown, G.W., Andrews, B., Harris, T.O., Adler, Z .
& Bridge, L. (1986) Social support, self-esteem and depression. PsychologicaZMedi~ine,16,813831. Brown, G.W., Bifulco, &Harris, A. T.O. (1987)Life events, vulnerability and onset of depression: British ~ o ~o~Psych~try, ~ a l 150, some refinements. 30-42. Brown, G.W., Adler, 2.& Bifulco, A. (1988) Life events, difficulties and recovery from chronic depression.British Journalof Psychiatry, 152,487498. Brown, G.W., Bifulco, A., Veiel, H. & Andrews, B. (1990) Self-esteem and depression. 2: Social correlates of self-esteem. Social Psychiatry 8 Psychiatric Epidemiolo~, 25,225-234. Brown, G.W., Lemyre, L. & Bifulco., A. (1992) Social factors and recovery from anxiety and depressive disorders: a test of the specificity hypothesis. British Journal of Psychiatr~,161, 44-54. Brown, G.W.,Harris, T.O. & Hepworth, C. (1994a) Life events and ’endogenous’depression:a puzArchivesof GeneralP s y c ~ ~ t r51, y, zle re-examined. 525-534. Brown, G.W., Harris, T.O., Hepworth, C. & Robinson, R. (1994b) Clinicaland psychosocial origins of chronic depression. 2: A patient en165, 457-465. quiry. British Journal of Ps y c h ia t ~, Brown, G.W., Harris, T.O. & Hepworth, C. (1995) Loss, h ~ a t i o and n entrapmentamong women developing depression: a patient and non-patient comparison. PsychologicalMedicine, 25,7-21. Brugha, T., Bebbington, P., Maccarthy, B., Sturt, supE., Wykes, T.& Potter, J. (1990) Gender, social port andrecovery from depressive disorders: a Ps~c~ological Medicine, prospective clinical study. 20,147-156. Calloway,P. & Dolan, R. (1989)Endocrine changes and clinical profilesin depression. In: Life Events and Illness (eds G. W. Brown & T. 0.Harris), pp. 139-160. Guilford, New York. Carney,M.W., Roth, M. & Garside, R.F. (1965)The diagnosis of depressive syndromes and the prediction ofECT response. British Jou~nalof Psychiatry,IIZ,659-674. Corney, R. (1987) Marital problems and treatment outcome in depressed women: a clinical trial of social work intervention. British JournaZ of Psychiatry, 151,652-660. Dohrenwend, B.P., Raphael, K.G., Schwartz, S., Stueve, A. & Skodol, A. (1993) The structured event probe and narrative rating method for measuring stressful life events. In: ~ a n d ~ofo o ~ Stress (eds L. Goldberger & S. Breznitz), pp. 174-199. Free Press, New York. Elliott, S.A., Sanjack, M. & Leverton, T.J. (1988)
~ ~ a p4: tAeps~c~osocial ~ ~ o ~of e~ el ~ ~ e s s i o ~ disorder: The Walthamstow Study.I: The role of lack of adequate parental care. Psychological Social Support: Formats, Issues, ProcessesEfects and Medici~e,16,641-659. (ed. B. H. Gottlieb),pp. 87-110. Sage, London. Harris, T.O.,Brown, G.W. & Bifulco, A.(1987) Loss Finlay-Jones, R. & Brown, G.W. (1981) Types of of parent in childhood and adultpsychiatric disstressful life events andthe onset of anxiety and order: The Walthamstow Study. 2: The role of depressive disorders. PsychologicalMedicine, 11, social class position and premarital pregnancy. 803-815. Psychological ~ e d i c i n e17,163-183. , George, L.K., Blazer, D.G., Hughes, D.C. & Fowler, Harris, T.O., Brown, G.W. & Bifulco, A.(1990~1) DeN. (1989) Social support and the outcome of pression and situational helplessness/mastery major depression. British Journal of Psychiatry, in a sample selected to studychildhood parental loss. Journal of Affective Disorders, 20,27-41. 1541 478-485. Goering,P.N., Lancee, W.J.& Freeman, S.J.J.(1992) Harris, T.O., Brown, G.W. & Bifulco, A. (199ob) Marital support andrecovery from depression. Loss of parent in childhood and adultpsychiaBritish Journal of Psychiatry, 160,7682. tric disorder:a tentative overall model. DevelopGoldberg, DX, Hobson, R.F., Maguire, G., P. et al. ment andpsycho pa tho lo^, 2,311-328. (1984) The clarification and assessment of Harris, T.O., Brown, G.W. & Robinson, R. (1996) a method of psychotherapy. British JournffZof Befriending as an intervention for chronic depresPsychiatry,144,567-580. sion among women in an inner city. Gorman, D.M. & Brown, G.W. (1992) Recent Herman, J., Perry, J. & van der Kolk, B. (1989) developments in life-event research and their Childhood trauma in borderline personality disorder. American Journfflof Psychiatry, 146, relevance for the study of addictions. British Journfflof Addictions, 87,837-849. 490-495. Harris, T.0.(1992) Some reflectionsof the process Hickie, I. & Parker, G. (1992) The impact of an of social support and nature of unsupportive uncaring partner onimprovement in non-melbehaviours. In: The Meaning and Measure~entof ancholic depression.Journal of Affective Disorders, Social Support (eds H. 0.F.Veiel & U. Baumam), 25,147-160. pp. 171-189. Hemisphere Publishing Corpora- Hobson, R.F. (1985) Forms of Feeling: The Heart of Psychotherapy. Tavistock, London. tion, Washington, DC. Harris, T.O. (1993) Surviving childhood adversity: Holmes, T.H. & b h e , R.H. (1967) The social readWhat can we learn from naturalistic studies?In: justment rating scale. Journal of Psychoso~atic Research, 11,213-218. Surviving Childhood Adversity: Issues for Policy and Huxley, P.J., Goldberg, D.P., Maguire, G.P. & Practice (eds H. Ferguson, R. Gilligan & R. Kincey, V.A. (1979) The prediction of the course Torode), pp. 93-107. Social Studies Press, Trinity of minor psychiatric disorders. Bri~ishJournaZ of College, Dublin. Ps~chiutry,135,535-543. Harris, T.0.(1996) Work, motherhood and mental B.J., state in transition: instability of indicator vari- Klerman, G.L., Weissman, M.M., Rounsaville, Chevron, R.S. (eds) (1984) Inte~ersonalPsychoables and implications for replication studies, of Depression.Basic Books,New York. t~erapy (Unpublished information.) Harris, T.O. & Bifulco, A.(1991) Loss of parent in Lam, D.H., Green, B., Power, M.J. & Checkley, S. (1994) The impact of social cognitive variables childhood, and attachment style and depression ~ t the Lije Cycle on the initial level of depression and recovery. in adulthood. In.A t t a c h ~Across Journal of Afective Disorders, 32,75-83. (eds C. M. Parkes & J. S. Hinde), pp. 234-267. Lazarus, R.S. & Folkman, S. (1984) S~ress,Appraisal Tavistock, London and Routledge, New York. and Coping. Springer Verlag, New York. Harris, T.O. & Brown, G.W. (19964 The origins of life events preceding onset of depression. 1: In- Leff, M.J., Roatch, J.F. & Bunney, W.E.(1970) Envidependent events and humiliation/entrapment ronmental factors preceding the onset of severe in the general population. (Unpublished depressions. Psychiatry,33, 293-311. Main, M. & Goldwyn, R. (1988) Adult A t ~ a c h ~ e n t information.) Class~cationSystem Version 3.2.University of Harris, T.O. & Brown, G.W. (199613) The origins California, Berkeley. of life events preceding onset of depression. 2: Contingent events and dramatic personality Mann, A.H., Jenkins, R. & Belsey, E. (1981) The disorder in a patient series. (Unpublished twelve-month outcomeof patients with neurotic ~ca~ illnessin generalpractice.P s y c ~ o ~ oMedicine, information.) Harris, T.O., Brown, G.W.& Bifulco, A.(1986) Loss 1% 535-550. of parent in childhood and adult psychiatric Miller, P.McC., Kreitman, N.B., Ingham, J.O. & Parent groups inpregnancy: a preventive intervention for postnatal depression.In:Marshalling
~ ~ a4: A ~ ~syc~osocial ~ e r model of d e ~ r e s s i o ~ Sashidharan, S.P. (1989) Self-esteem, life stress esteem: Society and the Adolescent Image. Princeton and psychiatric disorder. Journal of Afiective University Press, Princeton, NJ. D~orders,1 ~ ~ 6 5 - 7 6 . Rounsaville, B.J., Weissman, M.M., Prusoff,B.A. Oatley, K.& Hodgson, D. (1987) Influence of hus- & Herceg-Baron, R.L. (1979) Marital disputes bands on the outcome of their agoraphobic and treatment outcome in depressed women. wives' therapy. B r ~ t ~Journal sh of Psychiat~,150, Com~rehensive Psychiat~,20,483-490. 380-3236. Roy, A., Breier, A., Doran, A.R.& Picar, D. (1985) Parker, G.& Hadzli-Pavlovic,D. (1984) ModificaLife events and depression: relation to subtypes, tion of levels of depression in mother-bereaved ~ournalof Afiective Disorders,g, 143-148. women by parental and marital relationships. Rutter, M. (1987) Psychosocial resilience and proPsychological ~ e d i c i n e14, , 125-135. tectivemechanisms. AmericanJournal of Parry, G., Shapiro, D.A. & Davies, L. (1981) ReliOrthopsychiatry,57,316-333. ability of life event ratings: an independent Rutter, M. (1989) Pathways from childhood to adult life. Jack Tizard memorial lecture. replication. British Journalof Clinical Psychology, Jour~al of Child Psychology andPsychiat~,30,q-51. 20,133-134. Ryle, A. (1990) Cognitive Analytic Therapy: Active Patrick, M., Hobson,P., Castle, D., Howard,R. & Maughan, B. (1994) Personality disorder and theParticipation in Change. John Wiley & Sons, mental representationof early social experience. Chichester. Sarason, I., Johnson, J.H. & Siegel, J.M. (1978) ~ e v e l o p ~ eand n t psycho pa tho lo^, 6,375-388. Paykel, E.S., Myers, J.K., Diendelt, M.N., Klerman,Assessing the impact of life changes: development of the life experiences survey. Journal of G.L., Lindenthal,J.J. & Pepper, M.P. (1969) Life Consulting and Clinical Psychology, 46,932-946. events and depression: a controlled study. Seligrnan, M.E.P. (1975) Helplessness~On Depression, Archives of General Psychiat~, 21,753-760. Development and Death. W.H. Freeman, San Paykel, E.S., Prusoff, B.A. & Klerman, G.L. (1971) The endogenous-neurotic continuum in depres- Francisco. Stiles, W.B., Elliott, R., Firth-Cozens, J.A. et al. sion, rater independence and factor distributions. (1990) Assimilation of problematic experiences Journal of Psychiat~ic ~esearch, 8,73-90. by clients in psychotherapy. Psychotherapy, 27, Paykel, E.S., Emms, E.M., Fletcher,J. & Rassaby, 411-420. E.S. (1980) Life events and social support in P. & Hurry, J. puerperal depression. British Journal of Psychi- Tennant, C., Smith, A., Bebbington, (1979) The contextual threat of life events: the atry, 136,339-346. concept andits reliability.Psychological Medicine, Paykel, E.S., Rao, B.M. & Taylor, C.N. (1984) Life stress and symptom pattern in out-patient de9,525-528. Vaillant G. (1976) Natural history of male psychopression. Psychological Medicine,14,559-548. logical health. The relation of choice of ego Post, R.M., Rubinow, D.R. & Ballenger, J.C. (1986) Conditioning and sensitisation in the longitudi-mechanisms of defense to adult adjustment. Archives of General Psychiatry,33,535-545. nal course of affective illness. British Journal of van der Eyken,W. (1982) Home Star: A Four year Psychiat~,149,191-201. Pound, A.& Mills, M. (1985) A pilot evaluation of €valuation. Homestart Consultancy, Leicester. Newpin-home-visiting and befriending scheme Waring, E.M.& Patton, D. (1984) Marital intimacy and depression. British Journalof Psychiat~, 145, in South London.A,C.C.I? ~ e ~ s l e t t(4). er Quinton, D. & Rutter, M. (1984) Parents with chil- 641-644. dren in care. 2: Intergenerational continuities. Wing, J.K., Cooper, J.E.& Sartorius, N. (1974)The Measure~entand Class~icationof Psychiatric Journal of Child Psychology and Psychiatry, 25, Symptoms. Cambridge University Press, Cam231-250. bridge. Rosenberg, M. (1965) The measure men^ of Self-
93
Cognitive therapy has been seen as one of the most promising innovations in the treatment of depression in the past two decades. Its basic tenet the is cognitive theory in which emotions are seen to be influenced by meanings people ascribe to events in the external world. The depressed individual's cognitive triad (views about the self, the world and the future), negative thoughts and assumptions (rigid or overgeneralized rules) haveimportant contributions toour ~ d e r s t a n d i n g of how depressed patients construct their internal worlds. Clinically, therapists find that aspects of the theory such as the cognitive triad and the relationship between thought, mood and behaviour, canhelp depressed patients describe and make senseof their overwhelming state of mind. The contents of automatic thoughts also enable therapists to have window a into how patients construe their world and work out some general rules or assumptions that have implications for therapy. However, despite the effectiveness of cognitive therapy as a treatment for depression, empirical evidence forthe cognitive theory of depression is incomplete. For the sake of clarity Beck's cognitive theory for depression can be divided into the descriptive and causal aspects (Beck, 1967; Beck et al., 1979). This chapter consistsof three sections: first, an overview of the cognitive theory of depression is presented in order to give an overall picture; then more details of the descriptiveand causal aspectsof the cognitive theory are elaborated and explained; followed by a review of the empirical evidence. It can be argued that Beck (1983) intended his theories to apply to various levels of depression, both clinical and non-clinical. Hence in this review empirical evidence from both groups has been included. Furthermore, the clinical implications of the cognitive theory for the management of depression is discussed to promote the idea of clinical practiceguided by theory.
Beck's cognitive theoryof emotional states spellsout how clinical problems such as depression arise (Beck, 1967, 1983; Beck et al., 1979). Figure 5.1 is a
94
e.g. rejection or unrealistic pressure to achieve
Core belief(s1, i.e.
‘Iam unloveable or incompetent’
I
Adverse events relating to specific vulnerability
Fig.5.1 Cognitive theoryof emotional disorder.
linear summary of Beck‘s cognitive theory of emotional states. It is hypothesized that certain beliefs about ourselves, other people and events are developed from an early age. These beliefs about ourselves, ’core beliefs’, are regarded as truths and are encoded at an early age in linguistically simple forms such as ’I am unloveable or incompetent’. Beck (1983) hypothesized that the relevant core beliefsin therapy were relatedto the themes of loveability or competence (see Table 5.2,p. 114, for more details). It is believed that these core beliefs develop in relation to early experiences, suchas traumas and separation, patients’ perceived parental attitudes, relationships
C h ~ ~ t5: eA r cog~itivetheory of ~ e ~ r e s s i o ~
with peer groups and authority figures and any history of abuse. Beck(1967) hypothesized that early childhood experiences in the form of rejection or authoritarian pressure from significant others can lead to the development of a dependent personality or a value-achiever with little intrinsic sense of selfworth. Beck et al. (1979) proposed the term ’schema’ as thestable cognitive patterns that ‘maintains a depressed patient’s pain-inducing and self-defeating attitudes despite objective evidence of positive factors inhis life’ (Beck et al., 1979, p.12). Schema was defined as functiona~ a unit of info~mationprocessing ’for screening, coding, and evaluating the stimuli that impinge on the organism. It is the mode by which the environment is broken down and organized into its many psychologically relevant facets. On the basis of schemas, the individual is able to ..Categorize and interprethis experiences in a meaningful way’ (Beck, 1967; p. 283). Both core beliefs and dysfunctional assumptions are hence ’schemata’. Krantz (1985) refined the definition of the schema as the basic rules that are relatively independentof current life circumstances. Furthermore, Beck (1987) hypothesized that ’at the deepest level the conditional phrases starting with ”since” are dropped out of the rule, leaving the ”core schema” consistingof an absolute concept such as’’I am a fraud orincompetent”’. Dysfunctional assumptions are seen as a permutation of the basic core beliefs in order to cope with the internal world the ~ d i v i d u alives l in. Hence themes tend tobe related to loveableness, acceptance, achievement, or control (Beck, 1983). This, in turn, leads to the vulnerability postulated by cognitive theory. These assumptions are dysfunctional because they are rigid and so prevent the individual’s goal a t t a i ~ e nin t life; for example, a man who experiences unrealistically high parental pressure to achieve may believe that he is incompetent and develop the assumption that, in order tobe accepted, he has to be successfulall the time. It is thought that the onset of emotional disorders is often preceded by situations relating to individuals’ specific vulnerabilities: interpersonal eventsin dependent individuals and achievement events in achievement-orientated individuals (see Chapter 4). In the case of the man mentioned above, problems may ariseif he is trapped in a situation in whichhe cannot meethis lofty goals. Violationof the rigid rules or assumptions tends to be associated with extreme or excessive emotion. Depression comes about, for example, when adversity occurs that matches the person’svulnerabilit~threatening the person’s sense of self-worth (Beck, 3983). m e n in a highly emotionalstate such as depression, individuals tend to haw negative automatic thoughts which reflect the content of the idiosyncratic dysfunctional assumptions. Once depression sets in, the individual presents with the classic cognitive, behavioural and affective symptoms, Negative cognitions, depressive behaviour and depressive mood interact with eac
~ h a ~ t5e:A r cognitive theory of ~ e ~ r e s s i o n
other to maintain depression; for example, a depressed patient has negative thoughts such as ’I am lazy’, ’It’s useless trying anything’ and ’Everything is just too overwhelmingly difficult’. He lacks motivation and becomes inactive. As a result he has more negative self-denigrating thoughts. Hebecomes hopeless and more depressed.
ssi m e descriptive aspectsof the cognitive theory for depression consist of: the cognitive triad, thinking errors, and the reciprocal relationship between thought, mood and behaviour. These descriptive aspects of the theory simply propose that a groupof phenomena tends toexist in such a way thatit constitutes a syndrome. These phenomena disappear when patients are in remission and do not have any causal implications. In this section the details of these three concepts are explained, withan emphasis on the clinical implications of each.
The cognitive triad consists of the individual’s idiosyncratic view of self, the et al. (1979)postulated that in depression an indiworld and the future. Beck vidual often views himself as ’defective, inadequate, diseased or deprived. He tends to attribute his unpleasant experience to a psychological, moral or as physical defect inhimself’ (Beck et al., 1979, p. 11).Often he views himself et al.(1979,p. 11) described the depressed perworthless or inadequate. Beck son as someone who ’sees the worldas making exorbitant demands on him and/or presenting insuperable obstacles to reaching his life goals. He misinterprets his interactions with his animate or inanimate environment as representing defeat or deprivation’. Lastly Beck et al. (1979)described the depressed individual as someone who ’makes long-range projections, he anticipates thathis current difficulties or suffering will continue indefinitely. He expects unremitting hardship, frustration and deprivation’ (P. 11).The future seems hopeless andgloomy.
Cli~ical i~~lic~tio~s Clinically the cognitive triad provides very useful information for therapists. By paying attention to the cognitive triad, therapists can obtain a sense of how patients see themselves. One rule of thumb in cognitive therapy suggested by Safran et al.(1986)is that the important issues or automatic thoughts are those relatedto the senseof self and self-evaluation. By paying attention to the depressed individuals’ world, therapists can obtain much information
97
about interpersonal issues, and in viduals’ patterns of interpersonal tions can be relevant the to can interferewith therapy other peopleas critical may perceive the the this mayhinder the establishmentof predictor of suicide (Beck et al., 1975 84) and the person’s view of the future c oreover, if the individual conv~ced is that and will not change, he/she is not going to carry ou ment or even come back labelled and tackled first. In tack1 the patient to effect some positive change in one area, which is, in turn, used to challenge the patient’s belief that nothing will change.
There is much empirical evide cognitive triad; for example, de controls (Blatt et al., 1982); the ives (BradleyBr Mathews, 198 and they report lower self-este 1988). Furthermore, Bradley art ~dividuals’ negative viewsof and Myers et al. (1989)repor have less negative views of self comes from Brewin et recalled more negative tha depressed subjects were a adjectives applied to the rated themselves curre equally in both positi ’patient’s world view’ scored higher than no ’world’ subscaleof (1989)found that depressives endorsed less positive depressiort-relev~tworld adjectives and more negativedepression-releva~tadjectives comparedwith non-depressed controls. With respect to the future, depressed ~dividuals were found to be more hopeless about their future than controls or remitted depressives (HamiltonBr Abramson, 1983; Blackburnet al., 1986; Lam et al., 1987).
C ~ ~ ~5: tA ecognitive r t ~ e o r of y ~~ression
Beck (1967) hypothesized that systematic thinking errors of depressed patients maintain the patients’ validationof their negative concepts. This bias can occurdespite objective evidence to the contrary due to individuals making logical errors in their thinking. These errors include: selective abstraction (selecting abstract negative elements of a situation while ignoring other more salient featuresof the situation); arbitrary inference (drawing a conclusion in the absence of evidence or even when there is evidence to the contrary); overgeneralization; magnification(of negative aspects)and minimization (of the positive aspects); personalization; and dichotomous (black and white) thinking.
In cognitive therapy, patients are oftentaught that, for reasons which are not entirely clear, their thinking tends to be idiosyncratic and biased towards negatively evaluating themselves or construing their internal world. Patients are taught to catch and report their negative automatic thoughts. They aretaught how to cut through this strong emotional state and step back toevaluate these negative automaticthoughts. They areshown ways to examine evidence objectively for and against their negativethoughts and to look for alternative ways to explain upsetting situations. Cognitive therapists do, however, acknowledge that depressed patients can have real-life difficulties and so problem-solving is often employed when working with overwhelmed depressed patients with real-life difficulties. €mpirical evidence
The evidence that depressed individuals display greater negativity in their thinking is ample; for example, the using Automatic Thoughts Questionnaire (Hollon & Kendall, 1980), depressed subjects scored consistently higher than non-depressed subjects (Blackburn et al., 1986; Dobson & Shaw, 1986; Hollon et al.,1986; Lamet al., 1987). Gotlib (1981) found that depressed patients were significantly more negativeabout the feedback they had received than psychiatric non-depressed controls. DeMonbreunand Craighead (1977) found that depressed subjects significantly underestimated positive feedback. However, Beck’s theory does not just hypothesize that depressed individuals are negativein their thinking, it also proposesthat they make idiosyncratic thinking errors. The evidence for this is relatively scanty because it has been difficult distinguish to reliably between different types of evaluative distortion. For example, Krantzand H m e n (1979)reported that independerit
99
Chapter 5: A cognitive theory of depression judges found it difficult to distinguish reliably between certain depressive cognitive errors postulated by Beck such as arbitrary inference, selective abstraction and overgeneralization. Similarly,Cook and Peterson (1986)reported similar difficulties and they had to collapse together selective abstraction, overgeneralization and magnification and/or minimization into one category. However, there is some evidence of idiosyncratic thinking errors. Depressed subjects are less likely to give depressed-distorted responses according to the Cognitive Bias Questionnaire (Krantz & Hammen, 1979)~to justify their causal attribution logically and are more likely to make logical errors in making attribution to personal negative events in the past year (Cook & Peterson, 1986). Lefebvre (1981) measured four types of cognitive errors: catastrophizing, overgeneralization, personalization and selective abstraction. Lefebvre reported that depressed patients scored higher on cognitive errors. Wenzlaff and Grozier (1988) asked depressed and non-depressed students to indicate the extent to which they possessed anxious personality traits and the importance of these traits on a trait scale. Then subjects were given predetermined feedback of success or failure on an ostensible test of social perceptiveness. After feedback, the trait scale was re-administered. Depressed students inflated the importance of social perceptiveness after they had been told that they had done poorly on a test that purportedly measured social perceptiveness. Furthermore, they generalized from the social perceptiveness test results to an overall lowering of their general proficiency. This was taken as evidence of depressed individuals‘ tendency to self-derogate and to magnify their failings. Golin (1989) asked depressed and non-depressed students to judge the probabilities of positive or negative inferences on 10hypothetical situations being correct. The situations were other-inferencesor self-inferences. Depressed students judged negative self-inferences as more plausible than positive self-inferences. However, one criticism of the above studies is that the experimenter provided dimensions as the basis for inferences. This paradigm can be criticized as artificial and inducing such introspection may disrupt the typical negative, automatic or reflexive processes postulated by Beck et al. (1979).
Cognition, mood and behaviour In his cognitive theory, Beck (Beck et al., 1979) acknowledged that negative cognition may not be the ultimate ’cause’ of depression. In fact negative cognitions can be seen as a depressive symptom on a par with affective, behavioural and physiological symptoms. However, an attempt is made to ’make some sense out of the highly diverse phenomena of depression by arranging them in a coherent logical sequence’ (Beck et al., 1979; p. 18). Hence the patient’s negative constructions of reality can be postulated to be the first link 100
in the chain of de ressive symptoms. In the cognitive theory, CO (thoughts/images),mood and behaviour affect each other. Negative cognitions can maintain, if not lead to, depressive affect. Likewise, behaviour can distract from negative rumination and hence havean impact on c o ~ i t i o nan behaviour.
Theoretically, as mentionedabove,(1987) denied makin tion that owever, general1it is negative can and epression be an handle to deal with patients’ depression. In therapy,patients are important aspect of the theory in order to understand how emo nition or behaviour. Theyare asked to identify the when depressed in order to make sense of their em0 as described below, to examine idios~cratic the negativity of Similarly, patients are also taught to use certain behavioural skills such as pleasant activity scheduling in order to lift theirdepressed mood.
Evidence to support the viewthat mood can affect cognition comes from clinical studies and experimental studies that manipulate non-clinical jects’ mood states. Lloyd and Lishman (1975)reported that in depressed patients the latency for retrieval of unpleasant experiences correlated negatively with their level of depression as measured by Beck Depression I Clark and Teasdale (1982) took advantage of the diurnal variation of depressed patients to testthe effect of mood on recallof personal pleasant or unpleasant memories. Each patient was tested twice, once when in a more depressed phase and oncein a lessdepressed phase of the diurnal variation. Subjects were asked to retrieve a memory when with presented a neutral wor It was found that unpleasant memories were more likely to be retrievedin the moredepressed phase. Conversely, negative memories were less likely to be retrieved in the less depressed phase. Experimentalstudies generally make use of some mood induction procedure to induce a desired mood state and then to test subjects for recall bias. For example, Teasdale and Fogarty (1979) reported that after the subjects’ mood wasmanipulated by a moodinduction procedure in a student population, the latency of retrieval of pleasant events was significantly longer than unpleasant events when subjects were depresse Teasdale and Russell (1983) presented student volunteers with a list containing an equal number of positive, neutral and negative personality trait words. Then the subjects’ mood wasmanipulated by a mood induction procedure.
e authors found that subjects recalled more positive personality trait words hen inan elated mood and more negative personality trait words when in a
imilarly there is evidence that cognition can affect mood; for example, mood induction procedures such as reading standardized, negative selfreferent statements (Velten, 1968) can lead to a sad mood in 50--70% of subjects, depending on howstrict the criteria were for sadness. This can be seen as tentative evidence to support cognition leading to mood. Clark (1983), how ever, pointed out that in these studies typically subjects were not directly asked how they achieved their sad state. Furthermore, Teasdale an reported that depressed patients became more depressed ins~uctedto thinkof personally relevant negative thoughts. Conversely, there is evidence that when patients were distracted from negative cognitions, they ecame less depressed, For example, Teasdale and Rezin (1978) reported sinle casesof reducing frequencyof negative thoughts leading to a significant reduction of depressed mood. A later study (Fennel1 et al., 1987) replicated Teasdale and Rezids findings. In both studies, the reduction of depressed mood was associated with the reduction in frequency of negative thoughts, and patients who scored low on the Newcastle Endogenous Scale seemed to ave benefited more. Certain behaviours are reported tobe ‘antidepressant’ in nature. People are generally fairly prompt to offer adviceof ’the thing to do when you are depressed’. Rippere(1979) summarized her findings that there seemedbe to etween’ and ‘within’ individual differences in terms of helpfulness of antidepressant behaviours. Certain behaviour such as ’see a friend’, ’do something I enjoy’, ’talk to someone about how I feel’, ’keep busy’, seemedbeto generally helpful. However, ’take antidepressant’, ’crawl away on my own’, ’take a rest’, and ’read about depression and how it’, to seemed cure tobe less helpfu~.Furthermore, Rippere (1981) reported that subjects generally had a good idea of the extent to which people find certain antidepressant behaviour as helpful. Citing evidence that there is a large between-subject of each antidepressant activity, Rippere (1981) difference as to the helpfulness speculated that this common pool of knowledge was likely to be socially transmitted.
The descriptive aspectof the cognitive theory generally has good empirical support. However,it does not make any causal implications. It simply postulates that a group of affective, cognitive and behavioural symptoms coexist in depression. In this sense, negative emotion can be part of the depression syndrome. There has been much debate about the primacy of cognition in 10
the role of secure atter m ~ t ~ r iitnysocial
~ h ~ ~5: tAec or ~ n i ~ itheory ve o~~e~ression Main et al. (1985, pp.66-67) further refined the a~achment conceptby elaborating the ’internal working theory’. This was defined as ‘a set of conscious and/or unconscious rules for the organization of information relevant to attachment and for obtaining or limiting access to that information, that is, to information regardingattac~ent-relatedexperiences, feelings and ideations’. The general assumptionis that negative childhood experiences would serve as an internal theory or representation of self and others,which provide some sort of a ’blueprint’ for later relationships. It is presumed that people with attachment problems will have problematic relationships in later life.
The empirical evidence and clinical applications of attachment andearly adverse experience leading to vulnerability in childhood have been reviewed et al., 1978; BelskyNezworski, 1988; extensivelyelsewhere(Rajecki not repeated here. In fact Rutter Greenberg et al., 1990; Rutter, 1995) and willbe (1995) commented that most of the componentsof attachment concepts have been found to receive empirical support in studies of children. Furthermore, the general notionof insecure attachment leading to later adult problems is vague and difficult to test empirically.It needs to be better operationalized to yield to empirical testing. Clinically the most important concept of the attachment theory is the ’internal working theory’ that serves as a blueprint of further relationships in later life. At one level it makes sense that we have remembered and learned from past experiences which impinge on present relationships, at least in the behavioural repertory of interaction with others in our day-to-day contact and cognitive expectationsof people occupying different social roles. However, as Hinde (2988) concluded, problems arise in translating this general notion into more specific hypotheses that are amenable to empirical research.
A more specific and perhaps testable set of theses came from psychotherapeutic observation. Adverse early experiences in the form of perceived parenting style is linked with personality types that are hypothesized to be vulnerable to the development of depression when faced with stressful events in the environment. Before discussing the link between perceived parentin style and vulnerable personality types,it is important to clarify these personality types. or (1974) described two types of vulnerable individuals: those with anaclitic dependent traits, and those with introjective orself-critical traits. Similarly, both Arieti and Bemporad (1978) and Beck (1983)postulated two broad
C h ~ ~ t5:e A r co~nitivetheory of ~ ~ r e s s i o n subtypes of individuals at risk; those who focus on the interpersonal domain and those who focuson work or a similar achievement-related domain. The former is labelled as the 'dominant other' type and the latter as the 'dominant goal' type by Arieti and Bemporad (1978),and 'sociotropic' and 'autonomous' individuals, respectively,by Beck (1983). ~ t ~ as ~ 'dependent ~ ~ c on these Beck (1983) described the s ~ c ~individual social inputs for gratification, motivation, direction and modification of ideas and behaviour. The motif of this cluster is receiving' (p. 272). Arieti and Bemporad (1980) emphasized the failure of the 'dominant other' type to individuate and develop autonomy: They do not experience satisfaction directly from effort but only through an intermediary (anotherperson) who gives or withholds reward. They have formed an imagined agreement with the ... ,in important otherthat may be called a bargained relationship which the individual foregoes the independent derivation of gratification in return for the continuance of nurturance and support of the esteemed other. (p. 1360) Hence the'sociotropic', 'anaclitic or dependent' and 'dominant other' individuals are greatly reliant on close others as a sourceof meaning andselfesteem. Beck (1983) also described the a ~ tindividuals ~ ~ as attaching ~ ~ an irra~ ~ s tional amountof importance 'in preserving and increasing his independence, rhobility, and personal rights; freedom of choice, action and expression; protection of his domain; and defining hisboundary' (p. 272). Arieti and Bemporad (1980) also outline the 'dominantgoal' as individuals (who) invest their self-esteem in the achievement of some lofty goal and shun any other activities as possibly diverting them from this quest...The individual believes that the goal will transform his life and possibly himself. Attaining his desired objective will mean that others will treat him in a special way or that he will finallybe valued by others. (p. 1361) A premium is placed on meeting very stringent standards to inmainorder tain high self-esteem. Beck (1983) believed patients can change modes from issues of unloveability to incompetence between and even within episodes. Beck et al. (1990)labelled thesebeliefs of basic self-concept as core beliefs. Linguistically these are believed tobe encoded in the formof 'I am unlovable or incompetent'. All the above constructs attempt to spell out the characteristics of individuals who are vulnerable to depression and to types of stress which may precipitate a depressive episode in a vulnerable individual. The constructs appeared tobe similar even though the terms are different. More specific hypotheses relating to pathological parenting styles contrib-
~ h a ~p:tAeco~nitive ~ theory of ~ e ~ ~ e s s i o n uting to the v~lnerablepersonality types in later life have been proposed. cGlelland et al. (1953) pointed ‘out that ’value-achievers’ se ds and are significantlyin~uencedby social speculated that value-achievers develop these attitudes from ‘authoritarian pressure from parents to be ambitious and the resultant motives which origi~atedin external sources show itself as a fear of being unsuccessful’ ,419). Similarly, Gotlib and Hammen (1992) h ~ o t h e s i z e dthat individuals with insecure attachments may ependent personalities and overvaluing interpe~sonaldomains. A rent-child experiences also contribute to the ac~uisitionof ne ative self-schema, characterized by low self-worth. Moreover, these negative self-sche~alead to the development of further negative schemas of expectation of interpersonal negativity and rejection. These hypotheses are more in tune with the cognitive theory relating early experiences to later vulnerability.
The empirical evidence linkin e parenting with later adult epression has been review r~peating empirical the evidenc roblems in childhood and non-specific negative perceived concentrates on the evidence wh with later vulnerability to selfpendency in adulthood, assug As mentioned above, the vu ( ~ 9 7 4Arieti ) ~ and Bemporad(I pirically there is evidence that the variousins~umentsdesigned to measure theseconceptsoverlapto a l extent.Thethreebetter h o w n andmore
et al., 19 and htrojective DysAutonom~Scale (Beck ~ c t i o n aAttitudes l subscales by As 89). all thee scales are usedin the studies reviewed in this section and the following section on congruent events,it is worth examining how they relate to each other. latt et al., 1979) yielded three factors: de y. The dependency factor contains the donment, feeling lonely and wantingbetoclose to and dependent on others; the self-criticism factor contains themes about feeling guilty, failing to meet expectations and standards, and being unable to assume responsibility; the efficacy factor contains themes about high standards and personal inner strength, feelingsof independence and pride andsatisfac~onin one’s acco~plishments.The ~ociotropy-Autono~y Scale (SAS) (Beck et al., 1983)
yielded two factors of sociotropy and autonomy: the sociotropy factor relates to themes about disapproval, attachment and pleasing others; and the autonomy factor contains themes about achievement, freedom from control, and preference for solitude. The Anaclitic (dependent) and Introjective (selfcritical) Dysfunctional Attitudes subscales were taken from the original Dysfuntional Attitude Scale (DAS) ( ~ e i s s m a n& Beck, 1979) by ong grain and Zuroff (1989). Robins (1985) found significant correlations between the SAS sociotropy and the DEQ dependency (r=0.68); and SAS autonomy and DE criticism (~=0.39).Furthermore, Blaney and Kutcher (1991) fou intercorrelations between the DEQ, SAS and AnacliticDAS factor concerning dependency or sociotropy. However, the intercorrelations between the three scales assessing individuality or self-criticism were less clear. The correlation between Introjective DAS and DEQ self-critical factor was substantial and significant. However, the SAS autonomy factor did not correlate significantly with the DEQself-criticism.The authors speculated that theSAS a~tonomy factor appearedto measure counter-dependency rather than individuality. There is, surprisingly, very little evidence supporting this partof Beck's theory from the traditional cognitive therapy research. Mostthe ofevidence linking perceived problematic parenting and vulnerable personality types in later life comes from research using psychoanalytical perspectives. Given the absence of empirical evidence from the traditional cognitive therapy school and the high intercorrelations between the various instruments, this section concentrates on the more specific hypothesis of perceived parenting and vulnerable personality typesin later life using instruments suchas the DE Table 5.1 summarizes the studies examining specific perceived parenting relating to self-criticismor sociotropy. All six studies included assessed perceived parenting retrospectively. Two studies used longitudinal design to control for levels of depression and perceived parenting (Gotlibet al., 1988) and to assess trait self-criticism (Brewin et al., 1992b). Of thesix studies, five used non-clinical populations and there was only one study using a clinical population (Gotlib et al., 1988): this did not link perceived parenting with later adult vulnerability is type. included It for two reasons:first, it was longitudinal in design and hence could investigate whether retrospective recall of perceived parenting was a function of the level of depression; and second, the study is the only one using a clinical sample. The main findings were positive; for example dependent subjects remembered their mothers as dominating and controlling, emphasizing conformity (McCraine & Bass, 1984); maternal overprotectiveness significantly correlated with DEQ dependency in women (Blatt et al., 1992); dependency score in the DEQ was related to perceptionof a distant relationship with fathers durin
.l S u m a r y of studies ~ v e s t i g a perceived ~g parenting and vulnerable personality types. Sample
Authors
Inst~ment
Outcome
McCraine & Female nursing Depressive Experience Dependent subjects remembered their Bass (1984) students Questionnaire (DEQ; mothers as dominating and controlling, Blatt et al., 1979); Strict emphasizing conformity. Self-critical Control, Conformity and subjects remembered both of their Achievement Control parents as equally dominant, emphas~ing subscale from the Parentalachievement and performance, Behaviour Form (Kelly& demonstrating their love inconsistently Worrell, 1976); Parental Inconsistency of Love Scale (Schwarz& Zuroff, 1979); SchwarzGetter Interparental Influence Scale (Schwarz& Zuroff, 1979) Gotlib, et al. Longitudinal (1988) study of 30month followup of 25 postpartum, depressed and 25 matched postpartum, non-depressed women
Beck Depression Inventory (Becket al., 1961); Parental Bonding Instrument (Parkeret d . , 1979)
At bothTl and T2 depressed and remitted women reported higher overprotection scores than non-depressed women; at T1 and T2 non-depressed and reinitted women reported lower parental caring scores than depressed women
Frost et al. (1991)
41 women graduates and their parents
Multidimensional Perfectionism Scale (Frost et al., 1990); Fatherlmother Trait Scale (Steketeeet al., 1985); Brief Symptom Inventory (Derogatis& Melisaratos, 1983)
Mother’s perfectionism, correlated with the perfectionism in their daughters; daughters’ perceptionof fathers’ harshness correlated with daughters‘ perfectionism. A combination of mothers’ own perfectionism scores and mothers’self-reported harshness accounted for30% of the daughters’ Perfectionism. Mothers’ perfectionism, was associated with daughters’ psychopathology
Blatt et al. (1991)
Non-clinical young adults
Beck Depression Inventory (Beck etal., 1961); Parental Bonding Ins~ument(Parker et. al., 1979); Depressive Experience Questionnaire (DEQ; Blatt et. al., 1996); Zung Depression Scale (Zung, 1965)
In women, caringby mother and father had a significant negative correlation with DEQ self-criticism; maternal and paternal overprotection correlated with DEQ self-criticism andBD1 and Zung depression scores; maternal overprotectiveness significantly correlated with DEQ dependency
In men, mother’s caring significantly
correlated with in only the negative direction; paternal caring correlated significantly only with DEQ selfcriticism in negative direction; Maternal overprotection correlated with DEQ self-criticism (notso clear in men)
C h ~ ~ t5:e A r cognitive theory of d ~ r e s s i o n
Authors
Instrument
Outcome
Brewin et al. 75 medical (1992b) students assessed twice with an average interval of 30 months
Depressive Experience Questionnaire(DEQ; Blatt et al., 1976); Family Attitudes Questionnaire (Thomas & Duszynski, 1974);Parental Bonding Inventory (Parkeret al., 1979) and Symptom Checklist-90 (Derogatis et al., 1983)
After controlling for mood and social desirability, high levels of trait selfcriticism were related to retrospective reports of less satisfactorymaternal parenting in terms of level of care and overprotection as well as measures from the Family Attitude Questionnaire
Rosenfarb
Parent-Child Relations Questionnaire (Siegelman & Roe, 1979); Depressive Experiences Questionnaire (DEQ; Blatt et al., 1976)
Dependency scorein theDEQ was related to perceptionof distant relationship with fathers during development, perceptionsof increased parental attention and overindulgence. Self-criticism scoresin theDEQ were associated with perceptionsof difficulties of affective bond with fathers and peers during childhood, and perceptions of increased paternal power andcontrol
et al.
(1994)
Sample
132 female subjects
development and perceptions of increased parental attentionand overindulgence (Rosenfarbet al.,1994). In relationship to self-criticism, self-critical subjects remembered both of their parents as equally dominant, emphasizing achievement and performance and demonstrating their love inconsistently (McCraine & Bass, 1984); mothers’ perfectionism correlated with the perfectionism of their daughters, and daughters’ perception of fathers’ harshness correlated with daughters’ perfectionism (Frost et al., 1991); maternal and paternal overprotection correlatedwith DEQ self-criticism (Blattet al., 1992); high levelsof self-criticism were related to retrospective reports of less satisfactory parenting and less satisfactory relationships with their parents (Brewin et al., 1992b); and self-criticism scores in theDEQ were associated with perceptions of difficulties of affective bonds with fathers and peers during childhood, and perceptions of increased paternal power and control (Rosenfarb et al., 1994). In theirstudy, Brewin et al.(1992b)investigated self-criticismas a trait and found that subjects who were highly criticalat both time 1 and time 2 had significantly worse maternal relationships than other subjects. The authors also reported that the correlation between perceived parenting and selfcriticism was much reduced when the level of depression was controlled. However, the authors raised the issue that, if early adverse parenting influences both depression and self-criticism in later adult life, controlling for depression would underestimate the relationship between perceived problematic parenting and self-criticism.
C ~ ~5: A ~ cognitive t e ~ theory of ~ e ~ ~ e s s i o n
There was some evidenceof a gender-specific effect. Threestudies used 8t Bass, 1984; Frost et al., 1991; Rosenfarb exclusively female subjects (McCraine et al., 1994). Blattet al. (1992) investigated the gender effect and commented that the relationship between perceived parenting style and a lateradult vulnerable personality typein male subjects was not so clear in their study. In men, mothers’ caring significantly correlated with the Beck Depression Inventory (BDI)in the negative direction; paternal caring correlated si~ificantly only with DEQ self-criticism in negative direction, and maternal overprotection correlatedwith DEQ self-criticism. The evidence so far has supported the link between perceived problematic parenting style and later vulnerability in adulthood. However, the evidence reviewedhas come from retrospective recall of perceived parenting. In retrospective studies, there is a possibility that depressed subjects might report negative parental behaviour as a function of depressive mood. However, it has been argued (Johnsonet al.,1982)that depressivesdo not have a global and indiscriminate tendency to report their childhood experiences in a negative way. This is borneout by the fact that dependent and self-critical individuals did not report indiscriminately on global negative perceived parent In the study by Gotlib et al. (1988), overprotection and the level of parental care did not change when subjects’ depressionhad remitted. In a review article on a reappraisal of retrospective reports (Brewin et al., 1993)~ it was argued that the general unreliabilityof retrospective reports was exaggerated. Furthermore, one could argue that it was the perception of one’s own parents, rather than the actual behaviour of one’sparents, that form one’s internal working theory, which, in turn, affects one’s interaction with the world. The major problems with the majority of the studies reviewed include the lack of evidence from clinical subjects,the comparative lack of control for mood state and a lackof investigations intowhether these vulnerable personality types are truly trait and not state findings. Despite these weaknesses, the findings between perceived parenting and later adult vulnerable personality types seems robust. ~~lnera~le
p e r s types Q n a and l i ~ c Q n ~ ~ eevents nt
Beck (1983)hypothesized that people weavean environment around them in which they can cope. For example, highly sociotropic individuals would invest to the best of their abilityin interpersonal domains toensure that there is someone closein order to feelvalued. Similarly, highlyautonomous individuals would invest a lot in achieving whatever important is to themin order to be self-sufficient. Furthermore, they develop certain beliefs in response to these core beliefs. Clinical problems are seen as a result of an event, either social or 110
Cha~tey5: A cognitive theory o ~ ~ ~ y e s s i o n cognitive, which impinges on a specific vulnerability relating to either issues of loveability or competence:the congruent hypothesis.Both the core belief and conditional beliefs may be activated and further mould subsequent interpretations of an event. This will be followed by intense and emotionally laden automatic thoughts and the person may develop clinical depression.
Table 5.2 summarizes the empirical studies that involved identifying vulnerable personality typesand investigating whether undesirable events relating to the particular vulnerable personality type (congruent events) lead to emotional problems. Fourteen studies relating to vulnerability types and congruent events are summarized. Eight studies use student samples. Of the eight student sample studies, four are prospective studies (Hammen et al., 1985; Zuroff et al., 1990; Lakey & Thomson-Ross, 1994; Robins et al., 1995). The study by Rude and Burnham (1993) is a cross-sectional study because there was no measurement of depression symptoms at T (at recruitment). In thestudy by Hammenet al. (1985), in addition to the BDI, a formal diagnostic interview was conductedand depression was diagnosed using DSM-I11 (American Psychiatric Association, 1980) Major Depressive Disorder criteria. However, all the other seven studies used either BD1 or the DEQ (Blattet al., 1976) asan indication of depression. Moreover, the instruments used to measure vulnerable personality types ranged from recall procedures by Teasdale et al. (1980); Sociotropy-Autonomy Scale (Beck et al., 1983); to different versionsof a Dysfunctional Attitude Scale. Taken as a whole, the 14 studies summarized in Table 5.2 support the notion that sociotropy or dependency is related to depression. The only exception was the study by Hammen et al. (1992) in which bipolar patients were used. Only two studies (Hammen et al., 1989; Segal et al., 1992) found interaction between autonomy, negative achievement events and relapse. The study by Hammen was a small sample and the number of subjects relapsed was even smaller. Of more interest is thestudy by Segalet al. (1992) which needsfurther replication. The empirical evidence of congruent events and vulnerable personality types seems more complicated. The timing of the eventsappears important for the two different typesof event. Segalet al. (1992) found that relapse was related to the interaction of self-criticism scores and negative achievement events over the whole year but the predicted effect for dependency-orientated patients was only valid when events inthe 2.monthsprior to onset were taken into account. The authors’ interpretation was that the impactof interpersonal events was more immediate, whereas achievement-related events might involve worsening of ongoing difficulties over a period of time. l11
~ognitivetheory of depression Table 5.2 Summaries of studies investigating congruentevents and depression. Authors
Subjects
Instrument
Results
Hammen et al., 1985
4-month follow-up of a student sample of 46 dependent and 32 self-critical schematic individuals
Recall procedures by Teasdaleet ai. (1980); Beck Depression Inventory (Beck, 1967)
Dependent subgroup showed strong associations between depression and schema-relevant interpersonal events. The predicted patternfor self-critical schematics was observed but was less often statistically significant
SociotropyAutonomy Scale (Becket al., 1983); Beck Depression Inventory (Beck, 1967)
Weak support for the specificity of type of event andsociotropy versus autonomy individuals. Autonomy was unrelated to depression and had shown no relationship to any type of event categorized a priori, whereas sociotropy was associated with depression level and served asa moderator of the relationship between depression andfrequency of negative social events. After partialling outthe main effect for sociotropyand for negative social events,the interaction between these two variables accounted for an additional7% of the variance in theBeck Depression Inventory (BDI), a significant increase in the variance accounted for. However,there wasalso a significant increasein theBD1 variance accounted for by the interactionof sociotropy and negative autonomy eventsafter partialling out their additiveeffects
Robins and Block Cross-sectional (1988) study of another student sampleof 45 men and 53 women
Hammen et al. Currently Sociotropy(1989) Autonomydepressed autonomousand Scale(Beck et al., 1983) sociotropic unipolar (22 patients) and bipolar (25 patients) depressed patients
Onset of relapse or exacerbation of symptoms in unipolar depression in autonomous subjects who had experienced more achievement events and sociotropic subjectshad experienced moreinterpersonal events. However, congruence hypothesis did not apply for bipolar patients
Segal et al. (1989)
6-month followDAS”need for up of 40 remitted approval depressed patients and self-critical factors
Congruent eventscorrelated significantly with depressive symptoms during follow-up for dependent subjects but no significant correlation forcongruent events for self-critical subjects
Robins (1990)
41 depressed patients and 44 schizophrenic patients
Highly sociotropic patients reported more recent negative interpersonal events than negative autonomy events. Autonomous subjects did not report No more negative autonomy events. evidence of any congruence among non-depressed schizophrenic patients
SociotropyAutonomy Scale (Becket al., 1983)
Subjects
Authors Zuroff et al. (1990)
66 female undergraduates retested 12 months later
Depressive Experience Questionnaire (Blatt et al., 1976); Dysfunctional Attitude Scale (Weissman & Beck, 1979); Beck Depression Inventory (Beck et al., 1961)
After controlling for initial level of depression, majorityof dependent and self-critical subjects’ worst period involved interpersonal events. Dependency predicted anaclitic state depression and self-criticism predicted introjective state depression
Clark et al. (1992)
Cross-sectional study of 600 psychology students
SociotropyAutonomy Scale (Becket al., 1983; Beck Depression Inventory (Beck et al., 1961)
For dysphoric subjects(BDI> 16), sociotrophy significantly interacted with negative social related events but not autonomously related events to predict dysphoria. Autonomy failed to show any significant relation with dysphoria or types of negative life events experienced
Segal et al. (1992)
A sample of 59 Dysfunctional remitted depressed Attitude Scale patients for 1 year (Weissman& Beck, 1979)
Relapse was related to the interaction of self-criticism scores and negative achievement events over the period prior to relapse. The predicted effect for dependency-orientated patients was only valid when events in the 2 months prior to onset were taken into account
Hammen et al. (1992)
Followed up a sample of 49 bipolar patients over 18 months
SociotropyAutonomy Scale (Beck et. al., 1983)
No effect of congruent events in subjects suffering from bipolar depression. However, symptom severity was significantly associated with sociotropy, interpersonal events and the interactionof the two but not for autonomy
Rude and Burnham (1993)
443 studentsdepression symptoms and life events were measured 4 to 6 weeks later. (A cross-sectional study because no measurement of depression symptoms atT1)
Depressive Experience Questionnaire (Blatt et al., 1976); SociotropyAutonomy Scale (Becket al., 1983); Dysfunctional Attitude Scale (Weissman & Beck, 1979); Beck Depression Inventory (Beck et al., 1961)
With the exceptionof DAS, all interpersonal/sociotropy scales showed interaction with frequency of interpersonal events but not achievement life events in predicting depressive symptoms. The achievement/autonomous scale yielded no significant interactions with achievement life event frequencies
Lakey and Thomson-Ross (1994)
133 college students in a prospective study of 10 to12 weeks follow-up
Depressive Experience Questionnaire (Blatt, et al., 1976); Sociotropy-
After controlling for initial level of depression, dependent subjects displayed greater increase of depressive symptoms after interpersonal event but not (Continued on p. 224.)
ults Instrument Subjects
Authors Autonomy Scale (Beck et aZ., 1983); College Student Life Events Schedule (Sandler & Lakey, 1982); Beck Depression Inventory (Beck et al., 1961)
achievement events. However selfcritical subjects also displayed an increase of depressive symptoms after interpersonal events. There was only a trend for self-critical subjects to display greater increase in depressive symptoms after achievement events. Dysphoric subjects displayed greater increases in depressive symptoms following interpersonal but not achievement events. When the effect of initial dysphoria and interpersonal events was controlled statistically, the power of cognitive variables to predict reactivity was reduced greatly
Bartelstone and 166 Trull(1995) undergraduates including 42 identified as depressed by BD1
Depressive Experience Questionnaire (Blatt et al., 1976); SociotropyAutonomy Scale (Becket al., 1983); Life Experience Scale (Sarason et al., 1978)
Evidence to support congruent event hypothesis by DEQ dependency and SAS sociotropy subscale only
Robins et al. (1995)
164 undergraduates in a 1-month longitudinal study
Revised Depressive Experience Questionnaire (Welkowitz, et al., 1985); SociotropyAutonomy Scale (Beck et al., 1983); Life Experiences Survey (Sarason et al., 1978); Zung Self-rating Depression Scale (Zung 1965); Beck Depression Inventory (Beck et al., 1961)
Depressive symptom change was predicted by interaction of sociotropy and frequencyof negative interpersonal events. However, sociotropy also interacted with ac~evementto predict depressive symptom change. Similarly autonomy also interacted with both achievement and interpersonal events to predict changes in depressive symptom level. On the whole, findings did not support domainspecific hypotheses. Negative interpersonal events were strongly associated with subjects high in either autonomy or sociotropy
Lam et al. (1996)
1-year follow-up of a sample of 37 depressed patients
Dysfunctional Attitude Scale (DAS; Poweret al., 1994); Roles and Goals Questionnaire (Lam & Power, 1991)
Adversity in the most invested domain (matching adversity) according to the Roles and Goals Questionnaire and the DAS dependency subscale contributed significantly both to whether or not subjects relapsed and to the number of weeks subjects survived before they relapsed. Most matching adversity were interpersonal events. DAS achievement subscale did not contribute significantly to relapses
Hammen et al. (1992) found that sociotropy may relate tosymptom severity. The authors argued that with bipolar individuals, the theory of congruent, meaningful, personal events and relapse may not hold. However, once depressed, patients with a sociotropic personalitytype who have experienced negative interpersonal events and who define their worth in interpersonal relationships may experience more severe depression symptoms, the whereas autonomous individuals may even minimize the severity of their depression. Another complication relates to the conceptual aspectsof vulnerable personality types. Haagaet al. (1991) raised the issue of whether sociotropy and autonomy should be construed as dimensional or categorical constructs. Apparently about 37% in the study by Hammenet al. (1992)and 65% in that by Goldberg et al. (1989) were ’mixed’ or unclassifiabl~.Cognitive theory does not make predictionsabout them. It could that be individuals who have invested in both interpersonal and achievement domains are less prone to developing depression. Linville (1985) proposed that people vary in the degree of self-complexity. Self-complexity is defined as a function of two factors: the number of self-aspects and the relatedness amongst them. Greater self-complexity is defined as ’cognitively organizing self-knowledge in terms of a greater number of self-aspects and maintaining greater distinction among self-aspects.’ The basic assumption is that greater self-complexity moderates the adverse impact of stressful events.For people who have more self-aspects and maintain more distinctions among self-aspects,the impact of negative events is likely to be confined to a smaller ofaspect their self-representation. Cognitive theoryof vulnerable personality types is based on clinical observation and may not generalize tothe non-depressed population. Furthermore, the measureof the stressor (negative events) can also be improved. One could measure the subjective perceptionof levels of stress, the frequency counts of these negativeevents or the contextual rating of events by the Life Eventsand Difficulties Schedule method (Brown & Harris, 1978). Furt~ermore,studies have theirown problems; for example, the problem with the early study by Hammenet al. (1985) is that the initialgroupings were based on reported previous significant upsetting events, When subjects were assessed 2 months later, the kappa coefficient was only 0.30, suggesting the classification was temporally unstable. Therewere also differencesin terms of the stringency with which the statistical analysis was carriedout. In the Hammen et al. (1985)study, the statistical analysisby Hammen’s team was of the frequency of events. Presumably theauthors could argue that it was not clear in Beck’s formulation what level of intensity of events was needed for the onset of depression. Whereasin the other studies (Robins & Block, 1988; Clark et al., 1992; Rude & Burnham, 1993) the authors used more stringent statistical testssuch as multiple regression analysis MANOVA or to examine the contribution measuresof autonomy and sociotropy and their interactionwith congruent events can make the to depression measures.
To sum up, the empirical evidence for congruent events and the onset of a depressive episode for sociotropic or dependent subjects is robust. The majority of the studies reviewed here have produced positive results. However, the evidence is not so clear for the concept of autonomy or selfcriticism; onlya handful of studies produced positive results. This could partly be due to the fact that the instruments, such as the SAS, are notvery good at measuring autonomy. However, all the studies involving patients are of small samples, ranging from 59 to 37 depressed subjects. Again, H a ~ e n et al. (1989) used a frequency count as opposed to the interaction of measures of vulnerable personality types and congruent events. Moreover, these studies examined the impactof congruent events on relapses or recurrences of depression as opposed to the onsetof the first episode of depression. Hence one can only conclude from these studies that a vulnerable personality and congruent events as a trigger of a depressive episode are only applicable to relapses. There is no evidence relating to the onset of a first episode of depression.
As mentioned above, dysfunctional assumptions are asseen a permutation of the basic core beliefs in order to cope with the internal worldin which the individual lives. Hence themes tendto be related to loveability, acceptance, achievement or control. Each patient has a unique set of dysfunctional assumptions which are trait-like and can be latent during a remission. These assumptions are dysfunctional because they are highly rigid and absolutistic and hence prevent the individual’s goal a t t a ~ e nint life.It is through these assumptions that ’the individual attempts to intergrate and assign valueto the raw data...In essence, these basic assumptions forma personal matrixof meaning and value, the backdrop against which everyday events acquire rel et al., 1979, p.244). Beck et al. (1979) evance, importance and significance’ (Beck hypothesized that people with rigid and extreme dysfunctional assumptions are more prone to depression when faced with relevant adversities.
There havebeen three waysto investigate the trait-like vulnerabilityof dysfunctional attitudes:(i) cross-sectional studies comparing remitted depressives with normalsubjects; (ii) longitudinal studies following a sample of remitted depressives and investigating the relationshipof relapse and dysfunctional attitudes; and(iii) treatment studies. Cross-sectional studies comparing normal controls and recovered depressives have largely yielded negative results. Hamilton and Abramson
(1983), Evans and Rush (1984), Silverman et al. (1984), Simons et al. (1984.) and lackburn and Smyth(1985) found no differencein the DAS scores between remitted depressives and those who had never been depressed. However, longitudinal studies have yielded more evidence for nerability theory (Beck et al., 1979). Dysfunctional attitudes we predict the subsequent relapse following the successful treatment of depression (Evanset al., 1985; Simonset al., 1986; Rushet al., 1986). Lamet al. (1996) reported that in their 1-year follow-up, high levels of dysfunctional dependency attitudes assessed during the index episode when subjects were mildly depressed were associated with shorter survival time. Subjects with higher levels of dependency attitudes on the whole relapsed sooner, even with the effect of adverse events controlled. One explanation for the different findings of cross-sectional and longitudinal studies is that these dysfunctional attitudes remain latent until there is a slight mood disturbance or stress amongst the ’reactives’ (Williams, 1984). Hence when subjects are fully remitted, they do not score differently from those who had never been depressed. However, most subjects who havejust recovered from an episode of depression have some residual symptoms.It is therefore possibleto identify subjects who are more prone to relapses. This explanation is consistent with findings which use mood induction. Miranda and Persons (1988) reported their useof mood induction procedures to distinguish ’reactives’ and ’non-reactives’. Mood inroduced a dysphoric mood in both never-depressed and remitted es. However, only the remitted depressed end sed higher scores on dysfunctional attitudes. These studies provide some evidence of dysfunctional assumptions being latent and then activated only by negative emotional states. Turning to treatment studies, Simons et al. (1984) found that when depressed subjects recovered, their change in cognitive measures, including their levels of dysfunctional attitudes, were similar irrespective of whether they received pharmacological treatment or cognitive therapy. The authors postulated that different treatment modalities may work by attacking different facets of a complicated interlocking picture. Williamset al. (1990) examined whether dysfunctional attitudes lead to a predisposition to longer-lasting episodes after the onset. They found that more patientsin the high dysfunctional attitude scores range remained depressed after 6 weeks. In summary, there is empirical evidence that dysfunctional assumptions can predict treatment outcome, subsequent relapse and survival time. Furthermore, there is tentative evidence of dysfunctional assumptions being traitlike and activated differentially. This has led Teasdale (1988) to suggest that vulnerability liesin the individual’s difference in the patternsof information processing activated in depression: the differential activation hypothesis. Events interpreted as highly aversive and uncontrollable would lead to
transient or mild depressed mood. However, thexninority who progress to develop depression would access memories or representations of depressing experiences, and the negative interpretative categories and constructs. Hence their depressionis often enhancedby negative nations even though there is no immediate environmental input.
Cognitive therapists use a variety of cognitive and behavioural techniques. However, irrespective of the therapeutic techniques used, it is the cognitive theory that distinguishes cognitive therapy from other of psychotherapy. types it^ this theory trainee therapists are taught to conceptualize patients’ problems as soon as they begin to work with them. Clinically the causal aspects of the cognitive theory are used for case conceptualization. Case conceptualization is specific, although the actual content is individualized. Person (1994) defined case conceptualization as ’the therapist’s hypothesis about the nature of the psychological mechanisms underlying the patient’s difficulties’ (p.33). Therapists make use of the case conceptualization in order to have access to the patient’s internal world, to provide acdurate empathy, to guide intervention strategies and to predict the patient’s response to intervention. Concep~alizationconsists of making sense of all the longitudinal and cross-sectional da the patient presents in order to understand his/her overt problems. The theory postulates that core beliefs are developed in relation to early experiences, particularly early salient events. Hence therapists take note of any salient childhood events such as traumas and separation, patients’ perceived parental attitudes, relationships with peer groups and authority figures and any history of abuse. History of insecure attachment sug dependency or sociotropic issues, whereas authoritarian pressure from significant figures tobe ambitious in earlylife may lead therapists to hypothesize autonomous vulnerability. The therapist also tries to identifythehow patient has always viewed him/herself,what sortof problems the patient has experienced so far as a result of this corebelief and how thepatient has coped with difficulties. Furthermore, the cognitive theory of emotional state describes how clinical problems arise, The onset of emotional disorders is often preceded by situations relating to individuals’ specific vulnerabil~ties; for example, interpersonal events in dependent individuals and achievement events in achievement-orientated individuals. Hence, to understand how patients have reached the stage they are at, therapistsalso need continuously to ask questions about vulnerabi~ities and undesirable events. In cognitive conceptualization, the undesirable events do not have to be major social 118
adversities. They can be cognitive events that carry special meanings for patients. Dys~nctionalassumptions are important aspects of the cognitive theory as they represent part of the patient’s vulnerability to emotional problems. Safran et al. (1986) described a useful technique for eliciting core cognitive dysfunctional assumptions: the vertical exploration technique. Vertical exploration is the explorationof the individual’s idiosyncratic wayof construing events and their perception of the self. Hence therapists may want to explore the meanings of automatic thoughts as well as challenging them. Another important wayto elicit dysfunctional assumptionsis to lookout for patterns of automatic thoughts central for the sense of self and how patients interact with their environment. The process tends be to collaborative. As therapists and patients work collaboratively to agree on the idiosyncratic dysfunctional assumptions, attempts are made to modify these rigid and absolutistic rules intellectually. Furthermore, patients are encouraged also act to against these so that thereis rules in such a way as to have an experience of a different kind behavioural, emotional and intellectual learning.
The causal aspects of the cognitive theory for depression have been reviewed. The evidence supporting a pathogenetic parenting style and the formation of vulnerable personality types has mainly come from researchersof the psychoanalytic perspective and from retrospective recall of perceived parenting of non-clinical populations. The findings seem robust and consistent. However, most studies did not control forthe effect of the level of depression on such retrospective recall. Hence it is not clear whether such recall is a of result mood. Turning to studies of vulnerability types and congruent events, thereis enough empirical evidence tentativelyto support the notion that congruent events are more likely to lead to relapses in depression. Despite the fact that instruments used in the studies reviewed above to measure autonomy and sociotropy were not identical, the effect seems be robust. to The expansion of the cognitive theoryof congruent events has advanced the causal theory to an interaction between the intrapsychic and the environment. However,it is still hard to determine the extentto which subjects contribute to the occurrence of an adversity. Perhaps subjects with higher levels of dysfunctional This speculation attitudes are more prone to bring about their own adversity. is supportedby the findingof Simons et al. (1993) that elevated scores on DAS were associated with the number of objectively defined events occurring prior et al. (1987) have found that the to the onsetof depression. Furthermore, Lam level of dysfunctional attitudes predictedsubjects’ length of survival before
relapsing due toan adverse event. However, most of the evidence to support so far are fromevisociotropy and autonomy as vulnerable personality types dence of relapses in patient populations.It is hard to generalize such findings to the onsetof a first episodeof depression. It is possible that the experience of functioning so much that the of depression has lowered the person’s level person can only invest in one aspect of his/her life. To sum up, the bulkof the evidence supporting the causal aspects of Beck‘s cognitive theory for depression either come from cross-sectional or retrospective studies. To test the causal aspects of the cognitive theory, longi~dinal studies with long-term follow-up of never-depressed subjects are required.
Belsky, J. & Nezworski, T. (eds). (1988) ClinicalImpljcation of Attflchmen~.Erlbaum Associates, Hillsdale, NJ. Blackburn, I.M. & Smyth, P. (1985) A test of cognitive vulnerability in individuals prone to Depression: The Psychotherflpeutic Approflch. depression. British Journal of ClinicfllPsychology, Tavistock, London. 24, 61-62. Arieti, S. & Bemporad, J. (1980)The psychological Blackburn, I.M., Jones, S. & Lewin, R.J.P. (1986) organisation of depression. American Journfll of Cognitive style in depression. British Journfll of Psychiatry, ~7,1360-1365, Clinical P s y c ~ o l o25,241-251. ~, Personality, life Blaney, P.H. & Kutcher, G.S. (1991) Measures of Bartlestone,J.H. & Trull, T.J. (1995) events and depression. ~ournfllof Personfl~ity depressive dimensions: are they interchangeable? Journal of Personality Assessment, 56, 502Assessment, 64,279-294. Beck, A.T. (1967)Depression:Clinicfll,Experimental, 512. and T~eoreticfll Aspects. Harper & Row, New York. of object representation in Blatt, S.J. (1974) Levels Beck, AT. (1983)Cognitive therapy of depression: anacliticand introjective depression. Psyc~oflnflEn: Trefltmentof D~ression:Old New perspectives. lytic Studies of the Child, 29,107-157. Controvers~s and NewA ~ r o a c(e& ~ s P.J. Clayton Blatt, S.J. & Homman, E. (1992) Parent-child in& J.E. Barrett),pp. 5-37. Raven Press, New York. teraction in the etiology of dependent and self12, Beck, A.T. (1987) Cognitive models of depression. critical depression.Clinical Psychology Review, ~ o u ~off lCognitive l Psychotherflpy,I,5-37. 47-91? Beck, A.T., Ward,C.H., Mendelson, M., Mock,J. & Blatt, S.J., DAfflitti, JP. & Quiulan, D.M. (1979). Erbaugh, J. (1961)An inventory for measuring Depressive Experience Questionnflire. Unpublished manual. Yale University,New Haven, CT. depmsion. Archives of General psych^^, 4, 53-63. Beck, A.T., Kovacs, M. & Weissman, A. (1975) Blatt, S.J.,Quinlan, D.M., Chevon, E.S., McDonald, Hopelessness and suicidal behaviour: an overC. & Zuroff, D, (1982) Dependency and selfview. Journfllof the American Medical Associution, criticism: psychologicaldimensions of depresl and ClinicalPsychology, sion. ~ o u ~off lConsul~ing 234, 1146-1149. Beck,A.T., Rush,A.J., Shaw, B.F. &Emery, G. (1979) 50, 113-124. Cognitive Therflpyfor Depression. John Wiley & Blatt, S.J.,Quinlan, D.M. & Bers, S. (1992) DepresSons, New York. sion and the representation of parental behavBeck, A.T.,Epstein, N., Harrison, R.P. & Emery, G. iour in normal young adults. In: Pflrent-ch~~d (1983) Development of the Sociotropy-Au~onomy interaction in the etiology of dependent and selfScale: A Measure of Personfllity Factors in Decritical d ~r e s s io nClinical . Psychology Review, 12, pressjon. University of Pennsylvania Press, 47-92. Philadelphia. Bowlby, J. (1969) Attflchme~t and Loss, Vol. 1. AtBeck,A.T., Freeman, A. et u1. (1990) Cognitive New York. t ~ c h m(md ~ t edn, 1982). Basic Books, Therflpy of Personality Disorder. Guilford Press, Bowlby,J. (1973)Attflchmen~ and Loss, Vol. 2. SeparaNew York. tion, Anxiety and Anger.Hogarth Press, London.
American Psychiatric Association (1980) Diagnos tic and S~fltisticflr Manual mental Disorders, 3rd edn. American Psychiatric Association, W a s ~ g t o nDC. , Arieti, S. & Bemporad, J. (1978) Severe and Mild
l20
~ h a p ~5:e A r ~ o ~ n theory i ~ i ~o~~epression e Bowlby, J. (1977) The making and breaking of Derogatis, L.R. & Melisaratos, N.(1983) The Brief affectionate bonds: Part 1, Etiology and psychoSymptom Inventory: An introductory report. PsychologicalMedicine, 13,595-605. pathology in light of attachment theory. British Journal of Psychiat~,130,201-210. Dobson, K.S. & Shaw, B.F.(1986) Cognitive assessBowlby, J. (1980) Attachment and Loss, Vol. 3. Loss, ment with major depressive disorders. Cognitive Separation and Depression. Hogarth Press, LonTherapy and Research, 10~13--29. don. Dyer, J.A.T. & Kreitman, N. (1984) Hopelesness, Bowlby, J.(1988) A Secure Base:CZinicalImplications depression and suicidal intent in parasuicide. of Attachment Theory. Routledge & Kegan Paul, British Journal of Psychiatry, 114, 127-133. London. Evans, G. & Rush, A.J. (1984) Cognitive patterns in symptomatic and remitted unipolar major Bradley, B. & Mathews, A. (1983) Negative selfschemata in clinical depression. British Journal depression. Journal of Abnormal Psychology, 83, 31-40. of Psychiatry, 130,201-210. Bradley, B.& Mathews, A.(1988) Memory bias in Evans, M., Hollan, S.D., DeRubies, R.J., Piascki, recovered clinical depressives. Special Issue: In- J.M., Tuasson, V.B. & Vye, C. (1985) Accounting formation processing and emotional disorders. for relapse in a treatment o u t c o ~ study e of depression. Paper presented at the Annual Meeting of Cognition and ~ m o ~ i o2,235-245. n, the Association for the Advancement of BehavBrewin, CR., Smith, A.J., Power,M. & Furnham, iour Therapy, November. A. (1992a) State and trait differences in depressive self-perceptions.Behaviour Therapy and Re- Fennell, M.J.V., Teasdale, J.D., Jondes, S. & Damle, A. (1987) Distractions in neurotic and endogsearch, 301555-557. Brewin, C.R., Firth-Cozens, J., Furham, A. & enous depression: an investigationof negative McManus, C.(1992b)Self-criticism in adulthood thinking in major depressive disorder. Psychoand recalled childhood experience. J o ~ r n a of logy Medicine, 17, 441-452. l Abnormal Psychology,101,561-566. Frost, R.O., Marten, P., Lahart, C.M.& Rosenblate, Brewin, C.R., Andrews, B. & Gotlib, I.H. (1993) R. (1990) The dimensions of perfectionism. CogPsychopathology and early experience: a reap- nitive Therapy and Research, 14, 449-468. praisal of retrospective reports. PsychoZog~cal Frost, R.O., Lahart, C.M. & Rosenblate, R. (1991) Bulletin, 113,82-98. The development of perfectionism: a study of Brown, G.W. & Harris, T. (1978) Social Origins of daughters and their parents.Cognitive Thera~y Depression.Tavistock Publications, London. and Research, 15,469-489. Clark, D. (1983) On the induction of depressed Goldberg, J.O., Segal, Z.V., Vella, D.D. &Shawl B.F. mood in the laboratory: evaluation and compari-(1989) Depressive personality Milton Clinical son of the Velten and musical procedures. AdMulti Axial Inventory profiles of sociotropic and autonomous subtypes. J o ~r n aofl Personality Disvances in Behaviour Research and Therapy, 3,27-49. Clark, D.A. & Teasdale, J.D.(1982) Diurnal variaorder, 3, 193-198. Golin, S. (1989) Schema congruence and deprestion in clincal depression and accessibility of memories of positive and negative experiences. sion: Loss of objectivity in self- and otherinferences. Journal of A b ~ o r ~Psychology, Journal of Abnormal Psychology,91, 87-95. al 98, Clark, D.A., Beck, A.T. & Brown, K. (1992) 495-498. Gotlib, I.H. (1981) Self-reinforcement and recall: Sociotropy, autonomy and life event perceptions in dysphoric and nondysphoric individuals. Differential deficits in depressed and Cognitive Therapy and Researc~,16,635-652. nondepressed psychiatric inpatients.~ournalof Clifford,P.I.& Hemsley, D.R.(1987) The influence A b n o ~ apsycho l lo^, go, 521-530. of depression on the processing of personal at-Gotlib, I.H. & Hammen, C.L. (1992) Psychological tributes. British Journal of Psychiatry, 150, 98-103. Aspects of Depression: Towards a CognitiveInterpersonaZ Integration. John Wiley & Sons, Cook, L.M. & Peterson, C.(1986) Depressive irrationality.Cognitive~ ~ e r aand Chichester. p y Research,lo, 293Gotlib, I.H., Mount, J.H.,Cordy, N.I. & Whiffen, 298. V.E. (1988) Depression and perceptions of early Cope, J.C. & Gotlib, I.H.(1983) The roleof cognition in depression: A critical appraisal. parenting: a longitudinal investigation. British Psychological Bulletin, 94, 472-505. Journal of Psychiatry,~ 5 2 ~ 2 4 - 2 7 . DeMonbreun, B.G.& Craighead, W.E. (1977) Dis- Greenberg, M.S.& Beck, A. (1989) Depression vs. anxiety: a test of the content-specificity hypotortion of perception and recallof positive and neutral feedback in depression. C o ~ ~ t iTherapy ve thesis. Journal of Abnormal Psychology,9 8 , 9 1 3 . Greenberg, M.T., Cicchetti, D. & Cumings, M. and Research, I,311-329.
C h ~ ~ t5: eA r ~ o ~ n i t itheory ve of ~ ~ r e s s i o n (eds) (1990) Attachment inPreschool Year: Theory, chology, 88,611-619. Research andlntervention. University of Chicago Lakey, B. & Thomson-Ross, L. (1994) Dependency Press, Chicago. and self-criticism as moderators of interpersonal Haaga, D.A., Dyck, M.J. & Ernst, D. (1991) Emand achievement stress: the role of initial pirical statusof cognitive theoryof depression. dysphoria. Cognitive Therapy and Research, 18, Psychological~ulletin, n o , 215-236. 581-599. Hamilton, E.W. & Abramson, L.Y. (1983) Cogni- Lam, D.H. & Power, M.J. (1991) A questionnaire tive patterns and major depressive disorder: a designed to assess roles and goals: a preliminary longitudinal studyin a hospital setting.J o u ~ a ~ study. ~ r i ~ i~s ho u ~ofaMedical l P s ~ c h o l o64, ~, of A b n o ~ aPsychology, l 92,173-184. 359-374. Hammen, C., Marks, T., Mayol, A. & de Mayo, R. Lam,D.H.,Brewin,C.R., Woods,R.T. & (1985) Depressive self-schemas, life stress, and Beddington, P.E. (1987) Cognition and social vulnerability to depression. Journal of A b n o r ~ a l adversity in the depressed elderly. Journal of Psychology,94,308-319. A b n o ~ aPsychology, l 96,23-26. Hammen, C., Elliott, A., Gitlin, M. & Jamieson,K.R. Lam, D.H., Green, B., Power, M. & Checkley, S. (1989) Sociotropy/autonomy and vulnerability (1996) Dependency, matching adversity and to specific life events in patients with unipolar relapse in major depression.Journul of Afective depression and bipolar disorder.Journal of AbDisorders, 3 7 , 8 1 9 0 . normal psycho lo^, 98,154-160. Lefebvre, M.F. (1981) Cognitive distortion and Hamrnen, C.,Elliott, A.& Gitlin, M.(1992)Stressor, cognitive errors in depressed psychiatric and sociotropy and autonomy: a longitudinal study low back pain patients. Journal of Consulting and Clinical Psychology,49,517-525. of their relationship to the course of bipolar disorder. Cognitive ~~a~ and Research,16,409-418. Linville, P. (1985) Self-complexity and affective Hinde, R.A. (1988) Continuities and disconextremity: don’t put of allyour eggs in one cogtinuities. Conceptual issues and methodologinitive basket.Social Cognition, 3,94-102. cal considerations.In:Studies of Psychosocial Risk: Lloyd, G.G.& L i s h a n , W.A. (1975) Effects ofdeThe Power of ~ o n ~ t u d i nData a l (ed. M. Rutter), pression on the speed of recall of pleasant and pp. 367-383. Cambridge University Press, Cam- unpleasant experiences.Psychologicul Medicine, bridge. 5, 173-180. Hollon, S.D.& Kendall, P.C. (1980) Cognitive self- Main, M., Kaplan, N. & Cassidy, J.(1985) Security statements in depression: development of an in infancy, childhood and adulthood: a move to the levelof representation. In:Growing Points in Automatic Thoughts Questionnaire. Cogniti~e (eds I. Bretherton Attachment Theory and Research Therapy and Research, 4,383-395. Hollon, S.D., Kendall, P.C. & Lumry, A. (1986) & E. Waters), Vol. 50, Society for Research in Specificity of depressiotypic cognitions in clini- Child Development Monographs, pp. 60-104. Chicago University Press, Chicago. cal depression. Journal of Abnormal psycho lo^, McCauley, E., Mitchell, J.R., Burke, P. & Moss, S. 95, 52-59. (1988)Cognitive attributes of depression in chilJohnson, J.E., Petzel, T.P., Dupont, M.P. & Roarnno, dren and adolescents.Journal of Consul~ingand V.M. (1982) Phenomenological perceptions of Clinical Psychology,56,903-908. parentalevaluations indepressedand nondepressed college students.Journal of Clini- McClelland, D C , Atkinson, J.W., Clark, R.H. & Lowell, E.L. (1953) The Achievement Motives. cal Psychology,3 8 , 5 6 6 2 . Appleton-Century-Croft,New York. Kazdin, A.E., French, N.H., Unis, AS., EsveldtDawson, K. & Sherick, R.B.(1983)Hopelessness, McCraine, C.W. & Bass., J.D. (1984) Childhood depression and suicidal intent among psychiat- family antecedents of dependency and selfcriticism: implications for depression. Journal rically disturbed inpatient children. Journal of of A~normalPsychology,93,3-8. Consulting and Clinical Psychology,51,504-514. Kelly, J.A. & Worrell, L. (1976) Parent behaviours Miranda, J. & Persons, J.D. (1988) Dysfunctional related to masculine, feminine and androgenous attitudes aremood dependent.~ournalof Abnormal Psychology,9 7 , 7 6 7 9 . sex role orientations. Journal of Consulting and Mongrain, M. & Zuroff, D.C. (1989) Cognitive Clinical psycho lo^, 44,843451. vulnerability to depressed affect in dependent Krantz, S.E. (1985) When depressive cognitions and self-critical college women.~ournalof Perreflect negative reality. Cognitive Therapy and sonality Disorders,3,240-251. ~esearch,g, 6 1 7 7 . Krantz, S. & Hammen, C. (1979) Assessing cogni- Myers, J.F., Lynch, P.B. & Bakal, D.A. (1989) Dysthymic and hypomanic self-referent effects tive biasin depression.Journal of Abnormal Psy122
~ h a p t 5:e ~A co~nitivetheory o ~ ~ e p ~ e s s i o n associated with depressive illness and recovery.Scales. Cognitive Therapy and Research, 17, 531Cognitive Therapy and Research, 13,195-209. 549. Parker, G., Tupling, H. & Brown, L.B. (1979) A Rude, SS., Krantz, S.E. & Rosenhan, D.L. (1988) parental bonding instrument.British Journal of Distinguishing the dimensions of valence and Medical Psycholo#, 52, S-so. belief consistency in depressive and non-depresParry, G.& Brewin, C.R. (1988) Cognitive style and sive information processing. Cognitive Therapy depression: symptom-related, event-related or and Research, 12,391-407. independent provoking factor? British Journal Rush, A.J., Weissenburger, J. & Evans, G. (1986) of Clinical Psychology,27,23-35. Do thinking patterns predict depressive sympPerson, J. (1994) Case conceptualisation in cogni- toms? Cognitive T ~ a ~ aResearch, n d lo, 225-236. tive behavioural therapy. In:Cognitive Therapy Rutter, M. (1995) Clinical implications of attachin Action (eds K. T. Kuehlwein & H. Rosen). ment concepts: retrospective and prospective. Jossey-Bass Publisher, San Francisco. Journal of Child Psychology and Psychiatry, 36, Power, M.J., Katz, R., McGuffin, P., Duggan, C.F., 549-571. Lam, D. & Beck, A.T. (1994) The Dysfunctional Sandler, I.N.& Lakey, B. (1982) Locus of control as Attitude Scale (DAS): a comparison of forms a a stress moderator: the role of control percepand b and proposal for a new sub-scaled vertions and social support. American Journal of sion. Journal of Research in Personali~y,28, 263CommunityPsychology,lo, 65-80. 276. Sarason, I.G., Johnson, J.H. & Siegel, J.M. (1978) Rajecki, D.W., Lamb, M.E.& Obmascher, P. (1978) Assessing the impact of life changes: developTowards a general theory of infantile attachment of the Life ExperiencesSurvey. Journal of ment: a comparative review of aspects of the Consultingand Clinical Psycholo#, 46,932946. social bond. Brain and Behavioural Sciences, I, Schwarz, J.C.& Zuroff, D. (1979) Family structure and depression in female college students: 4'7-436. Rippere, V. (1979) Scaling the helpfulness of effects of parental conflict, decision making antidepressive activities.Behaviour Research and power, and inconsistency of love. Journal of Abnormal Psychology, 88,398-406. Therapy, 17' 439-449. Rippere, V. (1981) Depression, common sense, andSefran, J.D., Vallis, T.M., Segal, Z.V.& Shaw, B.F. psychosocial evolution.~ r i t i s hJournal of Medi(1986) Assessment of core cognitive processes cal Psychology,54,379-387. in cognitive therapy. Cognitive Therapy and Robins, C.J. (1985) Construct validation of the Research, 10,509-526. Sociotropy-Autonomy Scale: A measure of vulner- Segal, Z.V., Shaw, B.F. & Vella, D.D. (1989) Life ability in depression. Paper presented at the stress and depression: a testof the congruency annual meeting of the Eastern Psychological hypothesis for life event content and depressive Association, Boston. subtype. CanadianJournal of Be~aviouralScience, Robins, C.J. (1990) Congruence of personality and21,389-400. life events in depression. Journal of A~normal Segal, Z.V., Shaw, B.F., Vella, D.D. & Katz, R. (1992) Cognitive and life stress predictors of relapse in P s y c h o l o ~99,393-397s f Robins, C.J.& Block, P.(1988) Personal vulnerabilremitted unipolar depressed patients: a test of ity, life events and depressive symptoms. Jourthe congruency hypothesis. Journal of Abnormal Psychology,xox,26-36. nal of Personality and Social Psychology, 54, 846862. Siegelman, M.& Roe, A. (1979) Manual: The Robins, C.J., Hayes, A.M., Block, P.,Kramer, R.J.& Parent-Child Relations Questionnaire: 11. City Villena, M. (1995) Interpersonal and achieveCollege of NewYork, New York. ment concerns and the depressive vulnerability Silverman, J.S., Silverman, J.A. & Eardley, D.A. and symptom specificity hypotheses: a prospec- (1984) Do maladaptive attitudes cause deprestive study. Cognitive Therapy and Research, 19, sion: misconception of cognitive theory. Archives 1-20. of General Psychiatry,41,1112. Simons, A.D., Garfield, S.L. Rosenfarb, I.E., Becker, J., Khan, A. & Mintz, J. & Murphy, G.E. (1984) (1994) Dependency, self-criticism, and percepThe process of change in cognitive therapy and tions of socialisation experiences. Journal of pharmacotherapy for depression. Archives of Abnormal Psycholo#, 103,669-675. General Psychiatry,41,45-51. Rude, S S . & Burnham, B.I. (1993) Do interpersonalSimons, A.D.,Murphy, G.E., Levine,J. & Wetzel, and achievement vulnerabilities interact with R.D. (1986) Cognitive therapy and pharmacongruent events to predict depression? Comcotherapy for depression. Archives of General parison of DEQ, SAS, DAS and Combined Psychiatry,43,43-50.
Simons, A.D., Angell, K.L., Monroe, & S.M. Thase, Thomas, C.B.& Duszynski, K.R. (1974) Closeness in a proM.E. (1993) Cognition and life stress in depres- to parents and the family constellation Johns sion: cognitive factors and the definitions, ratingspective study of five disease states. ~ o p ~ iMedical ns Journal, 134, 251-270. and generationof negative life events. JournaZ of Velten, E. (1968) A laboratory task for induction Abnor~aZpsycho lo^, 102,584-591. of mood states. gehaviour Research and Therapy, Steketee, G., Grayson, J. & Foa, E. (1985) Obsessive-compulsive disorder: differences between 6,473-482. nt washers and checkers. Behaviour Research and Weissman, A.& Beck, A.T. (1979) ~ e v e l o p ~ eand ~aliaationof the ~ y s ~ n c t i o n a l A t ~ i Scale: fude A Therapy,23,197-201. Paper presented at Teasdale, J. (1988) Cognitive vulnerability to per- Preli~inary ~nvestigafion. the annual meetingof the American Education sistent depression. Cognifion and ~ ~ o t i o2,n , Research Association. 247-274. Teasdale, J. & Bancroft, J. (1977) Manipulation of Welkowitz, J., Lish, J.D. & Bond, R.N. (1985) The thought contentas a determinant of mood and Depressive Experiences Questionnaire: revision and validation. JournaZ of PersonalityAssess~ent/ corrugator EMG in depressed patients.Journal of A b n o r ~ apsycho l lo^, 86,235-241. 49, 89-94. Teasdale,J. & Fogarty, S. (1979) Differential effects Wenzlaff, R.M.& Grozier, S.A. (1988) Depression and the magnification of failure. Journal of Abof induced mood on retrieval of pleasant and unn o r ~ apsycho l lo^, 97,90-93. pleasant events from episodic memory. Jou~aZ of Williams, J.M.G. (1984) The PsychologicalTr e a f~e n t A b n o r ~ apsycho l lo^, 88,248-257. of ~ ~ r e s s i oRoutledge, n. London and New York. Teasdale,J. & Rezin, V. (1978) The effects of reducing frequencyof negative thoughts on the moodWilliams, J.M.G., Healy J.D., Teasdale, J.D., W'hite, W. & Paykel, E.S. (1990) Dysfunctional attitudes of depressed patients: tests aof cognitive theory and vulnerability to persistent depression. of depression.Bri~ishJournal of Social and CfiniPsychological~ e d i c i n e20,375-381. / cal psycho lo^, 17, 65-74. Teasdale, J.D. & Russel, M.L. (1983) Differential Zung, W.W.K. (1965) A self-rating depression scale, , effects of induced mood on the recall of positive,Archives of General P s y c h i a f ~12,63-70. Zuroff, D.C., Igreja, I. & Mongrain, M. (1990) negative and neutral words. British Journal of Dysfunctional attitudes, dependency, and selfClinical psycho lo^, 22,163-172. criticism as predictors of depressive mood Teasdale, J., Taylor, R.& Fogarty, S. (1980) Effects of induced elation-depressionon the accessibil- states: a 12-month longitudinalstudy. Cognitive and Research, 14,315-326. ity of memories of happy and unhappy experi- Thera~y ences. g e ~ v i o uResearch r andT h ~ a p18,339-346. ~/
Psychoanalytic (Abraham, 1911; Freud, 19x7; Cohen et al., 1954; Mendelson, 1960) and interpersonal t~eorists(Coyne, 1976a; Brown & Harris, 1978; Klerman & Weissman, 1982; Lewinsoh, 1985) have pointed to the importance of interpersonal processes in promoting and maintaining depression. While the psychoanalytic theorists have described how early childhood experiences of inte~ersonalprocesses are important precursors of depression, particularly the quality of the child-parent (usually mother) relationship, interpersonal theorists have tended to focus on the functional role of depression. Interpersonal theories emphasize how problematic interpersonal interactions develop when a person becomes depressed. Psychoanalytic theories have described how parental rejection and lack of warmth and availability can lead to a~bivalentfeelings towards important figures which,if reactivated in adulthood, resultin regression toearlier beliefs of helplessness and hopelessness. Self-esteem and mood are then lowered, feelings of inadequacy and anger are turned inwards, and a depressive episode maybe experienced. Interpersonal theorists,in contrast, focus on the current relationship difficulties of the depressed individual. Such difficulties lead to a reduction in social reinforcement (Libet& Lewinsoh, 1973; Weissman & Paykel, 1974) and the pattern of interacting between the depressed person and others becomes confused and unsupportive (Coyne, 1976a). espite the abundance of theoretical accounts of depression, research studies of the social and interpersonal context of depression have not been guided by a cohesive theory(Gotlib, ~ 9 9 2and ) ~ attemptsat synthesis of the research in this area (for example, Segrin & Dillard, 1992) have led to complex and often contradictory conclusions (Burchill & Stiles, 1995). In addition, many writers have not considered depression as a separate illness, but as one element of the array of disturbances a person may display when faced with interpersonal and social difficulties. Struppet al. (1982), for example, wrote, 'the objective of developing relatively specific treatments for relatively specific clinical conditions has been basically alien to psychoanalytic psychotherapy' (p. 217) Althougha psychoanalytic clinician observes and describes the same symptoms, mood and behaviours as clinicians froma cognitive or
biological background, theyunderstand them in a differentway. Part of the problem of producing an inclusive intra- and interpersonal theory of depression is that it requires multiple meanin~sto be attached to the same phenomena; as Coyne(1985)noted, there is noneutral language to describe depression, viewpoints are not distinct from their data. In this chapter attachment theory (Bowlby, 1969,1973, 1980) will be used to move beyond these conflicts in the literature. Attachment theory incorporates both in~a-individualor personality mechanisms involvedin the aetiology of depression and interpersonal factors that lead tothe maintenance and relapse of depression. Although attachment is conceived of as afundamental human motivation (Baumeister &: Leary, 1995), specific early attachment experiences are thought to predispose the individual to depression. These attachment experiences lead to dysfunctional working models about the self, others and relationships, which leave the individual vulnerable to depression when faced with life events that activate the attachment system. Early dystwo basic atfunctional attachment experiences, it is proposed, can lead to t a c ~ e nstyles t that, in turn, can lead to two types of depression based on threats to interpersonal relations and threats to self-esteem (Blatt&: Zuroff, 1992). Thus, attachment theory provides a conceptual model within which much of the psychodynamic and interpersonal work in this area can be organized.
Attachment theory originated from Bowlby’s seminal workwith infants on the impact of separation from their care-giver(Bowlby, 1969, 1973, 1980).In thee volumes entitled~ ~ ~ a cS e~~ a~~ ae~and n i o nLoss, ~ , Bowlby documented the emotional bond that infants develop with their primary care-giver and the serious consequences that follow if this bond is disrupted. Predictable stages of emotional responses were recorded following separation. These wer first, p ~ o ~ ewhen s ~ , infants cry, search for their care-giversand resist others’ attempts to soothe; second,despair, when infants are visibly sad and passive; and third, d e ~ a c ~ ~characterized en~, by a seeming defensive disregard thefor care-givers. From this work Bowlby argued that there was primary a motivational system to seek contact with and security from a care-giver. This is known as the attachment systemand is evidenced most clearly when situations are seen as dangerous or threatening, Considerable evidence of these elementsof attachment behaviour in children has been obtained from naturalistic and laboratory studies (Rutter, 1979; Maccoby, 1980; Bretherton, 1985; Erickson et al., 1985).
C ~ a ~ t6:e A r ~syc~odynamic-inte~ersonal model Bowlby listed the primary features of the attachment system as a bond with specific individuals who are seenas stronger and wiser, which is enduring and serves the function of protection and survival. The attachment bond has a strong emotional componentthat can be positive (such as when being renewed) or negative(when threatened orlost). In addition, because attachment behaviour driven is by a primary motivation, attachments develop irrespective of the abilityof care-givers to respondappropriately or consistently. and ex~loration
Inevolutionary terms, the attachment system is seen as providing infants with a means of obtaining protection fromdanger by keeping them close to their care-giver. If the primary care-givers are responsive and sensitive toa child's attachment signals, thenthat child will form a 'secure base' from which to manage emotional experience and handle distress (Bowlby, 1988). Such children will then be able to use their care-giver a base as from whichto explore their environment. Providinga secure base involves care-givers being both emotionally responsiveand providing physical security. The attachment system, or care-seeking behaviour, is complemented by what Bowlby called 'exploratory behaviour'. If children experience their caregivers as available and responsive, then they will increasingly move further and further away from their care-givers in order to explore and master their environment and establish friendships with a widening net of social contacts (Main, 1983). ~ttachment~ e h a ~ i oin u radulthood
Bowlby hypothesized that infants and children internalize their attachment experiences and form internal working modelsof themselves and their relationships with significant others. AccordingBowlby to these working models are central components of personality: ...confidence that an attachment figure is likely to be responsive depends on (a) whether the attachment figure judged is to be the sort of person whoin general responds to calls forsupport and protection; and (b)whether or not theself is judged to be the sort of person towards whom anyonein general, and the attachment figure in particular, is likely to respond in a helpful way.. as a result, the model of the attachment figureand the modelof the self are likely to develop so as to be complementary and mutually confirming. (Bowlby, 1973,p. 238) These internal working models are thought to beenduring structures and so they provide people throughout their lives with the experience of 'felt
security’ (or insecurity) (Blunstein et al., 1991). Social, interpersonal and exploratory behaviour willbe influenced by the experienceof security (provided in infancy almost totallyby the responsivenessof care-givers) and, as cognitive structures become more sophisticated, increasingly by felt security (which is a productof a person’s internal working models). The attachment pattern established in infancy was also by seen Bowlby as continuing throughout adulthood (Bowlby, 1979). This fundamental need to form and maintain interpersonal bonds, pervading adult life but beginning in infancy with one’s primary care-givers, also forms the basis of psychoanalytic thinking (Freud, 1930; Horney, 1945; Guisinger & Blatt, 1994). In a review of the literature, Baumeister and Leary(1995, p. 497) concluded that s~fficientevidence exists to support the hypothesis that the ’needto belong is a powerful, fundamental, and extremely pervasive motivation’. Hazan and Shaver (1987) argued that there was a strong case for conceptualizing adult romantic relationships as attachment processes. These separate strands of theoretical and empirical support began what is now a major interest in applying attachment theory to describe, explain and predict adult social behaviour within both normal and psychiatric populations ( Bartholomew &E Perlman, 1994).
According to attachment theory, psychopathology therefore results from deviant patternsof attachment behaviour established earlylife, in which persist through the influence of internal working models and affect interpersonal relationships in adult life. Bowlby’s original work on loss experiences has been used to clarify the processesthat underlie the developmentof depression. Threat of separation from or loss of an attachment figure results, initially, in anger and protest. It is thought that the function of anger is to achieve reunion with the lost figure (Parkes, 1972). If protests go unheeded, then anger gives way to despair and then finally to detachment. These processes are part of normal grieving, but if experienced early in life, or repeatedly without resolution, then despair and detachment become the customary response when a person (childadult) or is faced with threaten~g a situation. Such responses make the person vulnerable to a sense of helplessness and to depression (Bowlby, 1973). Seligman (1975) showed how this phenomenon of ’learned helplessness’is active in producing depressed states. Attachment theory therefore links depression to: (i) early experiences of care-giving; (ii) expectations and interpretations of others’ behaviour; and (iii) current relationship function~g(Carnelley et al., 1994).
~ h a p t 6: e ~A psychodynamic-interpersonal m o ~ e l iVi
Early attachment experiences c m provide childrenwith models of themselves as loveable and competent, and with models of others as supportive, warm and reliable. Such models are likely to lead children to develop a ofstyle what Ainsworth et al. (1978)described as ’secure attachment’. Children who experience theirimportant care-givers as cold, unavailable or inconsistent, in contrast, are likely to developpatterns of ‘insecure attachments’through their poor and inflexible working models of themselves in relationships with others. Psychod~amicunderstandings of depression also point to the central importance of early experiences. Freud (1917) originally saw depression as the result of the loss of either a loved one or a ’narcissistic need’ or an unrelin~uishedwish to be loved, superior or good (Rapaport,1959).A. basic ambivalence towards the lost person (i.e. theindividual feels both loveand hate towards the lost person), means grief is not worked through, but the individual maintains angertowards the lost ’object’(person or need), which inward towards the self. Sustainedself anger isthen a causeof guilt and depression.
Cognitive theories (seeChapter 5) focus on the depressed person’s viewof him/herself. Studies based on these theories have shown that depressed people, in comparison to non-depressed people, have a negative view of themselves, the world and the future (Beck, 1967). Some research into dele’s viewsof others, however, has produced conflicting findings. es havefound that depressed individuals have a negative view of others (~ietromonacoet al., 1 9 9 ~while ) ~ others have found people are more positive about others than about themselves 1988). These findin S are perhaps tapping two underlying dimensions of depression that wil be discussed later, and are based on the early experiences of attachment figures being either: (i) rejecting, resulting in people feeling negative about themselves, but positive about their attachment ure; or (ii) inconsistent, resulting in people feeling negative about themselves a~~ their attachment figures.
The role of intimate relationships in depression is seenas central by many writers from Cohenet al. (1954.)~through the experientialists (Rogers, 1959; Perls et al., 1951) to Coyne (197613). There is indeed convincing evidence to
~ h a p t e6 r:A p s~ c h o d y n ~ ~ i c - i n t e ~ e r ~ s o no a ~ l e l show that relationships involving depressed people are often distressed; for example, Gotlib and Hooley(1988) concluded that the marital relationships of depressed people tend to be conflictual and unhappy and are characterized by poor c o ~ u n i c a t i o nWeissman . and Paykel (1974) found that depressed women were impaired in their social f u n c t i o ~ gexpressed , hostility and dissatisfaction with their spousesor partners, and that the couples had poor levels of communication. Attachment theory would, of course, predict that the relationshipsof depressed individuals reflect theirpast attachment experiences, and arelikely to be insecure and problematic. Coyne’s (1976b) model of depression is also based on thec o ~ u ~ c a t i o n patterns between a depressed person and her/his close relationships.Goyne conceptualized depressionas a response to a disruption in a person’s social structure, which eventually comes tobe supported and validatedby others (Burchill & Stiles, 1995). When first depressed, a person obtains the sympathy and supportof their close ones.If the depression continues, then the depressive symptomatology no longer produces concernin others but frustration and a burdensome sense of undischarged responsibility for the depressed person. The spouse or close friend then feelsguilty as their concern comes to be tempered by anger, helplessness and a wish to avoid too much ortoo close a contact with the depressed person. This,in turn, reinforces the depressed person’s viewof themselves as unloveable andso a vicious circleis set up. This research has clear implications for choiceof treatments in the management of depression;for example, the Interpersonal Treatment (Klerman & Weissman, 1982) used in the National institute of Mental Healthstudy of depression (Elkin et al., 1989) and theP s ~ c h o d ~ a ~ i c - ~ t e ~ e rtreatment sonal (Hobson, 1985; Shapiro &: Firth, 1985) used in the Second Sheffield Psychothat depression therapy study (Shapiroet al., 1994) are based on the premise occurs in a social and interpersonal context that needs to be understood for improvement tooccur.
There is now substantial evidence to suggest poor that childhood attachment experiences lead to a vulnerability to depression in adult~ood(Basic Behavioural Task Force, 1996).This vulnerability canbe triggered by a numberof , Thesehave been grouped life experiences (see Herman & ~ e i s s m a n1982). into two main types: disruptionsof interpersonal relations and experiences of failure, guilt or loss of control (see Table 6.1). These, in turn, lead to two forms of depressive reactions, dependent and self-critical (Blatt & Homann, 1992). The dependent depressed individual is someone who relies heavily on others, is fearful of being alone and whose depression is characterized by feelings of loss, abandonment andinability to face the future using their own
~ h a p t 6: e~ A psychod;ynamic-interpersonalmod;el Table 6.1 Schematic summaryof literature on dependent and self-critical depression
styles. (FromBlatt, S.J. (1991) Depression and self destructive behaviour in adolescence. In:S e ~ ~ e~ g ~ e~anda~ ~~~ s ~k -~~Causes a ~ ~ and ~~nCg~ n ; s ~e ~ ~ e n(eds c~e s L.P. Lipsett ~ 8r ~ L. L. Mitnick)). Ablex, Norwood, NJ. Copyright 1991 by Ablex Publishing Corp. Reprinted by permission. Proximal Clinical precipitating events (life stress) disorders
Distal precipitating events (parental styles)
Personality organization (dysphoria)
1 Inconsistent, neglectful, rejecting, abandoning and/or overindulgent parents. Anxious/ ambivalent attachment
~n~erp~sonaZ Disruptions of Vu~nerabiZi~y interpersonal Dependent,affectivelyrelations,experiences labile, preoccupied of loss, abandonment, withattachmentandandrejection interpersonal relations. Use of avoidant defences (denial and repression)
~ ~ e n ~depression e n t Feeling sad and lonely, rejected, unloved, abandoned. Helpless searching for an object who will care, soothe, feed and provide love. Somatic preoccupation, eating disorders, sexual acting out. Suicidepassive
2 Caring and dependable parents who are able to establish appropriate and constructive limits, set appropriate standards and goals, and provide approval and criticism in constructive ways. Secure attachment
~ e ~ a t i~nvulnerability ve Able to form mature, mutual reciprocal relationships. Able to enjoy one’s talent and capacities and to accept comfortably one’s shortcomings and limitations
Normal grief and sadness (mou~ing) without protracted, prolonged severe depression
3 Intrusive,
Se~-evalu~tive vulnerabili~ Striving self-critical, overcompensating; preoccupied with issues of failure/ success, blame and responsibility. Hostile, critical, angry. Use of counteractive defences (projection, reaction formation, overcompe~ation)
controlling, judgemental, punitive parents. Avoidant attachment
Experiences of failure, guilt diminished selfesteem, lossof control and an impaired sense of mastery. Feeling criticized, unappreciated, and ridiculed
Se~-cri~ical depression Feelings of worthlessness selfcriticism and guilt. Constant striving to compensate for perceived feelings of guilt and inferiority. Serious attempts to hurt self and others. Antisocial delinquency, violence (aggression). Suicide-violent
resources. The self-critical depressed person,in contrast, is fearful of disapproval by others, strives for perfection and suffers from feelings of unworthitypes of ness, failure and guilt. The attachment styles that result in two these depression have been called, respectively, anxious/ambivalent and avoidant
~ ~ a6: A~ psychodynamic-interpersonal t ~ r moael
(Ainsworth et al., 1978; although see ain et al., 1985 for an alternative terminology). These attachment styles or interpersonal strategies develop from and reflect the responsiveness of the person’s care-giver. If the care-giver responds negatively to an infant’s demands, then the infant will learn to avoid expression of internal states, but will have a sense that the relationship with the care-giver is predictable, at least in a cognitive or causal sense. However, such an individual will not learn to use and under st an^ her/his emotions; an avoidant pattern of engagement with others is developed. If, on the other hand, a care-giver is inconsistent, the infant will be notable to organizeher/ his behaviour nor developa sense of mastery within interpersonal relationships, and will have intense feelings of insecurity for long periods. Such experiences lead to ambivalent interpersonal attachments,in which there is intense emotional expression, but little ability to observe and make use of causal relationships between events. links The between these two attachment n Table 6.1 (taken styles and dependent and self-critical depression are ishown from Blatt, 1991) and are discussed below. This summary of two types of depression represents the viewsof many interpersonal, psychod~amicand cognitive writers; for example, Arieti and Bemporad ( ~ 9 ~state $ ) that when depressed the person is trying either:(i) to (re)gain the love and attention of a ’dominant other’ on whom the individual feels dependent for support, self-esteem and well-being (anxious/ambivalent); or (ii) to be reassured of her/his worth and be free of guilt following failure to reacha self-imposed goalof impossibly high standards (avoidant). Arieti and Bemporad called these two forms of depression ’dominantother’ and ’dominant goal’, respectively. Blatt and Zuroff (1992) describe two forms of depression that result when there are di§~ptions to interpersonal relationships (anaclitic depression) or threats to self-int~grityand self-esteem (introjective depression). Finally, Beck (1983) disting~ishedbetween ’sociotropic depression’, where there aisv u ~ e ~ a b i l ito t yloss or disruptions in interpersonal relationships (anxious/ambivalent), and ’autonomous depression’, which is characterized by fear of failure and criticism and lossof control (avoidant). Although these writers have many different components to their theories, the basic distinction between the two forms of depression remains similar.
Early experiences of individuals with anxious/ambivalent attachments are characterized primarily by inconsistent parenting. Sometimes the child’s needs are attended to and sometimes not; sometimes quickly, sometimes slowly; sometimes overindulgently, but often not. The care-giver is frequently intrusive, but again canbe ~ p r e d i c t a ~absent ly or abandoning. Children of such
1 in.teracti0n.s with others an.
~ ~ a p6:tAepsychodynamic-intevpersonal ~ ~ o d ~ l are preoccupiedwith controlling events."hese individuals are often very selfcritical, preoccupiedwith success and feel guiltyand responsible for any failures they encounter. They tend to avoid close interactions with other people and invest in achievementsrather than relationships. Emotional expression is limited, but underneath the desire for success the is belief that this will bring approval and a senseof being valued by others. If such avoidant people experience failure or loss of control, then their fragile self-esteem is si~ificantlyimpaired and dysphoric mood results. The content of their depressivethoughts is of self-deprecationand guilt. They feel that they have not lived up to their own self-imposed standards, and they will inevitably lose the respect and love of those around them. Blattand Zuroff (1992) have called this form of depression self-critical or introjective.
The conceptsof self-esteem and affect regulation have often been referred to in the preceding descriptionsof attachment theoryand types of depression. Indeed, in both psychodynamic and interpersonal theories, loss of selfesteem is one of the central symptoms of a depressive episode and one of the vulnerability factors for depression.Affect, such as anger, disappointment, guilt and sadness, is also described in psychodynamic interpretations of depression. Experientialists, although they would reject many of the psychodynamic understan~ingsof such phenomena, see psychological distress, including depression, as a result of the meanings people attach to their mood or affect (Greenberget al., 1993). Self-esteem refers to our views of whether other people value, respect and like us. To have low self-esteem is to saythat we feel that others have little regard for us, that they do not like us or think we have little to offer to them or to society in general. The conceptinterpersonal is in that it reflects individuals' views of themselves in relation to others. This concept is, of course, part of what Bowlby called the 'internal working model'. It develops based our on experiences of how others, especially care-givers, have responded to us. In securely attached children, care-givers have generally responded positively, consistently and responsively to demands and needs, giving children the sense that they arevalued and loved and confident that their requests forsupport or help will be met. Such children develop and positive robust self-esteem. In contrast, childrenwho are unsure whether care-givers will respond helpfully and with affection, or who know that any displays of upset or need will be rejected, aref~damentally less sure about their place or survival in the world, As they grow, their self-esteem remains fragile. Although they may develop strategies for avoiding the pain of not feelingvalued, if these are challenged too strenuously then the early feelings of rejection and disappoin~entare
likely to resurface resultingin low self-esteem.~xious/ambivalentpeople and avoidant people, respectively, depend on closeness with others and success for self-esteem. There aresig~ficantdifferences in the emotional functioning of anxious/ ambivalent and avoidant individuals. Anxious/ambivalent people tend to show a great deal of emotional lability. They are anxious about being ignored or abandoned, need reassurance and demand emotional investment from other people. It is this that often leadsothers to feel annoyed and trapped, resulting in the depressed patternof interacting described by Coyne (197613). The emotions expressed by the ~ o u s / ~ b i v a l e n tdepressed ly person are what the experientialists call the automatic ‘activation of dysfunctional emotion schemas’ ( ~ r e e n ~ eetr gal., 1993). By this they mean that certain situations or actions automatically activate particular emotional responses without the individual consciously appraising the situation, for example, perceived loss or rejection will inevitably resultin feelings of anger and protest, whatever the context. Avoidant individuals avoid emotional expression both in themselves and from others. They do not utilize the emotional information available to them, and exhibit what the experientialists call ’dysfunction in symbolizing’. When faced with a potentially upsetting or difficult situation, avoidant people will ignore their physiological experiences of emotion (such as increased heart rate associated with anger). They have failed to develop an effective language for expressing how they are feeling and consequently, when depressed, experience emptiness and flatness, rather than grief or anger.
lie There now exists compelling evidence for many aspects of attachment theory’s hypotheses concerning the development of depression. Depression not only comprises specific symptomatology, but also affects of quality life pervasively, especially in the area of interpersonal relationships.It is a recurrent disorder, and the emotional and relationship deficits that occur within a depressive episode are likely to contribute to future relapse (Brown et al., 1977; Barnett&r Gotlib, 1988; Hooley& Teasdale, 1989; Marcus&r Nardone, 1992). Moreover, depression runs in families; children of depressed mothers are twice likely as to suffer from depression than children of non-depressed mothers( H a m e n , 1991). It can be concluded safely that vulnerability to depression begins in childhood andis dependent upon the contextor environment within which the at-risk person finds themselves. Whilst biological and cognitive factors may contribute to the aetiology and maintenance of the condition, interpersonal factors constitute a common theme across all models, whether this be at the stage of early life experiences resulting in a vulnerability for depression,
or the later family dynamics of a depressed person maintaining a depressive cycle. For the clinician interpersonal issues are therefore important to assess at three stages: vulnerability factors, mechanismsof change in treatment, and maintaining t r e a ~ e ngains. t Thesewill be considered in turn.
Using attachmentre~resentationsand style as a focus, the depressed person's vulnerability canbe assessed in two ways: their self concept and their interpersonal history. Poor relationships with care-givers in childhood may lead to insecure attachments. This, in turn, may leave the individual excessively dependent on others and fearful of rejection or abandonment (anxious/ ambivalent), or rejecting of close relationships and fearful of being controlled or hurt. Bothof these experiences lead to a fragile self-concept. The anx~o~s/ambivalent person when depressedis likely to have a lowered senseof self-worth and feel that they have little to contribute to the worl for example, they may feel that they'have noth of value to contribute to a relatio~ship, group or even societyat large' (Gi t, 1992).Following theloss of a confiding relationship (associated with onset of a depressive episode, Harris, 1978),anxious/ambivalent people often feel that they are not sufficiently interesting be toworthy of the attentionor concern of anyone, including thatof the person they have just lost. The anxious/ambivalent person needs closeness to others to maintain any sense of self-esteem. The avoidant person has feelings of equacy. To avoid these unrea~izable) feelin S, such a person invests in some demanding (and often epression follows failure to reach thisgoal, with the linked fear that other people will no longer show respect or liking. The avoidant person, when faced with failure, may also feelofloss belief in self-control or ability to change things andbegin to feel hopeless about the future. The interpersonal history of both anxious/ambivalent and avoidant individuals is likely to include a pattern of dysfunctional relationships. The mother-child interactions of depressed mothers, when compared to interactions with non-depressed mothers, areless positive, more criticalm d show 1990).A, depressed woman often chooses a depressed less involvement (Rutter, partner and the marriages of depressed people,on the whole, are more stressful, more conflictual and are more likely to break down (Gotlib & Hooley, ).A,ssessment of current attachment style and attac ent representations of care-givers can give the clinician a good picture of interpersonal functioning and of more generallife functionin (i.e. exploratory behaviour).
~ h a p t 6: e ~A psychodynamic-interpersonal ~
o
The aims of an interpersonal-psychodynamic treatment areu n tod e r s t ~ d and modify the problems that have arisen a result as of disturbances of significant personal relationships. The primary mechanisms through which change is assumed to occur are:(i) through the relationship and dialogue between the therapist and patient; and (ii) through the expression of feelings associated with difficult experiences from past, current and therapeutic relationships. ese mechanism are described in a generic modelof psychotherapy (Garfield ergin, 1986) in which the therapeutic contract, the~ e r a p e u t i c i n t e ~ e n tion (which in interpersonal therapies involves the expression of feelin relationship difficulties) and the therapeutic bond (pa~ent-therapistre ship) contribute to therapeutic self-rea~zations(such as insight, resolution of conflicts) and the patient’s self-relatedness (their personality, or their attachment internal working model). These, in turn, influence the outcome of psychotherapy. In treatment the inter persona^ issues facing the patient will be reflected in the relationship that develops between the patient and the therapist. This provides the therapist with material for exploring, understand in^ and modifying relationship difficulties. The methods for achieving this will not be described in this chapter, but involve the therapist developing conversational strategies that enable the patient to disclose and re-experience problematic feelings in relationships, develop a language for expressing these feelings, understand the nature, origins and consequences of the feelings and seek new solutions to the interpersonal problems (Rounsaville & Chevron, obson, 1985; Greenberget al., 1993). S
It has been argued throughout this chapter that depression a develo~menis tal disorder in the sensethat vulnerability is established early in childhood, patterns of relationship functioning andbeliefs about theself are developed from these early experiences through working models, and these stabilshow ity over time (McCrae & Costa, 1988; van Ijzendoorn, ~995).The close relationships of depressed individuals are more stressful and less secure than those of non-depressed individuals, leading to greater ~ulnerability to subsequent depressive episodes (Merikangas et al., 1985). However, if depressed people have supportive marriages, they are more likely to maintain treatment gains than depressed people with conflictual marriages, and single people (i.e. those without attachment figures) are the depressed group least likely to et al., 1996). maintain treatment gains (Hardy This points to the importance of considering marital and family relation-
137
d
C~lzp~er 6: A psyc~odynnmic-interpersonlzl model ships and support networks of the depressed person even when in remission (see Chapter 4 on social theoriesof depression). Equally, the interactive style of the depressed person will suggest what sort of life stressors will increase their likelihood for another depressive episode. Understanding the attachment process and increasing the flexibility of the attachment responsein the patient will reduce the likelihood of relapse. In addition, the clinician should be a responsive attachment figure for the patient both during treatment and through the follow-up period.
It is clear that interpersonal contexts, both historical and current, influence the nature and course of depression. ~ttachmenttheory provides a powerful framework for organizing current knowledge of these influences. Early caregiving experiences help determine the way people see themselves and of the their stylesof interacting with others. These ’internal working models’ self and others make individuals vulnerable to stressful events that threaten their interpersonal relationships or their self-esteem. It has been suggested that two typesof depression result from these different life events: dependent and self-critical depression. In addition, the depressed person’s interactions with others help to maintain depressive symptomatolo~,and increase the likelihood of future depressive episodes. by Bowlby,but expanded This conceptual framework, initially developed and further articulatedby many researcherssince, links togetherboth traditional psychodynamic and social psychological models of depression. It represents an overarching and powerfully practical approach ~tod e r s t ~ d i n g some of the causes and consequences of depression. The value of this approach for preventing and treating depression is great, although it is only recently that researchers and clinicians have begun to recognize how valuable it is.
Abraham, K.(1911)Notes on the psycho-analytic Arieti, S. & Bemporad, J.R. (1980) Thepsychological organization of depression. A ~ e ~ i c a n investigationand treatment of manic-depressive 137,1356-1365. J o u ~ nofa ~Psychiut~, insanity and allied conditions. Reprinted in: Context-dependent Essential Papers on ~ ~ ~ e s(ed. s ~J.C. o nCoyne) Bargh, J. & Tota, M.E. (1988) automatic processing in depression: accessi(1985). NewYork University Press, NewYork. bility of negative constructswith regard to self Ainsworth, M.D.S., Blehar,MC., Waters, E. & Wall, Journal of P e r s o n a ~and i ~ Social S.(1978) Patterns of A t t a c h ~ ~At :Psycholo~cal but not others. PsychOlo~,54,925-939. Study of the Strange Situation.Erlbaum, Hillsdale, Barnett, P.A. & Gotlib, I.H. (1988) Psychosocial NJ. functioning and depression: distinguishing Arieti, S. & Bemporad, J.R. (1978) Severe and Mild among antecedents, concomitants,and conse~ e pr e s s i o n :The T~erapeuticApproach. Basic quences. Psychological~ulletin,104~97-126. Books, New York.
C h ~ p6:~Ae psychodynnmic-interpersonal ~ ~ o ~ e l Bartholomew,K. & Perlman, D. (1994)attach men^ Processes in Adulthood:Advances in Personal Relat ~ o ~ s h i pVol. s , 5. Jessica Kingsley, London.
Burchill, S A L . & Stiles, W.B. (1995) ~ n t ~ e r s o n a l processes in depression: What do we know now?
Manuscript. Department of Psychology, Miami University, Oxford,Ohio. Basic BehavioralTask Force (1996) Basic behavioral science research for mental health: vulnerability Carnelley, K.B., Pietromonaco, P.R. & Jaffe, K. (1994) Depression, working models of others, and resilience. American Psychologist,51,22-28. and relationship f ~ c t i o n i n gJ.o ~ r n oaf~Personneed to Baumeister, R.F. & Leary, M.R. (1995) The ality and Social PsychoZo~/66,127-140. belong: desires for interpersonal attachments as a fundamental human motivation. Psychological Cohen, M.B., Baker, G., Cohen, R., FrommReichmann, F. & Weigert, E.V. (1954)An intenBullet~n,117,497-529. Beck, A.T. (1967)D~ression:Causes and T r e a t m ~ ~ . sive study of twelve cases of manic-depressive psychosis. Reprinted in: Essent~alPapers on University of Philadelphia Press,P~ladelphia,PA. Depression (ed. J.C. Coyne) (1985). New York Beck, A.T. (1983) Cognitive Therapy and the EmoUniversity Press, NewYork. tional Disorders. International Universities Coyne, J.C. (1976a)Depression and the response Press, New York. of others. Journal of A~normalPsychology, 85, Blatt, S.J. (1991)Depression and destructive risk186-193. taking behavior in adolescence. In: Self-regulatory Behavior and Risk-taking: Causes and Conse- Coyne, J.C. (197613) Towards an interactional de~, scription of depression. P s ~ c h i a t39,28-40. quences (eds L. P. Lipsett & L. L. Mitnick), pp. Coyne, J.C.(1985)Ambiguity and controversy:an 285-309. Ablex, Norwood, NJ. introduction. In: Essential Papers on Depression Blatt, S.J.& Homann, E. (1992)Parent-child inter(ed, J.C. Coyne), pp. 311-330. New York Uniaction in the etiology of dependent and selfversity Press, NewYork. 12, criticaldepression. Clinical Psychology Review, Crittendon, P.M. (1992) Treatment of anxious 47-91. Blatt, S.J. & Zuroff, D.C. (1992) Interpersonal attachment in infancy and early childhood. Develop~ental Psychopatholo~y, 4,575-602. relatedness and self definition: two prototypes for depression. ClinicaZ psycho lo^ Review, 12, Elkin, I., Shea, M.T., Watkins, J.T. et al. (1989) National Institute of Mental Health Treatment 527-562. of Depression CollaborativeResearch Program: Blunstein, D.L., Walbridge, M.M., Friedlander, general effectiveness of treatments. Archives of M.L. & Palladino, D.E. (1991) Contributions of General Ps y c h ia t ~, 46,971-982. psychological separation and parental attachment to the career development process. Jour- Erickson, M.F., Sroufe, L.A. & Engeland, B. (1985) nal of Counselingpsycho lo^/ 38,3950. The relationship between quality of attachment Bowlby, J. (1969) Attachment and Loss, Vol. I. and behavior problems in pre-school in a high Attachment. Basic Books,New York. s Societyfor Research risk sample. ~ o n o ~ a pofhthe in Child Develop~ent, 50,147-166. Bowlby,J. (1973)Attach~entand Loss, Vol. 2. Separation,Anxie#y and Anger. Basic Books,New York. Freud, S. (1917) Mourning and melancholia. Re~ a ~ on ~ ~ r e s s i(ed. o n J.C. Bowlby, J. (1977) The making and breaking of printed in:E s s e ~ tPapers affectional bonds. Part 1: Etiology and psychoCoyne) (1985).New York University Press, New pathology in the light of attachment theory. York. British ~ournalof P s yc ~ z i a t130,201-210. ~, Freud, S. (1930) Civilization and its is contents Bowlby, J. (1979) The ~ a k i n gand Breaking of (trans. J. Riviere). Hogarth Press, London. A ~ e c ~ ~ o Bonds. n a Z Tavistock Press,London. Garfield, S.L. & Bergin, A.E.(eds) (1986)~ a n d ~ o o k Bowlby, J. (1980)attach men^ and Loss, Vol. 3.Loss, of Psychotherapy and Behavior Change, 3rd edn. Separation andD~ression.Basic Books,New York. John Wiley & Sons, New York. Attach- Gilbert, P.(1992) Counsellingfor Depression. Sage Bowlby, J. (1988)A Secure Base:Parent~hild ment and Healthy Human Development. Basic Publications, London. Books, New York. Gotlib, I.H. (199~)Interpersonal and cognitive Bretherton, I. (1985) Attachment theory: retroaspects of depression. Current Directions in Psyc~ologicalScience, I, 149-154. spect and prospect. monograph^ of the Societyfor Research in Child Development/50~3-35. Depression and Gotlib, I.H.& Hooley, J.M. (1988) Brown, G.W. & Harris, T. (1978) Social Origins of marital distress: current status and future direcDepression. Free Press,New York. (ed. tions. In: Hand~ookof Personal ~ela~ionships Brown, G.W., Harris, T. & Copeland, J.R. (1977) S. Duck), pp. 543-570. John Wiley & Sons, New Depression and loss. British Jour~alof Psychiat~, York. 130.1-18. Greenberg, L.S., Rice, L.N. & Elliott, R. (1993)
Fflci~i~flting E~otional Change. Guilford Press, ality, 56,427-434. New York. Mendelson, M. (1960) PsychoanaZytic Concepts of Guisinger, S. & Blatt, S.J. (1994) Individuality and ~epress~on. Thomas, §pringfield, IL. relatedness: Evolution of a f~damental dialec- Merikangas, K.R., Pruscoff, BA., Kupfer, D.J. & tic. American Psychologist, 49,104-111. Frank, E. (1985) Marital adjustment in major deHammen, C. (1991) Generation of stress in the pression. ~ournflZ of A~ectiveDisorders, g, 5-11. course of unipolar depression. ~ournalof Abnor- Parkes, C.M. (1972) Studies of Grief in Adult L@ mal Psychology,100,555-561. International Universities Press, New York. H a m e n , C., Marks,T., Mayol, A. & deMayo, R. Perls, FS., Hefferline, R.F. & Goodman, P. (1951) (1985) Depressive self-schemas, life stress, and Gestalt Therapy: Excitement and Growth in the vu~erabilityto depression.~ournfll of Ab~ormfll ~ u ~ ~ersonali~y. f l n Julian Press, NewYork. Pietromonaco, P.R., Rook, K.S. & Lewis, M.A. &Stiles, W.B. (1996) The (1992) Accuracy of perceptions of interpersonal impact of ~ f l r i ~status f l l on the outcome of treatinteractions: effects of dysphoria, friendship, ment for depression. Manuscript. Psychological and similarity. ~ o u ~ofa Personfllity l and Social Therapies Research Centre, University of PSYChOlO~,63,247-259. Leeds. Rapaport, D. (1959) Edward Bibring’s theory of Hazan, C.& Shaver, P. (1987) Romantic love condepression. Reprinted in: Essential Papers on ceptualised as an attachment process.~ournalof D ~r e s s io n(ed. J. C. Cope) (1985). New York University Press, New York. Personfllityand Social psycho lo^, 52,511-534. of personalHobson, R.F. (1985) Forms of Feeling. Tavistock Rogers, C.R. (1959) A theorytherapy, ity, and interpersonal relationships, as develPress, London. oped in the client-centered framework. In: Hooky, J.M. & Teasdale, J.T. (1989) Predictors of relapse in unipolar depressives: expressed emo- Psycho1o~:A Study ofScience,Vol. 3,184-256 (ed. tion, marital distress, and perceived criticism. S. Koch). Basic Books, New York. ~ournalof Abnorma~psycho lo^, g8,229-235. Rounsaville, B.J. & Chevron, E. (1982) InterHorney, K. (1945)Our lnner Con~icts:A Construcpersonal psychotherapy: clinical applications. tive Theory of ~ e ~ r o sNorton, ~ s . New York. In: Short-termPsychotherapiesfor Depression (ed. Klerman, G.L.& Weissman, M.M. (1982) Interper- J. Rush), pp. 107-142. John Wiley & Sons, New sonal psychotherapy: Theory and research. In: York. Rutter, M. (1979) Maternal deprivation, 1972-1978: Short-~ermPsychotkerflpiesfor ~epression(ed. J. New findings, new concepts, new approaches. Rush), pp. 89-106. John Wiley & Sons, New York. Lewinsohn, J. (1985) A behavioral approach to Child develop men^, 50,283-305. depression. Reprinted in: EssentialPapers on DeRutter, M. (1990) Commentary: some focus and process considerations regarding effects pamof York ~ression(ed. J.C.Cope), pp. 150-180. New tal depresion on c h h li. D ~ e l ~ t Psychof l l logy, ~ 6 ~ 6 0 - 6 7 . Segrin, C. & Dillard, JP. (1992) The interactional depressed persons. ~ o u r ~ofa lConsulting and theory of depression: a meta-analysis of the Clinical psycho lo^, 40,304-312, research literature.~ournalof Social and Clinical Maccoby, E.E. (1980) Social Development:Psycho1o~PSYChOlO~,11,43-70. cal Growth and the ~flrent-child ~ e l a t i o n s ~ i p . Seligman,M.E.P. (1975)~ e ~ p l e s s ~On e s sDepress~on, : Doubleday, New York. Develop~entand Death.W.H. Freeman, San FranMain, M. (1983) Exploration,play, and cognitive cisco, CA. Shapiro, D.A. & Firth, J.A. (1985) Explorfltory functioning related to infant-mother attacht ~ehflvior and D ~ e l o p ~ e 6,167-174. nt, therapy ~ a n u a l f othe r S h e ~ie ldPsycho~herflpy plan, N.& Cassidy, J. (1985) Security Project (SAPU Memo 733). University of Shefin infancy, chil~hoodand adulthood: A move field. M., Rees, A.,Hardy, G.E., to the level of representation. M o n ~ ~ f lofp hthe Shapiro, D.A., Barkham, S o c i e ~ f o r ~ e s e ain r cChild h Develop~ent, 50~66Reynolds, S. & Startup, M. (1994) Effects of 104. duration and severity of depression on the effectiveness of cognitive-behavioral and Marcus, D.K.& Nardone, M.E. (1992) Depression psychodynamic-interpersonalpsychotherapy. and interpersonal rejection. ClinicuZ psycho lo^ ~ournfllof Consultingand Clinical P s Y c ~ o l o62, ~, Review, 12,433-449. McCrae, R.& Costa, P. (1988) Recalled parent-child 522-534. relations and adult personality. ~ o ~ r nofa lPerson- Strupp, H.H., Sandell, J.A., Waterhouse, G.J.,
~ h a p t 6: e~ A psychodynamic-infevpersonal model O'Malley, S.S. & Anderson,J.L. (1982) Psychodynamic therapy: Theory and research. In: Short-term Psyckotkerapiesfor Depression (ed. J. Rush). John Wiley & Sons, New York. van Ijzendoorn, M.H.(1995)Adult attachment representations, parental responsiveness, and
infant attachment a meta-analysis on the predictive validity of the Adult Attachment Interview. Psychological ~ u l l e ~ i117, n , 387-403. Weissrnan, M M . & Paykel, E.S. (1974) The Depressed ~ o ~A Study a n of~Social ~ e l ~ t ~ o n s ~ ~ p s . University of Chicago Press, Chicago.
efore the rise of cognitive-behavioural therapyin the 1970s, psychological t r e a ~ e n tof s depression were confined to the broadofrange psychodynamic, interpersonal and experiential approaches reviewed in this chapter. Although diverse, these typically share an emphasis upon empathic attentionto, and consideration of, the patient’s emotions and subjective experiences as being necessary to the alleviation of depression. Research has focused onbrief (a maximum of about 6 months of weekly sessions) formsof these treatments, with some attention also paid to extensions of treatment at a reduced frequency in an effort to prevent relapse or recurrence of depression. ~ s ~ c h o d ~ n atherapy m i c (Luborsky, 1984; Strupp (4r:Binder, 1984) derives from psychoanalysis,as developedby Freud and his associates. Key elements ng ~rocesses via i ~ t ~ r e t a t iand o n the emphasis upon include u n c o ~ ~ i~nconscious the r e - e n a c ~ e nof t significant processesin the trans~erence,or here-and-now relatio~~ between p patient and therapist. Theorigin of i n t e ~ ~ s o ntherapy al (Klerman et al., 2984) canbe traced to the work of Harry Stack Sullivan (Perry, conte~t,the 1982). Its central idea is that depression arises in an inter~ersonal c ~ a ~ ~ c a and t i o nrenegotiationof which will reduce the intensity of depressive ~ymptoms.Brief ands ~ c t u r e forms d of psychodynamic therapy increasingly of the desshare this emphasis, leading to the adoption for research purposes ignation, psychodynamic-intevpersonal therapy. As described by Hardy (see Chapter 6), this synthesis accords well with recent theoretical developments using attachment theory to account for psychopathology, includtherapy are its ing depression. The defining characteristics of e~perientia~ emphasis onthe facilitation of the patient’se ~ ~ e r ~ e ~ ascits i nkey g task, and on the thera~e~tic ~elationsh~p as curative. Rogers’(1951) Client-Centred Therapy, Perls et al.’s (1951) Gestalt Therapy, Gendlin’s (1981) Focusing, and Greenberg et al.’s (1993) Process ~ x ~ e r i e n tTherapy ~al are leading examples. A shared feature of all these methods is that their primary conceptual origins are with clinical practice rather than, asin the case of cognitive-behavioural methods (Chapter 8), an external, non-clinical body of scientific theory and knowledge. Until recently, high-quality controlled trials of these methods were rare. This lack is now being remedied. Trials with defined
diagnostic groups using manuals to specify the treatment procedures, together with ratings of therapists’ adherenceand competence to establishthat these have been deliveredwith integrity, are now feasible. Much of the evidential base for these procedures comes from process research analysing the components of treatment and their relationship toits effects (Elliott, 1995; Shapiro, 1996). Accordingly, clinical trials will be reviewed alongside studies of the mechanisms of change. A drawback of the lattermust be acknowledgedat the outset, however. The scientific focus of much process research has been on understanding the treatment rather than the disorder; consequently, diagnostic issues are frequently de-emphasized, although many studies include a substantial proportion of patients with at least mild or moderate depression. Strictly speaking, the specific applicabilityof many of these findings to formally diagnosed depression remains to be demonstrated, although there is seldom prima-facie evidencedoubt to this,
An overview of evidence from trials of psychodynamic, interpersonal and experiential therapies of depression can be gained from systematic reviews (meta-analyses), This approach to reviewingliterature was introduced to the health caredomain via reviewsof psychological treatmentsin mental health (Smith &:Glass, 1977;Smith et al., 1980; Shapiro&:Shapiro, 1982). Robinson et al. (1990)produced a landmark review of depression treatments. Psychodynamic and interpersonal treatments are discussed inthe practice guidelinesof the American Psychiatric Association (1993) and the US Agency for Health Care Policy and Research (Depression Guideline Panel, 1993), the latter guideline incorporating findingsof a new Systematic review. Roth and Fona were commissioned by the UK Department of Health to inform a review of psychotherapy services with ananalysis of outcome research, including a detailed analysisof depression treatments. Early meta-analytic reviews of depression treatments encompassing psychodyna~icmethods alongside other approaches produced mixed results. Neitzel et al. (1987) found no relationship between treatment type and outcome. h contrast, Dobson(1989)found ’other psychotherapies’, manyof which were dynamic therapies, to be inferior to cognitive therapy. Robinson et al. (1990) highlighted studies comparing four types of psychotherapy of depression. Their bibliography suggests that the ’general verbal’ category mainly comprised psychodynamic and experiential methods. This category yielded outcomes that were inferior to those for cognitive, behavioural and co~tive-behaviouralmethods. However, investigators’ allegiance to one or other of the therapeutic methods under test, assessed from cornrnents in the introductory section of each report, was strongly related to the outcome
143
of a given study: investigators tended to find in favour of their preferred method. Statistical correction for allegiance abolished the differences between treatment categories. This reflects the fact that most researchers who have compared psychotherapiesof depression are more expert and in, committed to, cognitive, behavioural and cognitive-behavioural methods than to psychodynamic, interpersonal and experiential approaches. The legitimacy of this statistical correction is open to debate, however. Proponents of cognitive and behavioural approaches (Giles, 1993) argue that investigator allegiance merely reflectsthe weight of objective evidence for their methods. Jacobson and Hollon (1996) attribute allegiance effects to differences in expertise or perspectiverather than any attempt to bias the nature of the findings. Thisshows the need for more research by investigators with allegiance to dynamicand interpersonal methods, or by teams comprising advocates of each of the treatments being compared, in order to level the playing field. The United States Public Health Service’s Agency for Health Care Policy and Research commissioned a review using Bayesian statistical methods to derive from published data the probability that each of several treatment methods will be successful in the treatmentof a depressed patient (Depression Guideline Panel, 1993). Overall efficacy percentages of 55.3, 34.8, 56.6 and 52.3 were obtained for beha~ioural,dynamic, cognitive, and erpe personal therapies, respectively. However, these percentages were based on just six trials of dynamic therapyand a single trial,the National Institute of Health (NIMH) study reviewed below,of interpersonal therapy, Behavioural and cognitive therapieswere represented by IO and 12 trials, respectively. A comparison, based on eight trials, of dynamic therapy with other formsof ~ s y c h o ~ e r a pfound y it slightly less effective (-7.8%). These figures suggest that dynamic therapy may be marginally inferior to other methods, but the extentof evidence available for inclusion was restrictedthe bydataanalytic strategy followed (Roth& Fonagy 1996).
Y Psychodynamic therapy applies the central principlesof psychoanalysis in a focused, relatively brief t r e a ~ e n ttypically , lasting for fewer than 30 weekly sessions (Alexander,1954; Sifneos, 3-972; Malan, 1976; Luborsky, 1984; Strupp & Binder, 1984). Psychoanalysis itself developedinto a long-term, intensive treatment aimed not at removing single symptoms or problematic behaviours, but attempting to restructure the entire personality (Roth &:Fonagy, 1996). Psychodynamic therapy, in contrast, assumes that increasing the patient’s insight into, or understanding of, emotionaldiff~culties and personal problems will resolve unconscious conflict, thereby alleviating symptoms. Insight is
2-44
~ ~ a ~7: ~ t es yr ~ ~ o d y int~r~ersonal n a ~ i ~ , and ex~eriential treat~ents achieved through interpretation of the patient’s behaviour and talk during treatment sessions. Problematic internal states, including feelings, desires and beliefs that are in conflict with one another, are believed to be encoded in patterns of interpersonal relationships which are brought to the patient’s attention within the treatment situation and in relation to the therapist. Unlike the stereotypeof the passive, often silent psychoanalyst, the modern psychodynamic therapist works actively to identify recurrent patterns of behaviour related to h ~ o t h e s i z e d unconscious conflicts. Conscious insight is viewed by contemporary psychodynamic therapists as less important than the emotional experience of the therapist’s acceptance of thoughts and feelings previously considered intolerableby the patient (Rothtk Fonagy, 1996).
Three reviews focusing on short-term, d ~ a m i c a l l orientated y psychotherapy 1992; have appeared since 1991 (Svartberg tk Stiles, 1991; ~rits-~hristoph, Lambert, 1995). Of these reviewers, Crits-Christoph (1992), whose inclusion criteria were the strictest, gave the most favourable verdict, and Svartberg and Stiles (1991) the least favourable. However, since each includes only a minority of studies of depression alongside trialsof other disorders, these studies arebest considered individually. Hersen et al. (1984) found that time-limited psychodynamic therapy was equally as effective as social skills training and amitriptyline. Therapists were all highly experienced and ’true believers’ in the method they used within the trial. None the less, fewer than half the patients in each treatment were deemed clinically improved after an initial 12-week period or following a further period of 6 months’ maintenance treatment. akstian (1979) found dynamic psychotherapy to be less effective than behaviour therapy on six out of 10 outcome measures, and less effective than amitriptyline on three measures, despite the use for both psychotherapies of experienced therapists working with their preferred treatment; forexample,onthe eck DepressionInventory (BDI), patients treated with dynamic and behaviour therapy recorded meansof 17 and 10 after treatment, both groups having started treatment with means of 27. Limiting the treatment to between eight and 12 sessions may have disadvantaged the dynamic treatment(Crits-~hristoph,1992). This is consistent with trends whose dynamic therapy patients in the data reported by Hersen et al. (1984)~ improved less markedly over the first 6 weeks of treatment but had caught up over 6 months of maintenance therapy. Kornblith et al. (1983) found group dynamic therapy as effective as behavioural self-control methods. Covi and Lipman (1987) found dynamic group therapy less effective than both group co~itive-behavioural therapy
alone and cognitive-behavioural therapy combined with medication. Three comparisons of psychodynamic and co~itive-behaviouraltherapies delivered in group formats to elderly, depressed patients have been reported. similar improvement over the course Gallagher and Thompson (1982) found of behavioural, cognitive, or 'relational/insight' therapy, but superior maintenance of gains over the year following treatment was detected amongst tho receiving the former two treatments, despite small samples of just IO patients in each group. Steueret al. (1984) found psychodynamic therapy less effective than co~tive-behaviouraltherapy on one of three outcome measures, the BDI, whilst Thompson et al.(1987) found no difference between psychodynamic, behavioural and cognitive therapy groups. In s~mmary,trials comparing brief psychodynamic therapy with other psychological treatmentsof depression have generally found its efficacy tobe similar to that of the comparison treatments, althougha minority of results appear slightly to favour the contrasted treatments. ore trials are urgently needed to guide the choice between psychodynamic treatments and other methods, especially medication.
nry et al.(1994) reviewed evidence onan issue that is central to the claims sychodynamic theory and practice in the treatment of depression as of other disorders: the effects of interpretations, specifically t~ a ~ s ~ e ~ e ~ responses to the therapy or therapist to ~ ~ t e ~ p ~ linking e t a t the ~ ~patient's ~ s core conflicts, which are the chief technical focus of psychodyna~c therapy. Several findings challenge the core assumptions of psychodynamic theory. The more recent, methodologically sound work, suggests that as far as transference inte~retationsare concerned, moreis not better and may even be damaging (Piper et al., qg1; Henry et al., 1993)~at least in brief therapy. Frequent transference~ t e r p r e t a ~ omay n s cause patients to feel criticized and so to withdraw. Four studies of patients' responses ttor ~ f e r ~ c e ~te~re~ti were consistent in showing that transference interpretations doelicit not the expected greater affective response or deeper experiencing on theofpart the atient. However, this may reflect the fact that most inte~retationsoffered by therapists are not judged by experts to be accurate. Recent evidence gests that transference interpretations may be only effective whenjudged to be accurate, and delivered relatively infrequently to more f highly ~ctionin~ patients (Henry et al., 1994). Henry et al. (1993) noted the perilsof inflexible, mechanistic adherence to treatment procedures such as transference ~ t e ~ r e t a t i oon n s the part of newly trained therapists. h contrast, experienced therapists have gained the opportunity to learn how to deliver treatment elements in a responsive manner (Stiles & Shapiro, 1989), attuned to the
patient’s subtlec o ~ ~ c a t i ofo readiness, n or otherwise, to absorb co~tru~ively the challenging information often contained within interpretations. The benefitsof accurate rather than inaccurate interpretations can only be properly assessed if the patient’s problems can be formulatedin a reliable and reproducible way. Progress is now being this made in direction; for example, a seriesof papers d e m o ~ ~ a t e d s uconvergence b s ~ t i ~ mong seven operational measures of ~~sference-related constructs applied to a single sample inte~iew (Luborsky et al., 1994); for example,the plan formulation method (Curtiset al., 1995) describesthe manner in whichpatient a will work in the therapeutic relationship to disconfirm individual pathogenic beliefs, overcome problems and achieve goals. Therapist interventions compatible with a plan formulated in object relations terms predicted in-session patient progress &(Tishby Messer, 1995).
A useful account of inte~ersonaltherapy (IPT) is given by ~ e i s s m a nand Markowitz (1994). Inaddition to the adult major depressionof interest tous here, these authors also review the developing use of IPT with related populations (including depressed adolescents, depressed patients with HIV, dysthymic patients), as well with as other psychiatricdisorders such as drug abuse and bulimia. IPT is based onthe assumption (Sullivan, 1953)that depression occurs in an environmental, more specifically interpersonal, context and that clarifyin and rene~otiatingthe context associated with the onset of symptoms is important to recovery ( ~ e i s s m a net al., 1981). This approach builds on extensive evidence linking stress and life eventswith the onsetof depression. It focuses on the patient’s present relationshipsrather than seeking originsof problems in the past, although overall patterns in the patient’s interpersonal relationships are identified by reviewing previous depressive episodes, early family relationships, and previous significant relationships and friendship patterns. The therapist is active, supportive and quite didactic in style.T r e a ~ e n t begins with a d i a ~ o s t i or c explanatory phase, h i g ~ i g h t i nthe ~ ways in which the patient’s current functioning, social relationshipsand expectations may have influencedhis/her depression. Thepatient is giventhe sick role, which excuses from overwhelming social obligations but requires the patient to work in treatment to return to full functioning. The approach is educational, linking depressive symptoms to one of four interpersonal domains: grief, interpersonal role disputes, role tran§itions orinterpersonal deficits. In the second phase of t r e a ~ e n the t therapist follows strategies specific to one of the four domains. If grief isthe focus, for example, the therapist aids m o u ~ g
~ ~ a p7:t~es y~c ~ o d y n a ~ i c , i n t e ~ pand e ~ sexpe~iential onal t~eat~ents and the uptakeof new relationships and activities to make good the loss. In the final phase of treatment, attention is focused on the gains achieved and a preventive approach to identifying and overcoming depressive symptoms arising in the future.
et al. Two major trialsof IPT as acute treatment have been reported. DiMascio (1979; see also Weissman et al., 1979) found that IPT and imipramine were equally effective in reducing the symptoms of unipolar depressives overall. At the level of symptom clusters, it was apparent that the drug was very rapidly effective in reducing sleep disturbance, whilst IPT was more rapid than pharmacotherapy in reducing mood disturbance and apathy. This suggests that patients with predominantly vegetative complaints may do bett with pharmacotherapy, whilst IPT may be especially useful for patients with predominantly emotional complaints and mood disturbance. The effectsof the two treatments in combination were essentially additive. One year later, most patients were asymptomatic, withno differences betweenthe groups, although many had received further drug and/or psychotherapeutic treatmen (Weissman et al., 1981). There was evidence, however, that IPT resulted in better social functioningat 1-year follow-up, suggestinga long-term benefit not apparent after4 months of treatment. The National Institute of Mental Health’s Treatment of Depression Collaborative Research Program (TDCW) has attracted widespread media and professional attention becauseof its unprecedented scale and, in some respects, disappointing results. Across three treatment sites, IPT was compared with Beck et al.’s (1-979)Cognitive Therapy(CT), the then standard tricyclic antidepressant imipramine, and a control condition in which placebo tablets were accompanied aby clinical management protocol identical to that used with the active drug. Patients met criteria for major depressive disorder. Elkin (1994) presentsan overview of the study and its findings. Therapists adhered to the requirements of the different treatment protocols (Hill et al., 3992). However, there were essentially no efficacy differences in respect of depressive symptoms, overall functioning (Elkin et al., 1989) or specific areasof functioning targetedby each of the psychotherapies (Imber et al., 1990)~as measured at the end of treatment between IPT, cognitivebehavioural therapy and imipramine. For example, IPT was no more effective thanthe other treatments in e ~ ~ c i patients’ n g social adjustment. ~terestingly, amongst patients whose social adjustment was relatively good at the outset of treatment, IPT was more effective than the other treatments, suggesting that good social functioning is required for patients to take advantage of et al., 1991). interpersonal strategiesto recover from depression (Sotsky
The response of depressive symptomsto i ~ p r a m i n was e somewhat more 1treatments ( ~ a t k i nets al., 1993). Twelve weeks medication weredoing s i ~ ~ c a n tbetter l y th psychotherapies, although this differencewas slight (and lly reliable)4 weeks later, Patients were subdivided into groups presenting with relatively severe vs. relatively mild depressive sy toms, and again into groups with relatively severe vs. relatively mild impairment of functioning (Elkinet al., 1989). These analyses revealedthat amongst relatively severely depressed patients, active medication was substantially more effective than placebo, with the two psycho~erapiesa&ev inte~ediateresultswhichremainedindis able from one another. The placebo condition yielded su~risinglygood outcomes amongst mildly depressed patients. The main findings were broadly confirmed by subsequent re-analysis using state-of-the-art statistical techniques to include alldata patients within a single analysis (Gibbons et al., 1993). However,the greater power of this analysis yielded somewhat stronger conclusions regarding the comparative efficacy of the treatment conditions among more severe those who were more severely depressed before trea~ent, and IPT were superior to GT and placebo; among those who were both more severely depressed and more functionally impaired before treatment, i ~ ~ r a m i was n e superior to both psychological treatments and the placebo condition (Elkinet al., 1996). In contrast to the modest effect of treatment conditions on outcomes in this study there were findings that perfectionistic, self-critical patients did less well in all treatments. In addition, the quality of the therapeutic relationship, measured earlyin therapy, predicted improvement, especially st those patients who were moderately “perfectionistic; the therape~tic hip did not strongly predict outcome amongst patients who were either highly self-critical or not at all self-critical (Blatt et al., 1996).~oderately self-critical patients may be thought of as those most ready to engage most fully in a therapeutic relationship whose goal is personal change,and it is these patients who are most affected by the quality of that relationship. Patients were followed for 18 months after endthe of treatment (Sheaet al., 1992a). h common with other studies of short-term depression treatments, relatively small percentages (19% of those receivingi ~ p r a m i n e20% , of those receiving placebo,26% of those receivingIPT, and 30% of those receivingGT) of those with follow-up data both met a stringent recovery criterion post-treatment and did not relapse at any time over the follow-up period. there were no statistically significant differences, CT appeared to do somewhat better overallthan the other treatmentsduring the follow-up. However, the results of such a ’naturalistic’ follow-up, in which patients are
free to take up further treatments, are notoriously difficult tointerpret. For example, focusing on the relapse rate of those initially making a good response to treatment may result in a biased comparison because different types of patients respond to eachtreatment. In addition, sample sizes were too small to detect modest effects with any confidence. Controversy concerning the TDCRP has arisen mainly from adherents of co~nitive-behavioural and pharmacological treatments, for each of whom findings could be considered somewhat disappointing. From a carefully reasoned cognitive-behavioural viewpoint, Jacobson and Hollon (199~) pointed to evidence suggestive of differential effectiveness of the psychological ~ e a ~ e natt sthe p a r t i c i p a ~ gsites, especially among the more severe patients. Although the anonymity of each site is protected for confidentiality reasons, therapists at the site with best CT outcomes were also rated (nonsi~ificantly)more competentin CT than those at the othertwo sites (Elkinet al., 1996). Were this site to be the one generally considered to have greatest expertise in, and commitment to,CT the results could be~terpretedto suggest that the efficacyof CT was underestimatedby the inclusionof the other two sites. From a pharmacotherapeuticperspective, Klein (1996) suggested that the TDCRIP’s pharmacotherapy condition was not optimally conducted, noting that patients who dropped out of the pharmacotherapy condition were rated by their therapists as having shown inferior compliance with medication. K and Ross’s (1993) reanalysis of TDCRP data found a clear advantage to IPT over CT in patients with more severe depression. Klein (1996) inte su~gestingthat psychotherfinding, alongside those of Sotsky et al. (1991)~ as apies may be counterproduc~veif they focus on areas of difficulty for the patient, and are at their most helpful when providing emotional support, usable coping skills and success experiences (Frank& Frank, 1991). A further piece of evidence concerning indications for IPT was obtained found that patients with endogenous depression by Prusoffet al.(1980). It was responded better to either pharmacotherapy alone or toIPT with pharmacotherapy than to IPT alone; whereas patients whose depression was not endogenous responded better to either IPT alone or the combined trea~ent than to pharmacotherapy alone. More generally, comparisons between psychotherapies and drug treatments become outmoded as new drug therapies are adopted. For example, if serotonin re-uptakeinhibitors are more widely tolerated because of lower side-effect profilesthan tricyclics, the rangeof patients who can be effectively treated by medication may be broader than it was when the drugpsychotherapy comparisonsof the 1970sand 1980s wereundertaken. Equally, of course, that literature will beoutdated by advances in psychotherapeutic treatment.
~ ~ a p7:tPsyc~odyna~ic, e ~ inte~ersonal and experiential t ~ e a t ~ e n t s
Studies of continuation or maintenance treatment (offered after remissionof the acute episode in order to prevent relapse or recurrence of depression) must be distinguished from studies (such as that by Shea et al,, ~ggza, discussed above) examining the extent to which gains made during the atment of a depressive episode aresustained over the months following treatment termination. ~terpersonaltherapy was first tested an $-month in c o n ~ ~ ~ ttrial i o nwith depressed women who had responded to 4-6 weeks of a m i t r i ~ ~ l i (Herman ne et al.,1974). Onentering the continuation phase, patients were randomizedto weekly IPT or a low-contact control condition. Two months later, patients were further randomized to continuing medication, or placebo no tablets. Thus there were six arms to the trial. Findings indicated that continuing pharmacotherapy prevented relapse and worsening of syrnptoms (Paykelet al., 1975)~ whilst IPT benefited social adjustment (Weissman et al., 1974). In a study by Franket al. (1990)~ patients in at least theirthird depressive episode were given a Combination of imipramine and PT. This combined treatment was continued for 17 weeks after symptomatic remission. In the maintenance phase, patients were randomized to one of five treatment conditions: a maintenance formof IPT-monthly sessions-was offered alone, or in combination with continued imipramineat the full dosage used in acute treatment (mean 200mg per day), or in combination with placebo; and medication clinic visits were offered with full imipramine medication or with placebo. Monthly IPT sessions substantially lengthened the time between episodes amongst patients not receiving active medication, represented by mean survival times of $2 and 74 weeks for the two groups receiving IPT vs. 45 weeks for those receiving placebo alone. However, monthly IPT had sessions little impact on those receiving active medication, whose survival meantimes were 131 and 124 weeks, respectively, with and without IPT. This result suggests that continuing medicationat the acute dosage level is the best way to prevent a recurrence. However, the alternative of monthly IPT sessions is worth considering as a less toxic alternative, although one carrying greater risk of recurrence. MonthlyIPT sessions did not appear to reduce the riskof recurrence amongst those receiving active medication.
Research from two major American groups has identifiedIPT as a promising treatment of the acute, continuationand maintenance phasesof depression. The numberof trials is less impressive than their quality and the consistency
of theirresults. In addition,the ’S TDCW suggestedthatIPT at least equals the more extensively researched cognitive-behavioural methods, although debate continues concerning this study’s delivery of CT.
There is promising evidence linking the technical integrity of IPT to its effects. O’Malleyet al. (1988) found that supervisors’ ratin treatment session, encompassing the quality of problem-ori quality of techniques, general IPT skills, and overall session quality, predicted depressed patients’ self-rated change and clinician ratings of the patient’s apathy at the end of treatment, over and above the effects of patients’ p r e t r e a ~ e ncharacteristics. t of the 1990 study (Frank et al., 1990) of Frank et al.’s (1991) further analysis maintenance IPT examined the contribution of the qualityof IPT sessions to the lengthof time that formerly depressed patients, not receiving active medication, remained well during the 3-year maintenance period. Sessions were rated for the specificity and purity of IPT ~terventions.Patients whose sessions were rated more specifically interpersonal did much better than those whose sessions were rated less so, with median times to recurrence of 102 vs. weeks, respectively.This finding could notbe explained in termsof other patient characteristics, such as severity of depression or comorbid personality disorder, that might be suspected of influencing both recurrence and therapists’ behaviour. Nor was it due to consistent differences between therapists in their use of IPT techniques withall their patients. Rather,it appeared to reflect the dyadic process between patient and therapist. It is of interest that the survival times of those with high-quality IPT sessions were quite similar to thoseof patients receiving medication.
As psychodynamic therapy has grown more focused and its emphasis upon interpersonal processes and problems has become more explicit, the term Psychodynamic-infeupersonalhas come into use to differentiate this approach more clearlyfrom earlier p s y c h o d ~ amethods ~c with their on intrapsychic conflicts; for example, this designation fits the methods of Lu~orsky(1984) and Strupp and Binder (1984). In the treatment of depression, the second Sheffield Psychotherapy Project(SPP2; Shapiro et al., 1994,1995) compared a psychodynamic-in~erpersonal(PI) method, based on Hobson’s (1985) ’Conversational Model’ withco~itive-behavioural(CB) therapy. The Sheffiel~PI treatment uses a psychodyna~ctheory of relations hi^ disturbances in which the therapeutic relationship is used as a vehicle for
understanding and modifying interpersonal difficulties which are viewed as primary in the origins of depression. The therapist’s contributions include adopting a negotiating style, developing a language ofm u ~ ausing ~ ~ metaphor , to enhance the immediacyof experienced affect, focusing on here-and-now experiences, and offering hypotheses to express understanding of the patient’s experiences. These hypotheses are based on psychodynamic inte~ersonal and concepts, butare offered tentatively as aids to the patient’s self-understandin rather than being delivered authoritatively as revealed truth. A positive bond between patient and therapist is maintained, and challenging or unwelcome ideas are not pursued at the expense of this. The patient-therapist relationship is seen as a microcosm of the patient’s relationships outside treatment, and a readily available source of data concerning the patient’s problems in relating to others and of hypotheses with which to build a new understan~g of these, For example, Kohut’s analysis of the role of narcissism in human development (Mollon & Parry; 1984) may enable the therapist to help a patient to understandhis or her apparent need for appreciative ’mirroring’ by others, including-during treatment sessions-the therapist. Similarly, attachment theory (see Chapter 6) serves the therapist asa source of hypotheses concerning the patient’s responses to experienced or feared loss. Manifestations of the patient’s attachment patterns within the therapeutic relationship can now be assessed reliably (Mallinc~rodtet al., 1995).
“)
Within SPP2, 117 patients completed either eight or 16 weekly sessions of either PI or CB therapy in a randomized design, Each patient scored 16 or more on the BDI, attained an Index of Definition on the Present-State Examination of at least five (indicative of current psychiatric disorder), and for had met RSM-I11 (American Psychiatric Association, 1980) criteriaa major depressive episode during the 3 months preceding treatment. Prior to randomization, patients were stratified into three levels of depression severity. To separate for scientific purposes the effects of the specific treatment methods from the influence of personal characteristicsof therapists choosing to practise each method, five research clinical psychologists sympathetic to both approaches were trained to carry out both PI and CB therapies. Their adherence to the contrasting treatment protocols was established via ratings of randomly selected sessions (Startup & Shapiro, 1993) On most measures, CB and PI therapies were equally effective, regardless BR1showed of severity of depression or the duration of treatment, however, the a modest advantage CB to over PI therapy. Irrespectiveof the treatment method *
~ h a p t e 7: r ~sychodynamic,interpersonal and e~perientia2 treatm~nts used, more severely depressed patients responded~betterto 16 than to eight sessions (Shapiro et al., 1994). A 1-year follow-up suggested that eight sessions of PI therapy may be insufficient to secure subsequent maintenance of gains, with patients in that condition reporting more depressive symptoms than those in the other three groups (Shapiro et al., 1995). Findings also indicatedthat not allpatients may benefit equally from PI therapy: consistent with a review by Sheaet al. ( q p b ) , those with comorbid cluster C personality disorders (avoidant, dependent or obsessive-compulsive personalities) did less well in PI therapy, whereas this comorbidity had no effect on response CB to therapy (Hardy et al., 1995a). Similarly, response to PI therapy was reduced amongst patients approaching treatment with a less psycholo~calapproach to the treatment of psychological problems, whereas these prior attitudes had no effect on response toCB therapy (Hardy et al., 1995b). echanisms of psychodynamic-interpersonaltherapy
In a strategic approach to delineating the mechanisms of psychotherapeutic change (Shapiro,~ 9 9 5both ) ~ quantitative and qualitative methods are used to identify factors contributing to the effects of PI therapy in depression, systematically contrasted with CB therapy. Before linking these process parameters to outcome, it is necessary to assess fully the similarities and differences between the two treatments across time: illustrative examples follow. Quantitative findings have included that by Agnew et al. (1997) in which the qualityof the relationship betweenpatient and therapist strongly predicted the extent of clinical improvement achieved by the patient. Stiles et al. (1997) plotted therapists’ intentions on a session-by-session basisin both PI and CB therapies. Distinctive featuresof PI therapy included an extended ’working phase’ in themiddle sessions of treatment involving a gradually increasing focuson encouraging changeand a gradually decreasing focus on affect, consistent with a gradual shift from exploration to implementation. Ongoing work on the Sheffield data is assessing the role of therapists’ responsiveness to the requirementsand characteristicsof individual patients in shaping the outcomes of treatment. For example,Hardy et al. (in press) found that therapists used more affectiveand relations~p-orientatedinterventions with patients whose attachment style (seeChapter 6) had been assessed as a~ous-ambivalent,particularly (but not solely) PI therapy. in Therapists used more cognitive treatment methods with avoidant patients in PI therapy, although not as markedly asCB intherapy These differing patterns of therapist response weremade without knowledge of the questionnairedata on which the attachment classification was based. Outcomes of the three attachment groups were approximately equivalent, consistent with a view that the
~ ~ a ~7:t~ es yyc ~ ~ d y n a m i c , i n t e ~ e yand s o nex~eyiential al tyea~ments differences in treatment~plementationreflected appropriate responsiveness to clients’ attachment styles. Reynolds et al. (1996)analysed session-by-session changes over the course of PI and CB therapies inthe patient’s post-session moodand evaluation of the treatment session itself. Early in treatment, PI sessions were less smooth (i.e. more tense and uncomfortable)and less focused on problem-solving, but PI sessions changed morerapidly than CB sessions on these dimensions, so that later in the treatment sessions of both, treatments were equivalently positive. Similarly,patients rated their relationshipwith their therapists slightly worse in early PI sessionsand slightly better in later PI sessions, in comparison with the correspondingCB sessions. These findings are consistent with the hypothesis, derived from the assimilation model (Stiles et al., 1990)~ that early sessions of PI treatment involvethe uncovering and exploration of potentially distressing, warded off or avoided material, whilst later in treatment both methods converge on applications of new ~derstandings-albeit different understandings-to solving the presenting problems. Stiles et al. (1995) also conceptualized the goal of the Sheffield PItherapy of depression as being~ to~ o ~theo~t ae t iassi~ilation ~ t ~ o~~~o~l~atic e~~iences (Stiles et al., 1990). This goal is achieved through the following five tasks. x Build a therapeutic relationship characterized bytrust and mutuality on the basisof which thepatient is encouragedand enabled to experience rather than to ward off problematic feelings. Promote more intense experiencing and more open expression and exploration of feelings. Draw attention to patient patterns of feeling and activity which result from ifficulties in experiencing, expression, interpersonal or relationships. Promote the patient’s self-~derstandingof difficulties by linking patterns of feeling and activity within the therapeutic relationship with patterns in significant relationshipsoutside therapy. Explore and experiment with alternative activities usingthe therapeutic relationship as the vehicle. Stiles et al. (1995) illustrated this analysiswith case material concerning an individual patient’s assimilation of a single theme-her difficulty in expressing feelings-that appeared to bean aspect of her successfully treated depression. The therapist’s pursuit of tasks 2,3 and 4 above,and the patient’s associated progress toward assimilation of the problematic theme, were revealed in the session transcripts.
Experiential approaches covered in a review by Greenberg et al. (1994) include client-centred therapy (Rogers, 1951), Gestalt therapy (Perls et al.,
1951), the experiential approaches of Gendlin (1981) and Mahrer(1983)~ and a et cluster of emotionally focused expressive approaches (Janov, 1970; Pierce al., 1983). Existential approaches (e.g. Bugental, 1978; Yalom, 1980) fall within All define the facilitatheir scope but coverage is limited by a lack of research. ~ ckey ~ ~ therapeutic g task, and view the tion of the patient’s e ~ p e ~ ~ase the therapeutic relationship as potentially curative. The relationship with the therapist provides the patient withnew, a emotionally validating experience and an opportunity to discriminate between past and present and to discover and own experience. In the past 15 years, new, more structured and active experiential therapies have emerged (e.g. Greenberg et al., 1993). These incorporate specific procedures, such as the Gestalt therapy technique of the ’emptychair’ in which the patient alternately takes the role of the self and of the other party to a problematic relationship. Such procedures are designed to promote the patient’s processing of problematic emotions, and their well-specified components lend themselvesto productive empirical research.
Greenberg et al. (1994; enlarged by Elliott, 1996) present a meta-analysis of research on the effectiveness of experiential therapies. This literature is not extensive, but its findings are broadly supportive. Unfortunately for present purposes, experiential therapies have traditionally neglected diagnostic issues and so in many studies an indeterminate proportion of patients are likely to have met criteria for major depression. Sixty-three studies were analysed, including 27 with waiting-list or no-treatment conditions, and 31 studies comparing experiential therapies with other treatments. These ranged from psychoeducational interventions or treatment-as-usual conditions to cognitive, behaviouralor psychodynamic therapies. A mean pre-post effectsize of 1.13 standard deviation units, and a mean control-referenced effect size of 1.04 standard deviation units, indicated substantial overall efficacy. Such effects place the average treated patient beyond the 80th percentile of control patients. Four studies of depression yielded pre-post effect sizes of 1.22, 1.35, 1.52 and 1.18 standard dqviation units, thoroughly representative of the overall, mixed-diagnosis population embraced by the meta-analysis. The average difference between experiential and other treatments was essentially zero. A more specific comparison was made between experiential and cognitive and behavioural methods. Seventeen studies yielded an overall effect size advantage of 0.23 standard deviation units in favour of cognitive and behavioural treatments over experiential treatments. Although sugges~ve, this difference was si~ficantly smaller than the 0.40 criterion for a clinicallym e a n i n ~advantage ~l of one ~ e a ~ eover n t another(Elliott et al., 1993).
Illustrative studies availablein English include a trialof the empty chair method’ whichde~onstratedits superiority to aps~choeducationalcontrol ically meaningful improvements in symptoms of dewell as interpersonal difficulties (Paivi S for major depressivedis y and Focused Expressive Psychotherapy g directed fantasy andthe-empty chair technique, achieved overall benefits similar to those of a non-directive condition in which own initiative to overcome patients were supported and enco~aged to use their p r o b l e ~via s weekly telephone contacts and self-help readings. I ~ p o r t a n thowever, l~ different patients prospered under each treatment, in accordance with the authors’ predictions based on theories of personality: behavioural and sy~ptom- orientate^ procedures of cognitive therapy were most helpful to ’externalizing’ patients, those who control stress poorly, are irritable and impulsive, and tendto blame others. In contrast, these patients were not helpedby the non-directive approach. Furthermore, ’internalizing’ patients (who blame themselves,as is typical of many with depression), did nitive therapy and best with non-directive therapy. In addition, patients showing high ’resistance potential’ faredbest with non-directive therapy and worst with focused expressive therapy.
~~s~ ~
~
~ is a ~ process y s ~research s strategy that involves the detailed study of
rocesses that individuals actually use to perform tasks.It focuses on understanding the processof task solution and building explanatory models of the mecha~smsinvolved therein. For example, Greenberg and Foerster (1996) focused on the resolutionof ’unfinished business’ (dissatisfaction about the nature of relationships in the past and associated feelings of disappointment, resentment, and grief) by the empty chair technique (Paivio & Greenber~, 1995). Greenberg and Foerster(1996) used a’ra~ional-empirical’methodology of repeatedly cycling between rational conjecture and empirical observations to develop an intervention manual and to identify the components of the patient’s resolution performance. A refined model of the change process developed in this fashion was validatedby comparing 11 successful and 11 unsuccessful attempts at resolution. Four components, including intense expression of feeling, expression of need, shift in the representation of the other person, and self-validation or understand in^ of the other, were found to discriminate between successful and unsuccessful attempts. This approach to experiential therapy in particular has,de~onstratedthe benefits of analysing this treatment as a non-uniform’ differentiated process
in which patient and therapist engage in varied activities with specific et al., 1994). impacts at different points (Greenberg
Greenberg et al. (1994)conclude that experiential therapies have promisein the treatment of depression. They also warnof hindering factors identified in the literature, including therapist intrusion or pressure, therapist misatunement, insufficient therapist direction, and patient concerns such as lack of perceived safety, confusion or distraction. They note that highly autonomous or reactant patients may react negatively to the more directive experiential therapies, whereas more dependent or externally orie~tated patients may react negatively to unstructured, non-directive therapies.
rc
Our understanding of what makes for effective psychological treatment of depression is usefully informedby considering the broader l i t e r a ~ r relating e the processes of psychological treatments to their effects. A comprehensive oi treatment toclinical outoverview of studies linking features or elements comes was conducted by Orlinsky et al. (1994).This review incorporated over zooo findings, and yielded consistent evidence identifying factors thatcontribute to treatment efficacy. These factors are highly congruent with the findings already reviewedin this chapter, and include the following points. z ~ ~ p l e ~ e n t a t i o n tkerape~tic o ~ t k e ’contract’:the patient’s~derstanding and acceptance of the rationale and procedures of therapy; the patient’s verbal activity; the therapist’s skill and adherence to the treatment model. These findings argue for clear, explicitly structured treatment approaches that are well explainedto the patient and competently followed by the therapist.This r e c o ~ e n d a t i o nis supported, for example,by Frank and colleagues(1991) findings on therapists’ contributions to the successof IPT. ~ r o c e ~ ~ r e s ~itk o lkin Z o~keyapy: ~ e ~ patients’ focus on life problems and core personal relationships, and therapists’ focus on patient problems, and, with sufficient tact and caution, on patients’ emotional reactions during sessions; therapists’ experiential confrontation and interpretation may alsobe beneficial, but this appears to depend on the patient’s readiness to accept these interventions; patients’ co-operation (as opposed to resistance) and positive affective arousal. These findings point to productive elements withi psychodynamic, interpersonal and experiential therapiesof depression and offer guidelines concerning their delivery. Tke ~ ~ a l ioft y~ e l a t i o n s ~ i ~ ~patient e t ~ e eand n tkeya~ist,or ’therapeutic alliance’, including role investment, coordination between the participants,
~ h a ~ t7:e ~r s y c h o d y n a ~ i c , i n t e ~ ~and r s o e~~eriential nal treat~ents communicative contact and affective attitude. These findings are especially powerful from the patient’s perspective, with a clear implication for management thatif the patientdoes not feelpositive about the relationship with the therapist, treatment will not succeed. Tke patien~~s openness (as opposed to defensiveness~:this implies that therapists must foster this by making therapyas ’safe’ as possible, confirmedby findings that critical or hostile responses by therapists interfere markedly with the treatment process and result in poor outcomes. These findings may explain why comorbid personality disorders impair the response of depressed patients to these treatments. 5 The pa~ient~s experiencin~ of positive i ~ p a c t s ~sessions, i t k i ~which therapists ( s i ~ i ~ c a n t lmay y ) fail to perceive although they are apparent to patients themselves and to outside observers. In summary, the circumstances in which psychological treatment of depression is likely to succeed include the following: a well-structured treatment, skilfully delivered in the contextof a safe, positive relationship between patient and therapist, in a manner responsiveto the patient’s requirements and productiveof positive feelings during sessions, that focuses on the patient’s key problem areas, will be effective with a patient who is open and ready to change.
There is reasonably good evidence for the efficacy of brief, structured forms of psychodynamic, interpersonal and experiential therapy in reducing the severity of depression or ending depressive episodes.This evidence comes trials, comparisons with other from a variety of sources, including controlled treatments of known efficacy, and single-group pre-post studies showing changes of comparable magnitude to those shown by other treatments independently supportedby controlled trials. The evidence is patchy rather than comprehensive, however.It is likely to remain so, unless resources comparable to those made availableby the pharmaceutical industry for drug trials h addition, thereis a become available for psychological treatment research. pressing need for exponents of these treatment methods to become more skilled in, and comitted to, the research enterprise. The use of these treatmentsin less-intensive formats over the months following successful treatment to prevent relapse or recurrence of depression warrants further study, based on promising results obtained with IPT. As notedby Persons et al. (1996), the American Psychiatric Association’s (1993) practise guideline offers an incomplete evaluation of psychodynamic therapy, overlooking the evidence of similarity in its effects to those of behavioural anddrug treatments. When directly compared in ’head-to-head’ trials
~ ~ a7: ~~ s y~ ~ e~ or~ y n a ~ i c ~
an^ i n e~xe ~r ~ ~ ri ~ ~s on~~ai larl
~
~
~
with cognitive-behavioural therapies or drug t r e a ~ e n t sthese , t r e a ~ e n t do s not fare badly. For example, co~tive-behavioural trea~ents may enjoy a slight advantage over the methods reviewed here, but any difference is small and may well reflect the preferences and skills of the ~ a j o r i t yof researchers et al. (1996) also note rather than substantially inferior effectiveness. Persons that the~ i d e l i noverstates e the argument that more severely or endogenously depressed out-patients require drug t r e a ~ e n(since t the evidence on this point is in fact mixed)as well as the case for combining antidepressant medication with psychot~erapy(unsupportedby the majority of comparative studies), Useful implications for the treatmentof depression canbe derived from the wider fieldof research on the processesof change in theset r e a ~ e n t as s applied to patient populations, including those not diagnosed with depression. For example, the evidence underscores the skill and sensitivity required for psychod~amicinterpretatio~sto be experienced as helpful rather than hindering or undermining. There is also evidence that the formation and maintenance of a positive relationship between patient and therapist is of paramount importance in d e t e r ~ n i n gthe outcome of treatment, and may outweigh the specific techniques used. Comparison, in the Sheffield depression research, with cognitive-behavioural t r e a ~ e n thighlights , many features in comxnon between these two approaches whilst also identifying specific points of difference. For example, it may be that the methods considered here are less universally acceptedby patients bringing differing attitudes and personality difficulties to the treatment of their depression than are cognitivebehavioural methods. From an empirical researcher's perspective, the future ~sychodynamic, of interpersonal and experiential therapies of depression looks rosier than ever before. "his is just as well, given the current demand for evidence-based treatment services (Avelineet al., 1995). Evidence alreadyto hand indicates that these methods have place a in the treatmentof depression; extension and refinement of that evidence over the coming decade will assuredly define more clearly the indications for these treatments and their modes of action.
Agnew, R.M., Stiles,W.B., Hardy, G.E.et al. (1997) American Psychiatric Association (1993) Practice guideline for major depressive disorder in Alliance structure and relations with psychoadults. A~epican ~ o ~of P~s ~ac hl ~ 150 a t ~(Suppl. , therapy outcome assessed by the Agnew Rela4), 1-126. tionship Measure( A R M ) . Alexander, F. (1954) Psychoanalysis and psycho- Anderson, E.M. & Lambert, M.J. (1995) Short-term therapy. Journal of the American Ps~choanul~tic dynamically oriented psychotherapy: a review and meta-analysis.C ~ i ~ i c a ~ PReview, s ~ c h15, o~o~ Associa~ion,11,722-733. American Psychiatric Association (1980) Diagnos503-554. tic and St~tistical~ a n u aofl Mental Disopdeps, 3rd Aveline, M,, Shapiro, D.A., Parry, G. & Freeman, F. (1995) Building research foundations for edn. American Psychiatric Association, Washpsychotherapy practice.In.Researc~ €ou~dations ington, DC.
for Psychotherapy Practice (eds M Aveline &
D. A.Shapiro),pp. 301-322. John Wiley & Sons, Chichester. Beck, A.T., Rush, A.J., Shaw, B.F. et al. (1979) Cognitive Therapy of ~epression.Guilford, New York. Beutler, L.E., Engle, D., Mohr, et D.al. (1991) Predictors of differential response to cognitive, experiential, and self-directed psychotherapeutic procedures. Journal of Consulting and Clinical ~ s y c h o l o59,333-340. ~, Blatt, S.J., Zuroff, DC., Quinlan, D.M. & Pilkonis, P.A. (1996) Interpersonal factors in brief treatment of depression: further analyses of the National Institute of Mental Health Treatment of Depression CollaborativeResearch Program. ~ o ~ r nofa Consulting l and Clinicalpsycho lo^, 64, 162-171. Bugental, J.F.T. (1978) Psychotherapy and Process: The Fundarnentals of an €xistential-hurnanistic Approach. Random House, New York. Covi, L. & Lipman, R.S. (1987) Cognitivebehavioral group psychotherapy combined with imipramine in major depression. Psycho23,173-176. pharrnacolo~ ~ulletin, of brief dyCrits-Christoph, P.(1992) The efficacy namic psychotherapy: a meta-analysis. Arnerican ~ournulof Psychiat~,149,151-158. Curtis, J.T., Silberschatz,G., Sampson,H. & Weiss, J. (1994) The plan formulation method. Psychotherapy Research,4, 197-207. D.Greenberg, Daldmp, R.J., Beutler, L.E., Engle, & L.S. (1988)Focused ExpressivePsychothera~:Freeing the Overcontrolled Patient. Guilford Press, New York. Depression Guideline Panel (1993) Cl~nicalPractice Guideline:~ ~ ~ e s siniPrirnary on Care,Vol. 2. Treatrnentof Major ~ ~ r e s s i oAgency n, for Health Care Policy and Research Publication no. 930551. US Department of Health and Human Services, Rockville, MD. B., Neu, C., DiMascio, A., Weissman, M., Prusoff, Zwilling, M. & Klerman, G. (1979) Differential symptom reduction by drugs and psychotherapy in acute depression. Archives of General Psychiat~,36,1450-1456. Dobson, K.S. (1989)Ameta-analysis of the efficacy of cognitive therapy for depression. ~ournalof C o n s ~ l t ~and n g ~1inicaZ psycho log^, 57, 414-419. Elkin, I. (1994) The NIMH Treatrnentof Depression Collaborative Research Program: where we began and where we are. In: Hundbook of Psychoth~apy and ~ehavior Change, 4th edn, 114-139 (eds A. E. Bergin& S. L. Garfield).John Wiley & Sons, New York. Elkin, I,, Shea, M.T., Collins, J.F. et al. (1989) Na-
tional Institute of Mental Health Treatment of Depression Collaborative Research Program: general effectiveness of treatments. Archives of General Psychiatry,46,971982. Elkin, I., Gibbons, R.D., Shea, M.T. et al. (1996) Science is not a trial (but it can sometimes be a ~Consulting and Clinical a tribulation). J o ~ r n of psycho lo^, 64,92--103. Elliott, R. (1995) Therapy process research and clinical practice:practical strategies. In: Research Practice, pp. 4 ~ 7 2 Foundationsfor Psychot~erapy (eds M. Aveline & D. A.Shapiro).John Wiley & Sons, Chichester. Elliott, R. (1996) Are~lient-centered/experiential therapies effective? A meta-analysis of outcome research. Paper presented at the Annual Meeting of the Society for Psychotherapy Research, Amelia Island, Florida, June 1996. Elliott, R., Stiles, W.B. & Shapiro, D.A. (1993) Are some psychotherapies more equivalent than (ed. others?h Handbook of E~ective P~chotherapy T.R. Giles), pp. 455-479. Plenum Press, New York. J.M.et al.(1990)Three Frank, E., Kupfer, D.J., Perel, year outcomes for maintenance therapies in r e c m n t depression.Archives of General Psychiatry, 47, 1093-1099. Frank, E., Kupfer, D.J., Wagner,E.F., McEachran, A.B. &Comes, C. (1991) Efficacy interpersonal of psychotherapy as a maintenance treatment of recurrent depression: contributing factors. Archives of General Psychiatry,48,1053-1059. Frank, J.D.& Frank, J.B. (1991)Persuas~onand Healing: A C o ~p a r is o n t o Ps y c h o ~h e r a p yJohns . Hopkins University Press, Baltimore. Gallagher, D.E. & Thompson, L.W. (1982) Treatment of major depressive disorder in older adult patients with brief psychotherapies. Psychotherapy: Theory, Research and Practice, 19,482-490. Gendlin, E.T. (1981) Focusing, 2nd edn. Bantam Books, New York. Gibbons, R.D., Hedeker, D., Elkin, I. et al. (1993) Some conceptual and statistical issues in analysis of longi~dinalpsychiatric data. Archives of General Psychiatry,50,739750. Giles, T.R.(ed.) (1993)Handbook of Efective Psychotherapy. Plenum Press, New York. Greenberg, L.S. & Foerster, F.S. (1996) Taskanalysis exemplified: the process of resolving unl Cons~ltingand finished business. J o ~ r n a of Clin~calpsycho lo^, 64,439-446. Greenberg,L., Rice,L.N. & Elliott, R.(1993)Facili-
tating E~otional Change: TheMo~ent-by-Mornen~ Process. Guilford, New York. Greenberg, L., Elliott, R. & Lietaer, G. (1994) Re-
search on experiential psychotherapies. In:
Handbook of Psychotherapy and Behavior Change (eds S. L. Garfield & A. E. Bergin), pp. 509-539. John Wiley & Sons, NewYork.
Clinical Psychology, 64, 81-87. Klein, D.F. & ROSS, D.C. (1993) Reanalysis of the National Institute of Mental Health Hardy, G.E., Barkham, M., Shapiro, D.A., Stiles, Treatment of Depression Collaborative Research W.B., Rees, A.& Reynolds, S. (1995a) Impact of Program general effectiveness report. NeuroCluster C personality disorders on outcomesof psychopha~acoZogy,8,241-251. contrasting brief psychotherapies for depres- Herman, G.L., DiMascio, A., Weissman, M.M., l and ClinicalPsychology, sion. ~ o uof ~ConsMlting Prusoff, B.A. & Paykel, E.S. (1974) Treatment of depression by drugs and psychotherapy. 63,997-1004. Hardy,G.E.,Barkham,M,,Shapiro,D.A., A ~ r i c a JoMrnal n of P s y c h i a ~131,186-191. ~, Reynolds, S., Rees, A. & Stiles, W.B. (1995b) Herman, G.L., Weissman, N.M.,Rounsaville, B.J. Client beliefs, expectancy and outcome in & Chevron, E.S. (1984) ~nterpersonalPsychocognitive-behavioural and psychodynamictherapy of ~ ~ r e s s i oBasic n . Books, New York. interpersonal psychotherapy. Brifish Journal Kornblith, S.J., Rehm, L.P., O'Hara, M.W. & of Clinical psycho lo^, 34, 555-569. Lamparski, D.M. (1983) The contributionof selfHardy, G.E., Stiles, W.B., Barkham, M. & Startup, reinforcement training and behavioral assignM. (in press) Therapist responsiveness to client ments to the efficacy of self-control therapy for interpersonal styles during time-limited treatdepression. Cognitive Therapy and Research, y, a ~ConsMlt~ng and ments for depression. ~ o u r n of 499-527. ClinicalPsychology. Luborsky, L. (1984)Principlesof PsychoanalyticPsy~ :Manual for SMpportive-~xpressive Henry, W.P.,S~pp,H.H.,Bu~er,S.F.,~acht,T.E. c h o t ~ e r a pA & Binder, J.L. (1993) The effects of trzaining in Treatment. Basic Books, New York. time-limited dynamic psychotherapy: changes Luborsky, L., Popp, C., Barber, JP. & Shapiro, D.A. (eds) (1994) Special issue: seven transferencein therapist behavior. ~ournalof ConsMlting and Clinical Psychology,61,434-440. related measures. ~sychotherapyResearch, 4, & Gaston, Henry, W.P., Strupp, W.H., Schacht, T.E. 151-286. L. (1994)Psychodynamic approaches. In: Hand- Mahrer, A.R. (1983) ~ x p e r i e n t iPsychotherapy: ~l book of P s y c ~ t h ~ aand p y BehaviorChange,4th edn Basic Practises. Brunner/Mazel, New York. (eds A. E. Bargin & S. L. Garfield), pp. 467-508. Malan, D.H. (1976) The Frontier of Brief Psychotherapy. Plenum Press, New York. John Wiley & Sons, New York. Hersen, M,, Bellak, A.S., Himmelhoch, J.M. & Mallinckrodt, B., Gantt, D.L. & Coble, H.M. (1995) Thase, M.E. (1984) Effects of social skill trainAttachment patterns in the psychotherapy relationship: development of the client attaching, amitriptyline, and psychotherapy in uniment to therapist scale. Jou~nalofcounseling Psypolar depressed women. Beh~viorTherapy, 15, 21-40. chology, 42,307-317. Hill, C.E., OGrady, K.E. & Elkin, I. (1992)Apply- McLean, P.D. & Hakstian, A.R. (1979) Clinical depression: comparative efficacy of outpatient ing the Collaborative Study Psychotherapy and Psytreatments.~oMrna1 of ConsMlting Clinical Rating Scale to rate therapist adherence in chology, 47,818-836. cognitive-behavior therapy, interpersonal therapy, and clinical management. ~ o u r n aof~ Mollon, P. & Parry, G. (1984) Thefragile setlf:narcissisticdisturbance and the protective function Consultingand Clinical Psychologyf60~73-79. l Psychoof depression. British ~ o u ~ nofa Me~ical Hobson, R.F. (1985) Forms of Feeling: The Heart of Psychotherapy.Basic Books, New York. k?Y, 571 137-145. Imber, S.D., Pilkonis, P.A., Sotsky,S.M.et al. (1990) Neitzel, M.T., Russell, R.L., Hemmings, K.A. & Mode-specific effects among three treatments forGretter, M.L. (1987) Clinical significance of psychotherapy for unipolar depression: a metadepression. ~ournalof Consulting and Clinical analytic approach tosocial comparison. ~ o ~ r n a ~ ~ ~ y c h o ~ o58,352-359. gy, of Consul~ing and ClinicalPsychology, 55, 156-161. Jacobson, N.S. & Hollon, S.D. (1996)Prospects for future comparisonsbetween drugs and psycho- OMalley, SS., Foley, S.H., Rounsaville, B.J. et al. (1988) Therapist competence and patient outtherapy: lessons from the CBT-vs.-pharmacome' in interpersonal psychotherapyof decotherapy exchange. Journal of Consulting and Clinical Psychology, 64, 104-108. pression. ~ o u r n a olf Consulting and Clinical Janov,A. (1970)The Primal Scream: Primal Therapy, Psychologyf56,496-501. Orlinsky, D.E., Grawe, K. & Parks, B.K. (1994) the Curefor Neurosis. Dell, New York. Process and outcome in psychotherapy: noch Klein, D.E (1996) Preventing hung juries about g einmal. In: H ~ n d b o o kof Psychotherapy and therapy studies. Journal of C o n s u ~ ~ i nand
a replication and refinement. P s y c h o ~ o g ~ c a ~ Behavior Change, 4th edn (eds A. E. Bergin & Bulletin, 92,581404. S. L. Garfield), pp. 270-376. John Wiley & Sons, Shapiro, D.A., Barkham,M,, Rees, A.et al. (1994) New York. Effects of treatment duration and severity of Paivio, S. & Greenberg, L.S. (1995) Resolving depression on the effectiveness of cognitive'unfinished business': efficacy of experiential behavioral and psychodynamic-inte~ersonal therapy using empty chair dialogue.~ournalof psychotherapy.~ournalof Consultingand Clinical Consultingand Clinical Psychology,63, 419-425. Psychology,62,522-534. Paykel, E.S., DiMascio, A., Haskell, D.& Prusoff, B.A. (1975) Effectsof maintenance amitrip~line Shapiro, D.A., Rees, A., Barkham,M. et al. (1995) and psychotherapy on symptoms of depression.Effects of treatment duration and severity of depression on the maintenanceof gains following PsychologicalMedicine, 5, 67-77. cognitive-behavioral and psychodynamic-interPerls, F.S., Hefferline, R.F. & Goodman, P. (1951) personal psychotherapy. ~ o ~of ~Consulting a l and Gestalt Therapy.Julian Press, New York. Clinical Psychology, 63, 378-387. Perry, H.S. (1982) Psychiatristof A ~ e r i c aThe : lye of Harry Stack Sullivan. Belknap Pressof Harvard Shea, M.T., Elkin, I., Imber, S.D. et al. (1992a) Course of depressive symptoms over follow-up: University Press, Cambridge, MA. findings from the National Institute of Mental Persons, J.B., Thase, M.E. & Crits-Christoph, P. Health Treatment of Depression Collaborative (1996) The role of psychotherapy in the treatment Research Program. Archives of General Psychiof depression: review of two practice guidelines. Archives of General P s yc h i a t ~53,283-290. , atry, 49,782-787. Pierce, R.A., Nichols, M.P. & DuBrin, J.R. (1983) Shea, M.T., Widiger, T.A. & Klein, M.H. (1992b) Comorbidity of personality disorders and E ~ o t i o n aExpression l in Psychotherapy. Gardner depression: implications for treatment.~ o ~ r n a l Press, New York. of Consulting and ClinicalPsychology, 60,857468. Piper, W.E., Azim, RA., Joyce, S.A.& McCullum, M. (1991) Transference interpretations, therapeuticSifneos, P.E. (1972) Short-ter~Psychotherapy and E ~ o t i o n aCrisis. l Harvard University Press, alliance and outcome in short-term individual Cambridge, MA. psychotherapy.Archives of General Psychiatry, 48, Smith, M.L. & Glass, G.V. (1977) Meta-analysis of 946-953. psychotherapy outcome studies. American Prusoff, B.A., Weissman, M.M., Klerman,G.L. & Rounsaville, B.J. (1980) Research diagnostic Psychologist,32,752-760. Smith, M.L., Glass, G.V. & Miller, T.I. (1980) The criteria subtypes of depression: their role as Benefiits of Psychotherapy. JohnsHopkins predictors of differential response to psychotherapy and drug treatment. Archives of General University Press, Baltimore. Psychiatry, 37, 796-801. Sotsky, S.M., Glass, D.R., Shea, M.T. et al. (1991) Patient predictors of response to psychotherapy Reynolds, S., Stiles, W.B., Barkham, M., Shapiro, D.A., Hardy, G.E. & Rees, A. (1996) Acceleration and pharmacotherapy: findings in the NIMH of changes in session impact during contrastTreatment of Depression Collaborative Research Program. A ~ e r i c a n ~ o u r nofa lPsychiatry, 148, ing time-limited psychotherapies. ~ournalof Consultingand Clinical Psychology,64, 577-586. 997-1008. Startup, M.J. & Shapiro, D.A. (1993) Therapist Robinson, L.A., Berman, J.S. & Neimeyer, R.A. treatment fidelity in Prescriptive vs. Exploratory (1990) Psychotherapy for the treatment of psychotherapy. British ~ournalof Clinical depression: a comprehensive review of controlled outcome research.Psychological ~ulletin, ~ ~ y ~321 h 443-456. ~ ~ ~ g y ~ Steuer, J.L., Mintz, J., Hammen, C.L. et al. (1984) 30-49. Cognitive-behavioral and psychodynamic Rogers, C.R. (1951) Client-Centered Therapy. group psychotherapy in treatment of geriatric Houghton Mifflin, Boston. ~ournalof Consul~ingand Clinical Roth, A.& Fonagy,P. (1996) at Worksfor ~ o ~ depression. : A Critical Review of Psychotherapy Research. psycho lo^, 52,180-189. Stiles, W.B. & Shapiro, D.A. (1989) Abuse of the Guilford Publications, New York. Shapiro, D.A. (1995) Finding out how psychodrug metaphor in psychotherapy process-outtherapies help people change. Ps~ckotherapy come research. Clinical Psychology Review, 9, Research, 5, 1-13. 521-543. Shapiro, D.A. (1996) Models of change in psycho- Stiles, W.B., Elliott, R., Lewellyn,S.P. et al. (1990) therapy. Current Opinion in P s y c k ~ t9,~177-181. , Assimilation of problematic experiences by Shapiro, D.A. & Shapiro, D. (1982) Meta-analysis clients in psychotherapy. Psychotherapy, 27, of comparative therapy outcome research: 411-420.
Stiles"W.B., Shapiro,D.A., Harper, H. &Morrison, two plan formulations. P ~ y c h o ~ ~ e rResearch, apy L.A. (1995) Therapist contributions to psycho5,7648. therapeutic assimilation: an alternative to the W a t h s , J.T., Leber, W.R., Imber, §.D. et al. (1993) drug metaphor. British ~ o u r n aof~~ e d i c a l NIMH Treatment of Depression Collaborative psycho lo^, 68,1-13. Research Program: temporal course of symptoStiles, W.B., Startup, M.,Hardy, G.E. et al. (1996) matic change. ~ournulof Consulting and Clinical Therapist session intentions in cognitiveP s y c ~61,858-864. ~ ~ , behavioral and psychod~amic-interpersonal Weissman, M.M. & Markowitz, J.C. (1994) Interpsychotherapy.~ o ~of rC on~ n~s e~l i ~ g ~ s y c h o lpersonal o~, es therapy: current status. A r ~ h ~ vof ~ n e r aPsychiut~, l 51, 599-606. vol. 43,4,402-414. Strupp, R.H. & Binder, J.L. (1984)Psychothe~apyin Weissman, M.M., Klerman, G.L., Paykel, E.S., a New Key: A Guide to T i ~ e - l ~ m i tDynamic ed Prusoff, B.A. 8r Hanson, B. (1974) Treatment Psycho~~erapy. Basic Books,New York. effects on the social adjustment of depressed ~, T ~ e o ~patients. Archives of General ~ s y c h i a t30,771Sullivan, H.S. (ed.) (1953) The ~nterpersona~ of Psychiatry. W.W. Norton, New York. 778. Svartberg, M. Csr Stiles, T.C. (1991) Comparative Weissman,M.M., Prusoff, B.A., DiMascio, A.et al. (1979)The efficacyof drugs and psychotherapy effects of short-term psychodynamic psychoin the treatment of acute depressive episodes. therapy: a meta-analysis. ~ournalof Consulting and C l ~ n ~ c a l P s y 59,704-714. ch~o~, ~ m e r ~ c~a on u r ~ofaPs l y c k ia t ~, 136,555-558. Thompson, L.W., Gallagher, D. & Breckenridge, Weissman, M.M.,Klerman, G.L., Prusoff, B.A. et al. (1981)Depressed outpatients:results one year L.S. (1987) Comparative effectiveness of after treatment with drugs and/or interpersonal psychotherapies for depressed elders. ~ o ~ r nofu l C o ~ s ~ l t iand n g Clinical psycho lo^, 55,385-390. ~, psychotherapy.~r c h iv eosf General P s y c h i a ~38, Tishby, 0. & Messer, S.B. (1995) The relationship 51-55. ent~al Basic between plan compa-tibilityof therapist inter- Yalom, I.D. (1980) ~ ~ ~ s ~ Psyc~otherapy. Books, New York, ventions and patient progress: a comparison of
ient learns some general principles about the way they may become and ain depressed which may help them to deal differently with problems in T is a short-term therapy, initially designed to be carried out over 12-20 sessions in 3-6 months. Recently, both briefer and longer, more intensive packages have been devised, in response to the varieties of need with which depressed patients present. GBT has been demonstrated to be a generally effective treatment for depression in the large number of studies that have accumulated since the original study of Rush et al. (1977). It has performed well in trials comparing minimal treatment controls and other treatments (Dobson, 1989). It has generally producedgreaterimprovementindepressivesymptomatolo than. no treatme~tor waiting-list controls (Comas-Diaz, 1981; Gall Thompson, 1982; Wilson,et al., 1983). Gomparisons with pharmacotherapy have yielded less clearcut results, principally because of methodological difficulties. Individual studies have involved relatively few patients and studies have been carried out in prestigo university centres where patients and therapists employed may not have bee representative of those in more usual settings. Many studies comparingG and pharmacotherapy have lacked drug-placebo controls, so it was not possible to d e t e r ~ n whether e the patients were pharmacologically responsive. F u r t h e ~ o r ethe , a d m i n i s ~ a ~ of o npharmaco~erapy has not been adgquatgly operationalized; in many studies patient compliance or plasma medication levels were not monitored, such that the adequacy of the pharmacotherapy was not checked (Hollon et al., 1991).As a result of these difficulties, although many clinical trials have demonstrated GBT and phar al., 197 comparable efficacy for depression (e.g. et Rush urphy et al., 1984; Hollon et al., 19cpa), to some critics this hypothesis has still not been ade~uatelytested. In addition, littleis known as yet about the effectiveness of CBT relative to other forms of psycho Robinson et al. (1990) carried out a comprehensive revi come research comprising all types of psychotherapy for depression using meta-analysis. Results indicated an advantage for CBT methods over strictly behavioural and'general verbal' psychotherapies. The authors raise the point, however, that thereis a relationship between'allegiance' to school of therapy and outcome. When this is corrected for, the comparative benefitCBT for disappears. Whilst we would agree that there might be a therapeutic impetus to be gained through therapist allegiance, we would not agree that this should necessarily be counted as error. The one recent and larger study which sought to control for of these some difficulties was the National Institute of Mental Health (NIMH) T r e a ~ e n t epression Collaborative Research Program (Elkin et al., 1989). This study compared imipramine, GBT, interpersonal psychotherapy and a
psychological placebo control therapy in treating depression in three different sites away from u ~ v e r sclinics. i~ This study found overall few significant differences between treatments, but for the relatively more severe patients those receiving medication did best, and patients with placebo only did worst, with the two psychotherapies intermediate. On several measures CBT did only as well as the placebo control and much less well than pharmacotherapy. However, the adequacyof the executionof the CBT at different sites has been inexperience^ questioned in this study. Many of the therapists were relatively and supervision was deliberately less intense. Large differences were seen between sites: at one site, pharmacotherapy outperformed CBT, but at another site the different treatments were equally effective. This suggests that there were differencesin the skill in application of CBT between sites. Thus there are possible methodological problems with the ad~nistrationof CBT in this study. In all, therefore, the relative performance of CBT and pharmacotherapy in treating depressionis still an open question. Where therapists have adequate two treatments will prove training and supervision,it is most likely that the to be of overall comparable efficacy. The more controlled studies, utilizing experienced pharmacologists and monitoring compliance and plasma medication levels, demonstrated little superiority for either CBT or pharmacotherapy in the treatment of depression ( M u ~ h y et al., 1984; Hollon et al., 1g92a). What is clear is that no controlled study of outcome of CBT in depression has had a negative result. Whatever the relative merits of CBT to pharmacotherapy, it is clearly an effective treatment. Perhaps the more important issues to address are the differences in responsiveness within a 'depressed' sample in different settings, the identification of the 'responder' and the developmentof methods to cope with non-responders. When examining the efficacy of CBT, two issues areof importance. First, a distinction needs to be made between CBT applied alone and~ B applied T in combination with other treatments, principally pharmacotherapy.when considering the efficacyof a treatmentin a researchstudy, comparison with another validated treatment may be most helpful, but in clinical practice bowing whether the addition of another treatment will produce further symptomatic improvement may be of more benefit. Therefore this chapter will draw attention to studies of CBT alone and CBT in combination with other treatments. Second, CBT is constantly evolving and new modifications are made for specific difficulties. When considering the efficacy of CBT in depression, a distinction is necessary between the original approach of et al., 1979) and more individual sessions in the Beck format (Beck specialist approaches. Some recent developmentswill be briefly mentioned at the endof this chapter.
toat least as effective In primary care settings T has been demonstrated be as other interventions ( kburn et al., ~~$31;Scott Freeman, 199a), and a ormal general practice ca perior to normal general ess makes up around 10%of consecutive cons~ltations Glare, 1987) and the more chance thereofisa chronict r e a ~ e n t - r e f r a c t depressive o~ disorder ing ( ~ c o tet t al., 1992), the increased a~ailabilityo may be a developmentto be actively considered. f the earliest investigations of CBT in the lackburn et al. (1981) compared G T and antidepressant ~ e d i c a t i o tients with major depressive disorder in psychiatric out-pa~entand ents recei~ingGBT in pressants. There were however. It has been suggested that the administration of medican ary care was inade~uate.Furthermore, patients were
ressive disorder. At completion of t r e a ~ e n tp, ntly less depressed than th tom severity made by PS ssion.Three out of 1
atientsgiventhecom14 improving substangroup, which included titioner (GP), contained only four epressed status, with five out nd the remain in^ eight ~atientsendi tate. Fennel1 and Teasdale (1987) ide ho produced a 50% im sessions only. These findings indicate both rences in response to ~ e a ~ eand n t the the combined approach. They reviated t r e a ~ e n t for s some
nts in primary care. oss andScott (1985) randomly all cticehealthcentreininnercity L 1 2 sessiQns)or a treat
ressedpatientsinageneral
,to indivi~ualGB
C h a ~ t8: e ~Cognitive-behavioural ~ did significantly better than the treatment-as-usual group. nificant differencesb e ~ e e the n gr ments were maintained at 12-mon In a more recent study which suggests an improved outcome (at least in dinb burgh) in the treatment of depression by GPs in the community, Scott and Freeman(1992) compared randomized allocation to an abbreviated individual CBT, amitriptyline, social work counselling and routineGP care for 121 major depressive disorder patients in 14 primary care practices in Edinburgh. They found that all t r e a ~ e ngroups t produced marked improvemen in depressive symptomsover 16 wee S, with only small advantages forS cialist treatments over GP care. Howe er, as Jan Scott points out (ScottJ. el ~ 9 9 4there ) ~ are some important questions about this study, ~ c l u d i n g failure to control for chronicity of depression (the CBT group si~ificantly had longer duration of depression), failure to monitor the adequacy of the CBTi n t e ~ e n tion and absenceof follow-up. Such shortcomings limit the conclusions beto drawn from thisstudy. There have been some other successful developments to produce abbreviated coursesof CBT for depression in primary care (Miranda eS Munoz, 1994; Scott C.S. et al., 1994). With the need for cost-effectiveness, short but effective ry care are likelyto become increasingly important. noz (1994) randomly assigned 150 public care medical patients in San Francisco to eitheran %week CBT course or a control condition. The patients fulfilled criteria for minor depression (depressed mood or anhedonia plus one or more depressive symptoms, insufficient to meet major depressive disorder). The CBT group showed reductions in symptoms which persisted over 1 year. They also reported less somatic symptomatolo missed fewer appointments with their ~rimary care providerin the fol year. Scott C S el al. (1994) have piloted a short C T consisting of six weekly by GPs themsessions each lasting 2o'min (a total of 2 h of therapy) carried out selves. With a sample of seven major depressive disorder patients (more severely depresse than in comparable studies) a mean of fall15 points (47%) was seen on the eck Depression Inventory (BDI) and a mean fall of 14.7 points (54%) seen on the amilton Rating Scale for Depression( H R S ~ ) of the subjects showed at least 50% change in depression ratings d treatment. Only oneof the seven patients was on antidepressant medication. Although small and lacking a control group, this study does demonstrate of depressed the potential forC T to make a large impact on the population primary care patients, within the financial and time constraints of such a caution is warranted in the interpretationof these e absenceof follow-up. Resultsof primary carestu summarized in Table 8.x.
h
~
Table 8.1 Studies of CBT in primary care. group
Study
Controls Treatment start n
HRSD at
. ~
Blackburn et al. (1981) 17
24
12-20 sessions
Tricyclic
Teasdale et al. (1984)19
34
12-20 sessions
TAU
Ross & Scott (1985)
51
12sessions individual or group
TAU
-
Scott & Freeman (1992) 18 121
-
Miranda & Munoz (1994) 150 Scott et al. (1994)
7
27
sessions Eight
Tricyclic SW counselling GP care
sessions Eight x2 h
WL
Six x 20 min sessions
Uncontrolled
HRSD, Hamilton Rating Scale for Depession; TAU, treatment as usual; SW, social worker; WL, waiting List.
Another development of interest is the evaluation of minimal contact bibliotherapy in the treatmentof depressed individuals in the community. A great deal of self-help written material is currently available, much of it espousing a cogni~ve-behaviouralapproach. One book about depression, Feeling Good by David D. Bums (1980) is commonly used as an adjunct to individual therapy and patients often find it useful. (A revised version is o o ~1995). , Jamison and now available called Tke Feeling Good ~ ~ n d ~Burns, Scogin (1995) were thefirst to conducta controlled trialof the effectivenessof the originalbook, as a stand-alone treatment witha group of 80 community volunteers diagnosed from the HRSD as suffering from major depression. The experimental group showeda statistically and clinically highly significant improvementin symptoms relative toa waiting list control group, and these changes were maintained at 3‘rnonths’ follow-up. At the end of the ’treatment’ period, 70% of the experimental group (in comparison with3% of the control group) no longer reached criteria for major depression. These findings, if replicated, could suggest a straightforward solution to the problem of scale in reaching the large numbers of depressed individuals presently untreatedin the community.
The majority of controlled outcome studies already quoted in this chapter, testify to the effectiveness of CBT for depressionin out-patient settings. CBT
Table 8.2 Studies of CBT in secondary care (out-patients). .~
Study group
Controls
nstart
Rush et al. (1977) 22
41
Backburn et al. (1981)
40
HRSD at Tricyclic
19
20 Murphy70 el. al. (1984)
Tricyclic Combination Tricyclic Combination Placebo
Beck et al. (1985)
33
21
Combination
Elkin et al. (1989)
250
19
Tricyclic IPT Placebo + CM
Hollon et al. (1992a, b)
24
Tricyclic Combination
64
Jacobson et 150 al. (1996)
18
BA AT
H E D , Hamilton Rating Scale for Depression; Combination, tricyclic -I- CT; IPT, interpersonal psychotherapy; CM, clinical management; BA, behavioural activation; AT, BA + work on automatic thoughts.
for out-patientshas been shown tobe effective in a range of countries: theUS (e.g. Rush et al., 1977); UK (e.g. Blackburnet al., 1981); Australia (e.g. Wilson et al., 1983); and Canada (Jacobson et al., 1996). Studiesof outcome for depressed patients in secondary care out-patientsettings are summarizedin Table 8.2. Later in this chapter further data will be discussed bearing upon extent of improvement and other questions of interest about this group of patients.
Controlled studies of CBT in in-patient settings are few as yet but again there are suggestions for the effectiveness of CBT. Patients on psychiatric wards are morelikely to have greater severity and chronicityof depression and more associatedpsychopat~ology.Examination of CBT with in-patients therefore also provides information about the ofuse CBT with more severely depressed patients. The studies of CBT with this populationdo not compare CBT alone with other treatments but are interested in whether CBT added to medication can significantly improve treatment outcome. The results are generally positive; the addition of CBT to standard in-patient treatment,
Although the greatest scientificattention has been given to major depression, CBT has been shown to be effective for ofadepressive range conditions: syrnptomatic community volunteers (Wierzbicki & Bartlett, 1987); adolescents (Reynolds & Coats, 1986);elderly populations (Gallagher & Thompson, 1982); minor depression (Miranda& Munoz, 1994);primary major depressive disorder (Blackburn et al., 1981; Kovacs et al., 1981; Elkin et al., 1989); residual symptoms of depression (Fava et al., 1994) and chronic drug-refractory depressed patients (Scott, 1992). The usefulness of CBT with bipolar depression and psychotic depressionis, as yet, little explored.Recent studies of CBT as an adjunctive treatmentwith medication in schizophrenic populations (e.g. Garety et al., 1994)would suggestthat these areas could be fruitfully investigated. Although CBT may not act on primary causes and symptoms of these disorders, it may be usefuldeal to with secondary problems. Attempts to determine whether CBT is more or less effective for parany ticular typesof depressed patient have not been particularly successful. Approaches tothe question of who will respond to CBT can be broadly split into two factors: ~ r o g ~ o sfactors-what t~c factors predict who will do best with CBT?-and ~ r e s c r ~factors-which ~t~~e treatment will best help this individual? The evidence gathered has tended to generate some prognostic factors but few prescriptive factors. The factors which have been considered in response to treatment range from sociodemographic variables (age, sex, marital status), to diagnostic and course of depression variables (endogenous, situational, severity, chronicity), to function and personality variables (IQ, personality disorder,attitudes to treatmentand self).
No sociodemographic factors have been demonstrated to favourthe choice of one therapy over another in a range of studies (McLean & Hakstian, 1979; Jarrett et al., 1991; Sotsky et al., 1991). Attempts to demonstrate that different typesof depression would respond better to CBT or pharmaco~erapyhave also been inconclusive. Beck’s copitive model for depression would seem to be most applicable to depressions that are reactive to life events. Gallagher and Thompson’s (1982)results which suggested that more non-endogenously depressed elderly patients (80~*)responded to CBT than endogenous depressed patients (33%) are consistent with this view. However, both Blackburn et al. (1981)and Kovacs et al. (1981) found that an endogenous/non-endogenousdistinction did not separate responders from non-respondersto CBT alone. Sotskyet al. (1991) found that endogenous depression predicted more improvement with all interventions
T, interpersonal therapy, imipramine and placebo control). Again, endogeneity did not distinguish CBT from the other ~eatments.Similarly, ressed patients with higher pretreatment stress levels do not respondbetter to CBT than toi m i p r a ~ n (Garvey e et al., 1994). erefore, overall, factors which would suggest a depression of a situational/reactivenature, do not favour a differential responsiveness to CBT. It may be that C ective of original causation ~ k r o its ~ ~i k~ ~~~0~ ~ cthet ~ i c i o ~ s cycles ~ k i c ~ ~ai~tai~ ~e~ressio~. Severity of depression has been considered as a possible prescriptive factor for different therapies, with pharmacotherapy recommended for more severe patients and CBT recommended for milder depression (Paykel, 1994). This suggestion carries some face value in that it may wellbe argued that for T to be effective, patients require some flexibility of thinking and mood and some willingness to attempt new behavioural tasks. In the most severe depressions, the rigidityof negative thinking and~ t r a c t a b i lof i ~depressed mood, coupled with fatigue and inertia, are likely to act against a ~ i ltoi ~ collaborate in CBT. In such cases, pharmacotherapy may act to potentiat condition whereGBT can be more effective. The differential response to and pharmacotherapy for severe depression is supported byE l h et al.’s (1989) finding that GBT was inferior to imipramine for more severely depressed patients. This, however, is not conclusive evidence when the methodolo pro~lems mentio~ed earlier are considered. Thaseet al. (x991a) have sp cally examined the effectsof CBT in an open controlled study with more of >20) and less severely depressed patients. severely (Hamilton Rating Scale Poorer response to CBT was seen in the more severe patients but both grou experienced similar, robust and clinically s i ~ i f i c a nreductions t in depressive symptoms leading the authors to the conclusion that treatment needs to be maintained over a longer duration for those patients who ~eatment start at a more severelevel of depression, rather than to the alternative view that CBT is ineffective. Thase et al. (199xb) report successful results with a modified intensive CBT for 16 endogenously and severely depressed in-patients treated without medication. Treatment consisted of a maximumof 20 sessions given on a daily frequency over 4 weeks.~ i r t e e patients n were characterized as responders making highly significant changes on measures of depression. The authors comment that relapse rate was high in individuals who did not receive continuation CBT therapy after discharge fron hospital. More resear is clearly needed to shed further light upon this question. It would seem that where therapists are well trained and supervised to maintain mor sistence with a severe group, highly clinically significant cha achieved. Thase and colleagues’(q91b) study is oneof the very few thathave utilized CBT without medication in a severely depressed group More frequently
a combined treatment is used, particularly in in-patient settings. This combination has been shownbeto effective, although questions are still unanswered as to whether a combined treatment is more effective than a single treatment. patients Miller et al. (1989) found that a combined treatment for depressed showed an advantage 4 months after discharge and out-patient treatment. Scott (1992) showed that combined CBT and pharmacotherapy with a chronic depressed group produced better improvement (42% change) in comparison to a study utilizing CBT alone with a similar group in primary care (Feme11 Teasdale, 1982;24% change). Milleret al. (1990) demonstrated that depressed in-patients with high pretreatment ~ y s f ~ c t i o nAttitude al Scores responded poorly to p h a ~ a c o ~ e r a p alone, y but responded much better to pharmacotherapy plusCBT. However, Beck et al. (1985) in out-patient samples found no difference between the effectiveness of CBT alone and CBT in combination with pharmacotherap~Murphy et al. (1984) and Hollon et al. (1992a) found no significant differences between CBT alone, pharmacotherapy alone and the combined treatment but non-significant differences did favour the combined treatment. This no^-statistically significant differenceis, however, of clinical significance. Perhaps with larger samples, there would be sufficient statistical power to demonstrate a statistical difference. Again, the setting and the severity or type of depression may explain the differences between these studies. In clinical practice, though, combined treatment probably has its benefits, particularly amongst the more severe patients, and there are no antagonistic effects.
Attempts to determine prognostic factors for CBT have been more successful, although manyof these factors are probably not specificCBT. to Most demographic factors do not predict response to CBT. Thase et al. (1994) dernonstrated that depressed men and women tended to have comparable responses to CBT. One factor that has been demonstratedas predictive in several studies is marital status. It appears that having a partner improves a patient’s response to CBT (Jarrett et al., 1991; Sotskyet al., 1991). The most prognostic factor that has been found is the severity of pretreatment depression (Blackburn et al., 1981; Elkin et al., 1989; Jarrett et al., 1991; Sotskyet al., 1991). The more depressed the patient, the less likely they are to respond to CST, although, ashas been suggested above, this may be an artefact of the chosen length or intensity of treatment. The effect of initial severity does not appear to be specific to CBT the more severe the depression, the less effective pharmacotherapy will also be. Hollon (as quoted in of Global Chronicity Fennel1 & Teasdale, 1982) developed a seven-item scale
Chapte~8: Cognitive-behavioural ~ h e ~ ~ p ~ and Severity which predicted 39% of outcome variance for depressed subjects treated either with medication or CBT in the Rush et aE, (1977) study. Scores of 4 or more predicted poor treatment outcome. The Global Chronicity/§everity Index comprises these seven items: x Beck Depression Inventory(BDI) 230; 4 months; duration of current episode is greater than inadequate response to previous treatment; previous episodes,at least two; associated psycho pa tho lo^; overall impairment is Moderate or Severe (rated by clinician); 7 poor estimated tolerance for life stress. This scale is clearly multifactor~~l and there is potential for further investigation of the weightof these different factors in predicting outcome in different samples. Some functional and personality factors are also important indicators prognosis forCBT. While many suppose that intelligence is required for C studies that have examined this variable have not found it to be predictive, although levelof education has been associated with drop-out (Blackburn et al., 1981). Much may depend upon the flexiblityof the therapist in devising techniques for those withless education. Considerable interest has emerged for the useof CBT in mildlylearning-disab1ed.patients(Williams &.Moorey, 1989). Whilst personality(as measured by scores on the Eysenck Personality &.Hakstian, 1979; estionnaire) did not predict outcome with CBT (McLean been reported as being a negavacs et al., 1981), personality disorders have tive indicator forstraightfo~ardCBT for depression (Tyreret al., 1993) and this viewis commonly promoted.This finding is probably related to diffithe culties in engaging such patients in therapy, the more extreme attitudes and behaviours they tend to display, the limited flexibility they show and their greater difficulty in reporting thoughts and feelings. However, et al.Shea (qgo), reporting the interactionof personality disorders with treatment for depresstudy, found that personality-disordered patients did as tter than those without personality disorders in the CBT group only. The specificityof this negative finding in the Shea study is interesting becauseit could be that CBT’s emphasis upon collaboration might be more acceptable for some patients with personality problems. concept of personality disorder clearly encompasses a dimension of severity; it may be that the most extreme cases are difficult to engage in any therapy. However, one might also argue that in such instances thedef~itionof satisof dysphoria norfactory outcome might need to take into account the degree mally experiencedby an individual with ahighly dysfunctional personality outside of a depressed episode. Various studies indicate the importance of patient’s attitudes prior to
treatrnent in predicting outcome. The Dysfunctional Attitude Scale (DAS), a 40-item questionnaire which measures u ~ e l p f uattitudes l c o n c e r ~ nper~ fectionism and a need for approval (~eissman, 1979)~ has been particularly implicated in predicting response to G T. The stronger the u ~ e l p f uattil tudes endorsedby patients, the less res nsive they are to CBT (Keller, 1983; Simons et al., 1986; Jarrettet al., 1991; Sotskyet al., 19 This cognitive measure does not, however, simply predict response to T; Sotsky et al. (1991) also found that low DAS scores predicted better responses to imipramine and placebo-control.It did appear, however, thatin addition to this general predictive effect, this measure was a more powerful predictor for GBT. the DAS score changedboth DeRabeis et al. (1990) demonstrated that whilst and pharmacotherapy, only in CBT did mid-trea~entchange on redict subse~uentchange in depressive symptoms. This finding is one of the very few todate that demonstratea specifically cognitive mediation for change in CBT. Another predictor for response to CBT is initial endorsement of the cognitive model and any subsequent expectation of improvement this may produce. Sotsky et al. (1991) found that high expectations of improvement predicted better outcome, whatever thet r e a ~ e n tMore . specifically,Fennel1 and Teasdale(1987) found that patients who showed a positive response toa s s a~ much ~ ~ , booklet o u t l i n ~ gthe CBT approach,~~~~~git^ ~ ~ ~ e showed greater improvement over the next 12 sessions of therapy. Thus the extent to which the CBT model is congruent with the individual’s own beliefs or the extent to which the individual is open to look at things in another way ificant. These hypotheses suggest fruitful avenues for further research. In conclusion, thereis little evidence for prescriptive discriminationsbetween GBT and other treatments for most depressed patients. Endogeneity/ severi~/situationalfactors do not predict which treatment will be more effective for patients. Psychotic depression and bipolar illness currently strongly indicate pharmacotherapy rather than CBT. Less severe depression, less dys~nctionalattitudes, greater expectationsof improvement and being married/having a partner areall factors which are related ato good outcome in CBT. These factors may partially explain why some persona1i~-disordered patients who tend to lack these factors, respond poorly to straightfo~ard GBT approaches.oreresearch into theidentification of characteristics of patients who are likely to respond to GBT is clearly needed. Current best practice relies upon a frank discussion of different approaches with the patient to arriveat a choice acceptable to the individual. h addition, therapists commonly take a ’let’s try it and see’ approach, offering a trial of five orsix sessions beforea review todecide on whether or not to takeit further.
The m a j o r i ~ of controlled outcome studieshave examined t r e a ~ e neffects t (BDI), a selfusing two measures. Thefirst is the Beck Depression Inventory ort, ;21-item questio~aire listing depressive symptomatology (Beck et al., 1961). The second is the Scale g for Depression ( H R S ~Ham; ilton, 1960)~ clinician-r a outcome then is reported in terms post-trea~enton avera p. Studies have demo 48 and 80% change in e in per~entagechange following CBT; b een milton Ratings (Rush et a1. , scores and between 64 and 77% c h a n ~ e phy et al., 1984; Becket al., 1985; Elkinet al., average scores change from around 30 o post-~eatmenton the BD1 and from between 20 und 7 post-treatment on the HRDS. Such a change scores is equi~alentto a change from severe-moderate depression to mild epression-euthymic. Thissize of therapeutic impactis sinxilar to that found der, 1996; and see Chapter11). t studies have also reported a ercentage treated to full r e ~ s s i o nas , S, and these range from 63 to 82% in the has been best executed, falling to51% in the Elkin et al. (1989) study, where, as has been observed, co~petencymay have been lower and variable between sites. Whilst we have expressedenth~siasmabout this approach, thereis clearly no room forcomplace~~y. The treatment effect sizes (0.85-0.96 reported in a meta-analysis carried out by less than forco~tive-behavioural treatobinson et al., 1990) are, on average, ment of anxiety andto obtain thebest results it would seem that considerable investment in training and supervision is required. Many have also reported that a small minority of patients do not respond all at or become worse with adly reflect uponthe outcome of their hugely expensive, ~ u l ~ c e ~ t r e studythat 16 weeks of thetherapiesstudied was insufficient for most patients to reach full recovery and lasting re~ission. That being said, it is clear that CBT has a treatment effectthat can produce r e ~ i s s i o in n at least one halfof depressed out-patient§ and § i ~ i f i c a nclinical t rovements in the majority of such patients. *
itive therapy has been found tobe an effective short-term treatment for ressed in-patie~tsand out-patients, producing effective results in
acute episodeof depression (Rushet a1. ,1977; lackburn et al., 1981; et al., 1984; Elkin et al., 1989; ollon et al., 1992a). The time-course of term change, where it has be studied, seems to be, on aver the time-course for p~armacotherapywith tricyclic drugs, a1 been observed, there is also a
S
As more studies have shown a significant short-term treatment effect defor pression, interesthas shifted to maintenance of t r e a ~ e neffects. t Ithas been ized that depression frequentlyreturns in affected individuals so the ay benot so much short-term improvement but long-term model for CBT would suggest that underlying attitudes patients learn skills that may help them deal to with future problems. There is some limited evidence that this hope may be justi A series of studies have followed up the controlled t r e a ~ e nSt 2years, using scores on the BD1 as the cr
e studies have methintenance medication group in the pharmacotherapy group and small samples. ~urthermore,the studies differ in assessment methods, timingof assessments and even inclusion criteria for follow-up groups. None the less, all thesestudies show that the CBT groups do better than other t r e a ~ e n t safter 1: or 2years followin termination of ther effect beyondthe e lapse. The Evans et al. (1992 recent and best d~signedin this series, included a group wh edication for 1year of the study beyond terminationof active therapy in the CBT groups. Using a definition of t , the relapse as a BD1 score of 16 or greaterand/or return to ~ e a ~ e nwithin 50%of 2-year follow-up period,21% of the C TT group relapsed, compared to a pharmacotherapy group without ntenance medication (i.e. active treatment halted at the point when CBT ended) and 32% of a pharma~otherapy roup with maintenance medication forthe first year. This mirrors many of orted results, with short-term G T producing about half the probability of relapse with short-term pharmacotherapy and approximately equal effects with long-term maintenanceof medication. These results indicate that a short-term therapy can produce some lasting change, comparable to continued pharmacotherapy, which persists in the absenceof the therapy. This is
consistent with the suggestion in the cognitive model of depression that C may alter underlying assumptionsthat produce vulnerability to depression. It is also consistent with the ideathat CBT teaches patients coping skills to handle future difficulties. Again, however, a cautionary note must be a in that in the above Evans study the percentages quoted refer only to those patients who were successfully treated and therefore entered the follow-up stage. If one takes the non-respondersinto account, the l o n g - t e ~ outcome would showthat many more patients from the sample originally enter treatment study relapseor remain depressed. on remains whether the reduced return of symptoms in e returnof se or reduced recurrenceof depression. ~ e l a ~iss the ociated with the current treated episode and~ e c ~ ~is ~thee ~ c e onset of a new episodeof depression (Klerman, 1978).An untreated moderate depression usually has a natural course of 6-9 months and therefore a lengthof time makes a suitable cut-off between relapse and recurrence. ing antidepressant medication within this initial 6-9 months, even follow rapid remission, carries a higher risk of a returnof symptoms than termin ing medication beyond this period. This is consistent with the concept of an rocess of depression, which medication suppresses rather than recause mostof the returnof symptoms in the studies quoted occurs in thefirst G'months after discharge,it has been argued that these studies have been assessing relapse rather than recurrence (Hollon et al., 1991,1993). The ~ ~ for e CBT ~ e for ~ tdepression. ~ u ~ results are indicativeof a ~ e l ~ ~ s e function However, Hollonet al.(1991, 1993) do raise one methodological problem with these results; because the follow-up studies have assessed patient completed therapy successfully, a filtering effect may occur, wi pharmacothe~apygroup successfully treating more difficult patients a subsequentl~showing less effectiveness in preventing relapse. This is one possible explanation of the results and untilit is tested, it cannot be for sure thatCBT has a differential relapse-prevention effect. None th is clear thatCBT does have a good record in reducing symptom retu 2 years after discharge. Ludgate (1991), in an uncontrolled retrospective study quoted in Ludgate (1994)~ followed up80 unipolar depressedpatients, 3 and 5 years after CBT. Using the criteria for affective disorderas a measure for recurrence,52% of the sample had a recurrence of depression within the 5 years. With a criterion of a 'return of symptoms of equal or greater severity to the indexepisode', 32% of the sample had a recurrence of depression over5 years. Thissu that the improved effects for CBT shown for preventing relapse are maintained over longer periods of time, indicating that CBT may also protect against recurrence. Thus during a long time period following discharge, patients undergoing CBT showed a return of symptoms comparableto those c o n ~ u i n g
~ h a ~8: tCognitive-behavioural e ~ medication for the y-year period. However, Ludgate’s patients attended the Centre for Cognitive Therapy~hiladelphia in and are a highly selected group, so genera~ationsfrom this sample must be guarded. Furthermore, thereis still much room for improvement because between one-third and of a patients half years. successfully treated with CBT, again suffered from depression 5within Table 8.3 shows results of follow-up studies of controlled clinical trials of CBT for depression. le 8.3 Follow-up studiesof CBT for depression. Study group Kovacs et al. (1981)
Months Tricyclic 12 82% vs.
Outcome YOrelapsed CT 56% (NS)
Blackburn et al. (1986)
12
Tricyclic 78% vs. CT 23% vs. Combination 12%
Simons et al. (1986)
12
Combination Tricyclic 67% vs. 18% (S)
Evans et al. (1992)
24
Tricyclic discountinued combination 15% (S) Tricyclic continued32%
Shea et al. (1992)
Tricyclic 18 50% 36% CT IPT 33% vs.
(S)
43% (S) vs. CT
50% CT vs.
+ placebo
21% ( S ) vs.
(NS)
P T , interpersonal psychotherapy.
Whilst manyof the key ideasof CBT (e.g. collaborative approaches, empirical testing, formulation of hypotheses, activity structuring and challengin thoughts) are relatively simple and likely to improve the general approa any mental health care professional to their patients, the effective practiceof CBT is much more complex. Effective use of CBT requires considerable training, practice and supervi~ion.The selection and monitoringof therapists to guarantee high levels of skill and experience is a core part of many of the t r e a ~ e nstudies t quoted.With less experienced therapists,less effective results wouldbe expected (cf. Elkin et al., 1989). Beck himself developed a traincourse for postqualification psychologists that runs full-time over a year. course in the UK offers such intensive training, most courses b release, professional development courses. Little is known about t tiveness of this relativelydilute dose of t r a i n ~ gfor the different professional groups that are trained. Economic pressures preclude the develo time-intensive courses at present, although basic training cou ning to corporate more of the principles of the GBT approach as the research evidence rows. Such courses that exist are scarce, with demand far
he^^^^
Gha~teu8: Gognitive-behaviouual t h e u a ~ ~ exceeding supply. In consequence, qualified therapists are few and far between, although the provisionof a register in the UK will at least allow monitoring and access to those who are trained. At present, access is relatively more available in some geographical areas of the NHS than others. This has implications for waiting-list time and ease of referral. With a longer waiting list, some depressed patients will have significantly improved before reaching treatment, others will have si~ificantlyworsened. In a time- and therapist-intensive therapy, such problems are clearly greater than in pharmacotherapy-based approaches. ~echanismsfor prioriti~ingpatients may be requiredunder these circumstances as wellas funding being made available fortraining. However, evenin an ideal economy, the prevalence of depression would suggest that CBT is never going to be a treatment for all. It would seem that many patients can benefit from bibliotherapy, abbreviated courses in primary care, group approaches, etc., allowing the more expert staff to devote time to those who are unable to use these less supported methods. ~ h i l s the t availabili~of training is scarce, therapists and counsellors abound who espouse the practiceof CBT. Thepractice of caveat e ~ ~ tshould or be borne in mind by purchasers and referrers, who are becoming more conscious of quality indices. Exposure toan inept therapist will do nothing for the optimism of a depressed patient and will add to the problems of future therapists for this patient. The managing clinician may also be faced with the problemof deciding whether the patient has received a proper trial. Detailed enquiry into the nature of the workundertaken in previous episodes of therapy may shed some lightupon this question.
The cost-effectivenessof CBT relative to other treatments is a highly relevant question, with an increasein market forces becoming apparent in the NHS. If CBT is of comparable effectiveness to pharmacotherapy, but requires the extra costsof training and paying for a number of sessions betweentherapist and patient, should it beused? A number of issues arepertinent to this question. First, the exact cost of CBT relative to other ~ e a ~ e nist unclear. s Ross and Scott (1985) concluded that CBT could be applied cost-effectively in general practice, whilst Scott and Freeman (1992) concluded that the additionakosts of any specialist treatments (at least twice as much as routine GP care) were not commensurate with their clinical superiority over routine GP care. However, in this latter study, the time difference between administering a course ofCBT and admi~steringa course of medication was only 3.6h. Second, aspatient choice in therapy has increased, a preference towards psychological therapies and
Chapter 8: C ~ ~ ~ i t i ~ e - ~ e htherapy avi~~ral away from pharmacotherapy has become evident. Indeed, patients’ dislike of medication maybe a major factor in attrition rates in drug studies. However, drop-out rates for CBT in the major outcome studies vary between 5 and 38% (Hollon et al., 1992a), suggesting that CBT also has a considerable attrition factor to address. Third,it is clear that CBT can help some patients who are not responsive or who are unable to take medication for some reason. Fourth, CBT appearsto have additional effects beyond treating the acute episode of depression, allof which have cost-benefit implications. Reducing somatization and failures to attendin a primary care setting arelikely greatly to improve the general efficiencyof a GP surgery. Most importantly, the relapse prevention effects suggested forCBT would reduce returnof depression compared to pharmacotherapy. This would reduce the costs involved in repeating t r e a ~ e n t and s reduce the financial effects of depression, for example, loss of man-hours and efficiency at work. Furthermore, successfulCBT appears to remove the need for the expenses of maintaining antidepressant medication for many years. A major factor in treatment cost is treatment duration.A major advantage of CBT is that it has a brief but flexible length. Substantial treatment effects can be seen within four sessions and often therapists do not exceed 20 sessions of treatment. However, with more difficult and intractable problems, longer courses of therapy can be utilized if required. Returning to the question of who responds well to CBT, perhaps the most useful practical predictor, is a patient’s response to the initial sessions. Patientswho respond to initial sessions are highly likely to respond much more. Using a timelimited contract can greatly enhance the efficiency of treatment and improve cost-effectiveness.
When considering CBT and its place in treating depression, the standard CBT approach as outlinedby Beck (Beck et al., 1979) has been the main focus of this chapter because it is the method that has received the most scientific attention. However,a number of developments in CBT, relevant to the treatment of depression, are evolving. Oneset of developments involve shifting CBT beyond just a ’here and now’ approach, to investigate more fully the patterns of thoughts and behaviours that characterize patients. By examining the models that people construct of themselves and the world from childhood onwards, better understanding of chronic problems can be achieved. Such approaches incorporateall the standard CBT repertoire but add more consideration of imagery, experiential techniques and interpersonal factors, drawing upon techniques from Gestalt and psychodynamic therapies. Others are developing the co~tive-behaviouralmodel itself and increasing
G h ~ ~ t8:e rGognitive-behavioural t h e r ~ ~ ~
our understandingof the conceptualbasis of CBT (Teasdale, 1993; Teasdale Barnard, 1993). Such approaches may be particularly useful for treating both chronically depressed and dysthymic patients and for mana disorders associated with depression and chronic dysphoria (Beck et al., 1990; Young, 1990; Linehan, 1993). Another set of developments concern the relapse prevention function of CBT. Fava et al. (1994) have demonstrated the value of CBT in reducing residual symptoms, following successful treatment with medication. Reducing these residual symptomsby CBT produced a lower rateof relapse at 2-year follow-up (15% vs. 3570, although not statistically significant). Hollon et al. ( q p b ) are investigating the useof CBT for euthymic populations to reduce the riskof future depression. Teasdale et al. (1995) have proposed an a~entional retraining programme to prevent relapse in remitted depressed patients. These developments will improve our ability to focusinte~entionon the problems most intractablein the treatmentof depression. A s these new approaches are refined and tested in controlled treatment studies, the efficacy of CBT for dealing with a wider range of aspects of depression will become clearer.
This chapter has focused upon researchin relation to the efficacyof CBT in the treatment of depression. We have argued for the general efficacyof the approach with a wide and expanding range of depressed patients and have pointed to some problems in its training and supervision implications with some suggested answers in the way expertise may become more focused upon resistant problems as knowledge grows about the treatmentof depression. Much of the researchto date has comparedCBT with pharmacotherap~the standard treatment approach. Ratherlittle is known about the relative efficacy of CBT and other psychological approaches, although procedural similarities between differing therapies are becoming more evident and some h suggested that CBT may becomean integrative therapeutic model which will come to embrace more and more techniques developed from differing theoretical perspectives. It will have become clearthat CBT is more than astatic series of operations but a whole scientific approach closely tied in to the developing psychology of human emotion; this relationship is leading to fruitful developments in method and process. Perhaps isthis the main secret of its continuing success beyond the first decade of its youthful enthusiasm and of the extent to which it continues to dominate the field of applied clinical psychology andto attract the interestof the mental health professions. Cognitive therapy based upon Beck’s theory of depression is now a standard treatment of depressive illness of mild to moderate severity presenting in primary and secondary care. In these patients and settings, CBT and
C h a ~ t8: e ~Cognitive-behaviouual t ~ e ~ a antidepressant medication may be of equivalent therapeutic efficacy. CBT may (unlike antidepressant medication) continue to have beneficial effects, even after treatment is withdrawn. Despite their apparent therapeutic equivalence, two treatment modalities varies and more research patient preference for the is needed into the indications for each form of treatment and the possible interactions between them. Much more research is needed upon the role of CBT in the treatment of severe depression and manic depressive illness for which CBT cannot yetbe recommended. All doctors treating rnildly and moderately depressed patients should have access to a CBT service because for some patients will thisbe the most effective and acceptable treatment.
Beck, A.T., Ward, C.H., Mendelson, M,, Mock, J. & Erbaugh, J. (1961)An inventory for measuring depression. Archives of General Psychiatry, 4, 561-571. Beck,A.T., Rush,A.J., Shaw, B.F. &Emery, G. (1979) Cognitive Therapy of Depression. Guilford Press, New York. Beck, A.T., Hollon, S.D., Young, J.E., Bedrosian, R.C. & Budenz, D. (1985) Treatmentof depression with cognitive therapy and amitriptyline. Archives of General Psychia~ry, 42,142-148. Beck, A.T.,Freeman, A., Pretzer, J. et al. (1990)Cognitive Therapy of Personality Disorders. Guilford Press, NewYork. Blackburn, I.M., Bishop,S., Glen, A.I.M., Whalley, of cogniL.J. & Christie, J.E. (1981) The efficacy tive therapy in depression: a treatment trial using cognitive therapy and pharmacotherapy, each alone and in combination. British Journalof Psychiatry, 1399,181-189. Blackburn, I.M., Eunson, K.M. & Bishop, S. (1986) A two-year naturalistic follow-up of depressed patients treated with cognitive therapy, pharmacotherapy and a combination of both. Journal of A~ectiveDisorders, lo, 67-75. Blacker, C.V.R. & Clare, A.W. (1987) Depressive disorder in primary care. British Journalof Psychiatry, 150,737-751. Bowers, W.A. (1990) Treatment of depressed inpatients: Cognitive therapy plus medication, relaxation plus medication and medication alone. British Journalof Psychiatry, 156,73-78. Burns, D.D. (1980) Feeling Good: The New Mood Therapy. Signet, New York. ~ a n d ~ oPeno~. Burns, D.D. (1995)The Feeling Good guin, New York. Comas-Diaz, L. (1981) Effects ofcognitive and be-
havioural group treatment on the depressive symptomatology of Puerto Rican women. Journal of Consul~ingand Clinical Psychology, 49, 627-632. DeRubeis, R.J., Hollon, S.D., Grove, W.M., Evans, M.D., Garvey, M.J. & Tuason, V.B. (1990) How does cognitive therapy work? Cognitivechange and symptom change in cognitive therapy and pharmacotherapy for depression. ~ournalof Consulting and Clinicalpsycho lo^, 58,862-869. Dobson, K. (1989)A meta-analysis of the efficacy of cognitive therapy for depression. J o ~ ~ nofa l Consulting andC~inicalpsycho lo^, 57,414-419. al.(1989)NIMH Elkin, I., Shea, M.T., Watkins,et J.T. Treatmentof Depression Collaborative Research Program. I: General effectivenessof treatments. Archives of General Psychia~ry,46,971-982. Evans, M.D., Hollon, S.D., DeRubeis, R.J. et al. (1992) Differential relapse following cognitive therapy and pharmacotherapy for depression. Archives of General Psychiatry,49,802-808. Fava, G.A.,Grandi, S., Zielezny,M., Canestrari, R. & Morphy, M.A. (1994) Cognitive behavioural treatment of residual symptoms in primary major depressive disorder. A ~ e r i c a nJournal of Psychiatry, 151,1295-1299. Fennell, M.J.V. & Teasdale, J.D. (1982)Cognitive therapy with chronic,drug-refractorydepressed out-patients: a note of caution. Cognitive Thera~y and Research,6,455-460. Fennell, M.J.V. & Teasdale, J.D. (19817)Cognitive therapy for depression: individual differences and the process of change. Cognitive Therapy and Research, 11,253-271. Gallagher, D.E. & Thompson, L.W. (1982) Treatment of major depressive disorder in older adult outpatients with brief psychotherapies. Psycho-
~ ~ a ~8: ~ t oe gr~ i ~ i v e - ~ e ~ atv ~ i o eu ~r aal ~ ~ therapy: Theory,Research andPractice,19, 482-490. tials in antidepressant pharmacotherapy. Garety, P.A., Kuipers, E.A., Fowler, D. et al. (1994) ~ o u ~ofaClinical l Psyckiutry,J7 (Suppl.2), 39-44. Cognitive behavioural therapy for drug-resist- Linehan, M.M. (1993) Co~itive-Behavioural Treatant psychosis. British ~ournalof Medical Psychoment of Bo r d ~linPersonality e Disorder. Guilford Press, New York. low167 (3111 259-271Garvey, M.J., Hollon, S.D.& DeRubeis, R.J.(1994) Ludgate, J.W. (1994) Cognitive behavioural Do depressed patients with higher pretreattherapy and depressive relapse: justified optiment stress levels respond better to cognitive mism or unwarranted complacency? Be~vioural therapy than imipramine? Journal of Afective and Cognitive Psyckofherupy,2 2 , l - m . Disorders, 32,45-50. McLean, RD. & Hakstian, A.R. (1979) Clinical deHamilton, M.(1960)Arating scale for depression. pression: comparative efficacy of outpatient ~ o u r nof~~l e u r o l o w , ~ ~ r o s uand r g Psychiatry, ery treatments. ~ o u r n aof~Consu~fingand C ~ ~ n ~ c u ~ 12,5242. Psychology,47, 818-836. Hollon, S.D., Shelton, R.C. & Loosen, P.T. (1991) Miller, I.W., Norman, W.H.,Keitner, G.I., Bishop, Cognitive therapy and pharmacotherapy for S.B. & Dow, M.G.(1989) Cognitive-behavioural depression. Journal of Consulting and Clinical treatment of depressed inpatients. Behavior Psycholow/ 59, 88-99. Therapy, 20~25-47. Hollon, S.D., DeRubeis, R.J., Evans, M.D. et al. Miller, LW., Norman, W.H. & Keitner, G.1. (1990) (1992a) Cognitive therapy and pharmacoTreatment responseof high cognitive dysfunctherapy for depression: singly and in combina- tion: depressed inpatients. C o n ~ e ~ p oPsychira~ tion. Archives of General Psychiutry, 49, 774atry, 30~62-71. 781. Miranda, J. & Munoz, R. (1994) Intervention for Hollon, S.D., DeRubeis, R.J. & Seligman, M.E.P. minor depression in primary care patients. Psy(1992b) Cognitive therapy and the prevention chosomaticMedicine, 56,136-142. of depression. Applied and Preventivepsycho lo^/ Murphy, G.E., Simons, A.D., Wetzel, R.D. & Lustman, P.J. (1984) Cognitive therapy and 1, 89-95. Hollon, S.D., Shelton, RC. & Davis, D.D. (1993) pharmacotherapy, singly and together in the Cognitive therapy for depression: conceptual is- treatment of depression. Archives of General Psysues and clinical efficacy. ~ournalof Consulting chiatry, 41,33-41. and C ~ i n i c a ~ P s y c k61,270-275. olo~, Paykel, E.S. (1994) Psychological therapies. Acta Psyckiatrica Scandinavica,89,35-41. Jacobson, N.S., Dobson, K.S., Truax, P.A. et al. (1996) A component analysis of cognitiveReynolds, W.M. & Coats, K.I. (1986) A comparibehavioral treatment for depression.lournal of son ofco~itive-beha~oural therapy and relaxaConsulting and Clinical Psycholow, 64,295-304. tion training for the treatment of depression in adolescents. ourn nu^ of Consu~t~ng Jamison, C. & Scogin, F. (1995) The outcome of and C l ~ n ~ c u ~ cognitive bibliotherapy with depressed adults. Psycholow, 54, 653-660. ~ o u r nofa ~Consu~tingand Clinical Psych~low,63, Robinson, L.A., Berman, J.S. & Neimeyer, R.A. (1990) Psychotherapy for the treatment of de644-650pression: a comprehensive review of controlled Jarrett, R.B., Eaves, G.G., Grannemann, B.D. & outcome research.PsychoZo~culBulleti~, Rush, A.J. (1991) Clinical, cognitive and demo108,30graphic predictors of response to cognitive 49. Ross, M. & Scott, M. (1985) An evaluation of the therapy for depression: a preliminary report. Psychiatry Research, 37,245-260. effectivenessof individual and group cognitive therapy in the treatmentof depressed patients Keller, K.E. (1983) Dys~nctionalattitudes and cognitive therapy for depression. Cognitive in aninner city health centre. ~ournalof the Royal ~ e r a p and y Research, 7, 437-444. College of General Practitioners,35, 239-242. Klerman, G.L. (1978) Long term treatment of Rush, A.J., Beck,A.T., Kovacs, M. & Hollon, S.D. (1977)Comparative efficacy of cognitive therapy affective disorders. In: Psychopharmucolo~; A and pharmacotherapy in the treatment of G e n e r a ~ i o nof Progress (eds M. Lipton, A. depressed outpatients. Cognitive Therapy and Dimascio & K. Killam), pp. 1303-1313, Raven Press, NewYork. Research, 1,17-37. Scott, A.I.F. & FEeman, C.P.L. (1992) Edinburgh Kovacs, M., Rush, A.J., Beck,A.T. & Hollon, S.D. primary care depression study: treatment out(1981) Depressed outpatients treated with cogcome, patient satisfaction, and cost after 16 nitive therapy or pharmacotherapy.Archives of General P s ~ c k i a t38,33-39. ~, weeks. British Medical ~ournal,304, 883-887. Lader, M.H.(1996) Tolerability and safety: essen- Scott, C.S., Scott, J.L., Tacchi, M.J. & Jones, R.H.
186
Teasdale, J.D., Segal,2.& Williams, J.M.G. (1995) (1994) Abbreviated cognitive therapy for depresHow does cognitive therapy prevent depressive sion: a pilot study in primary care.Behavioura~ relapse and why should attentional control and Cognitive Psychotherapy,22/57-64. s) help? An information Scott, J. (1992) Chronic depression: can cognitive ( m ~ d f u ~ e straining processing analysis. Behaviour Research and therapy succeed when other treatments fail? Behavioural Psychotherapy,20~25-36. Therapy, 33, 25-39. Thase, M.E., Sirnons, A.D., Cahalane, J., McGeary, Scott, J., Eccleston, D. & Boys, R. (1992) Can we J. & Harden, T. (1991a) Severity of depression predict the persistence of depression? British and response to cognitive behavior therapy. Journal of Psychiat~/ 161,633437. A~erican Journal of Psychiat~,148,784-789. Scott, J., Moon, C.A.L., Blacker, C.V.R. & Thomas, Thase, M.E., Bowler, K. & Harder, T. (1991b) CogJ.M. (1994) A.I.F. Scott and C.P.L. Freeman’s nitive behavior therapy of endogenous de’Edinburgh Primary Care Depression Study’. pression. Part 2: Preliminary findings in 16 British J o u ~ aofl Psychiat~,164, 410-415. unmedicated inpatients. Behavior Therapy, 22, Shea, M.T., Pilkonis, P.A., Beckham,et E. al. (1990) Personality disorders and treatment outcome in469-477. the NIMH Treatmentof Depression Collabora- Thase, M.E., Reynolds, C.F., Frank, etE.al. (1994) Do depressed men and women respond simitive Research Program. A~erican Journal of Psylarly to cognitive behavioural therapy?Americhi at^/ 149, 711-718. can Journalof Psychiatry, 151, 500-505. Shea, M.T., Elkin, I., Imber, S.D. et al. (1992) Tyrer, P,, Seivewright,N., Ferguson, B., Murphy, Course of depressive symptoms over follow-up: S. &Johnson, A.L. (1993) The Nottingham study findings from the NIMH Treatment of Depresof neurotic disorder: effect of personality status sion Collaborative Research Program. Archives on response to drug treatment, cognitive therapy of General Psychiat~,49 (lo), 782-787. and self-help over two years. British Journal of Simons, A.D., Murphy, G.E., Levine, &J.L. Wetzel, Psychiatry, 162,219-226. R.D. (1986) Cognitive therapy and pharmaWeissman, A. (1979) The Dysfunctional Attitude cotherapy for depression: sustained improveScale: a validation study. Dissertation Abstracts ment over one year. Archives of General PsychiInternational/ 40, 1389-1390. atry, 431 43-48. Sotsky, S.M., Glass, D.R., Shea, M.T. et al. (1991) Wierzbicki, M.& Bartlett, T.S. (1987) The efficacy Patient predictors of response to psychotherapyof group and individual cognitive therapy for mild depression.Cognitive Therapy and Research, and pharmacotherapy: findings in the NIMH Treatment of Depression Collaborative Research 1% 337-342Program. A ~ e r i c a n ~ o u r nofa lPsychiatry, 48, Williams, J.M.G. & Moorey, S. (1989) The wider application of cognitive therapy: the end of 997-1008. the beginning. In: Cognitive Therapy in Clinical Teasdale, J.D. (1993) Emotion and two kinds of meaning: cognitive therapy and applied cogni- Practice: An Illustrative Casebook (eds J. Scott, J. M. G. Williams & A. T. Beck), pp. 227-250. tive science. Behaviour Research and Therapy,31, Routledge, London. 339-354s Teasdale, J.D. & Barnard, P.J. (1993) Afect, Cogni- Wilson, P.H., Goldin, J.C.& Charbonneau-Powis, tion and Change. Lawrence Erlbaum Associates, M. (1983) Comparative efficacyof behavioural and cognitive treatmentsof depression. CogniHove and London. tive Therapy and Research,9, 111--1q. Teasdale, J.D., Fennell, M.J.V., Hibbert, G.A. & Cognitive Therapyfor Personality Amies, P.L. (1984) Cognitive therapy for major Young, J.E. (1990) depressive disorder in primary care. British JourProfesDisorders: A S c h e ~ a ~ o c u sApproach. ed sional Resource Exchange, Sarasota, FL. nal of P s ~ c ~ i a t r44, y , 400-406.
The term depression can apply to a transient mood, a sustained change in affect, a symptom, a syndrome or a psychiatric disorder. It is a common and chronic condition; most patients experiencing an initial episode of depression will succumb to a recurrence. It is symptomatically distressing, patients reporting an ineffable and all-pervading depth of despair and hopelessness far beyond normal experience, Physical, social and personal role functioning is greatly impaired, comparable to or even worse than the handicaps associated et with physical illnesses such as hypertension, diabetes and arthritis (Wells al., 1989). Suicide is often a risk. Major depression is often associated with other disorders suchas anxiety disorders, dysthymia and physical illnesses, and such comorbidity worsens the prognosis (Wells et al., 1992). Even in the milder cases, partial remission not full recovery is the usual outcome (Ormel et al., 1993). Depressive disorders canbe divided into five main conditions, although overlap may occur: major depression, dysthymia, bipolar-disorder depressed, cyclothymia and unspecified depressive disorder. This chapter will concentrate on major depression, although the drug treatment is quite similar for 11(see also bipolar depressives. Resistant depression is dealt with in Chapter Sokolov & Joffe, 1995). The toll exacted by depression necessitates the condition being taken seriously, especiallyat the primary care level whereit is often missed or its significance overlooked. Vigorous treatment must be i~stitutedearly, adequate dosages attained, compliance encouraged, and the course of treatment prolonged rather than curtailed.An increasingly wide range of antidepressant medication is becoming available although, as will be seen later, improvements relate to side-effect profile, tolerability and safety; efficacy remains on limited, and effectivenessin practice unchanged. The useof medication has always to be set into the context of wider management, with sympathetic counselling, ~mprovementof psychosocial circumstances, and more formal psychological treatments where appropriate and available.
In the late
I ~ ~ O imi S , odibenzyl derivatives were investigated
as antihistre-evaluated a few years later as found tobe disappointing t to have genuine antidepressant(as opposed to stimua r ~ e t e r of s efficacy were explored and the side-effect merous other similar tricyclic compou~dswere synntroduced over the years including amitriptyline, othiepin, ~lomipramineand lofepramine. Variants, the so-called ‘atypical’ appeared, witha range of dif~erent p~armacological properfor more selective compoundsled first to CO adrenergicm e c h ~ s m sfor , example mapro more selective for serotonin such as zimeldine (later withdrawn). A series have appeared, although of selective serotonin re-uptake ~ i b i t o r (SSRIs) s they vary in their selectivity. They have a different and somewhat bettertolerated side-effect rofile and are currently the focus of a vi concern~gtheir comparative risk/benefit ratios vis-&vis the TGAs and an controversy conce~ing relative cost-effectiveness (Edwards, oup of antidepressants is the monoamine oxidase inroniazid, was developed as an antituberculosis make patients euphoric. Antidepressant propbut the side-effects and hazardous drug and dietary ammelled the use of these drugs. More recently, selec0 have been developed and one, of imipramine and iproniazid, In the 35 or so years since the introduction much has been learned about theclinical use of antidepressant medication. e practical ~uestionsremain, relating to topics such as compliance and tility of plasma on cent ration monitoring, But the paramount puzzle en answered, namely, the mode of action of these drugs.A plethora ort- and long-term biochemical effects have been catalogueditbut r e m a ~ as complete mystery how these effects are translated into a clinical response. lead toa worthwhileimpro~ement in about70% of unselected depressive disorder. However, many of these patients would aneously or with help, support and counselling (Brown this, as has become apparent over the years, careful nt canm a x i ~ i z edr~g-placebo-non-drugtreatment differences. The use erful drugs like the antidepressants to treat minor fluctuations in mood hological problemsin general is tobe deprecated.
In practice, these mainly revolve around the usefulnessof mon.itoring plasma concentrations of the various antidepressants. This is still a controversial issue.,By and large, the relationship between daily doseand levels of drug and its major metabolites is atenuous one becauseof wide variations inthe metabolizing capabilities of individual patients (e.g. for clomipramine, ia et al., 1991). Low plasma concentrations are associated with poor response and high ones with adverse effects. What remains unclear is whether high concentrations are associatedwith a poor clinical response,in other words,whether a therapeutic 'window' can be discerned. Even where such a relationship across patients has been found, it is usually too weak statistically to influence a clinical decision. Estimation of the plasma concentration of an antidepressant in an individual patient remains a help in evaluating a side-effect or excluding total non-compliance, rather than in facilitating clinical response. ~ h a r ~ a c o ~ data e t i may c alsohelp obviate dangerous drag interactions. any antidepressants and other drugs are metabolized via common cytochrome P450 pathways so that potential interactions can be predicted (Taylor & Lader, 1996). This is particularlyimportant with some of the SSRIs, such as fluoxetineand paroxetine. The pharmacokineticsof the SSRIs has been studied in some detail but more with respect to potential drug interactions rather than the correlation. between clinical efficacy and plasma concentrations(van Harten, 1993). Data so far suggest that plasma concentrations bear little relationshipto clinical response for anyof the SSRIs. S
Brown and Khan (1994) recently averred, 'Although antidepressants are clearly effective, when used in. common practice their effectiveness is not a s t o ~ d i n g '(p. 342). Even in patients who meet criteria for a moderately severe depressive syndrome, the type of depression considered most responsive to antidepressants, the improvement rate with antidepressants is only around 70%. Furthermore, in these samepatients the response rate to placebo is between 30 and 40%. 'Improvement' and 'response' rate are not defined, and this uncovers fundamental a problem in assessing efficacyand effectiveness. Efficacy usually reflects the mean drug-placebo differenceon a standard rating scale, usually the Hamilton Depression Scale (Hamilton, 1960). A meta-analysis carried outin the early 1980s by the Australian and New Zealand College of Psychiatrists found that both the tricyclic and the newer antidepressants were associatedwith an effect sizeof 1.5,placebo with 0.8, in patients
.
with 'endogenous' depression (Quality Assurance Project, 1983). Thefigures for 'neurotic' depression were 1.5 and 1.1, respectively, Thus,antidepressants on average effect an improvement of much less than one effect size. (One effect size is one standard deviation of the initial severity levelratings (say, 4-6 points on the Hamilton).) However, clinicians are not primarily concerned with average responsesin groups of patients. They choosea treatment on the probability that it will resultin a clinically useful improvement-for example, recovery with discharge of the patient from in-patient, out-patient or primary care. The figures given earlierwould suggest that, evenwith carefully chosen patients,this probability with a TCA is around 2 in 3, compared with 1 in 3 for placebo. For less typically 'core' patients, the response probability rises from about 1 in 2 to 2 in 3, hardly an earth-shattering therapeuticmanoeuvre! The phenomenon of the placebo responseneeds further qualification. In most trials, thisepithet is applied to the improvement seenin the placebotreated group. However, it is comprisedof several elements of which the most important are natural remission, spontaneous fluctuations and true placebo response. Thus,80-90% of depressive episodesnaturally remit and the coincidence of this with treatment will overestimate the realefficacy. Secondly, patients tend to seek help when their condition is worsening, so spontaneous fluctuations in severitywithout genuine remission will tend to be towards improvement; again, efficacy is overestimated. Finally, the true placebo response is improvement consequent on the administration of a dummy medication and its psychological implications. It can only be assessed by including a 'no-treatment' group (Quality Assurance Project, 1983). "t.e newer antidepressantsdo not have increased efficacy. Indeed, theclhical impression is that they may have marginally less efficacy. This is probably a complex indirect reflectionof their better-tolerated side-effect profile. The argument runs as follows: SSMs are being perceived as having fewer and less severe side-effectsthan the TCAs. Therefore, manyprescribers willtry them in patients to whom theywould not normally giveTCAs, because of these side-effects. However, these patients, being less typical depressives, tend to be less responsivethan typical moderate to severely depressed patients and respond less. Certainly, inspection of the data bases for some of the newly marketed antidepressantssuggests that drug-placebo differences are narrowing. This is so for the newerdrugs butalso true to a lesser extent for the longestablished comparatorssuch as imipramine. Clinical trialsare generally carriedout on patients who meet criteria for major depression. This comprisesa syndrome including both psychological and somatic symptoms. Most important are a sustained depressed moodand loss of interest or pleasure. The syndrome persists for at least 2. weeks, although in practice most depressivesin clinical trials have been ill for much
longer than this. The third criterion is that the condition produces impairof functioning (Ameriment in social, occupational or other important areas can Psychiatric Association, 1994). Spurious estimatesof mean efficacy canbe obtained by relying too heavily on a single rating scale. For example, the Hamilton Depression Scale has several items suchas anxiety, psychic and somatic (numbers10 and 11) and i n s o ~ i (4,5 a and 6) which will respond to sedatives. Therefore, it is necessary to assessa medication on the core depressive items such as mood, guilt and suicide. An alternative is to use the Montgomery-Asberg Depression Ra ing Scale whichis a ’purer’ depression scale (Montgomery& Asberg, 1979). The depressive episode phase of a bipolar manic-depressive disorder also responds to antidepressants at about the same efficacy rate as major depressive disorder. However, bipolar depressives tend to remit spontaneously so the antidepressant seems to act to expedite that remission. Indeed, the ofrole in precipitating mania remains controversial (Wehr antidepressant medication & Goodwin, 1987; Sulzer& C u ~ i n g s1989). , At the extremeof the severity con ti nu^, psychotic depressives tend to do relatively poorly with antidepressants. Although the placebo response ra is verylow, only abouta third of such patients with hallucinations and delusions will respond toa TCA, but twicethat proportion to electroconvulsive therapy (ECT) or to the combination of a TCA andan ti psychotic drug ( et al., 1991). These patients are in severe anguish, function poorly if at all, and may develop physical complications because of self-neglect, so rigorous treatment is essential. At the other end of the severity cont~uum,milder case et al., 19 show little response to antidepressants (Stewart .Part of this reflects the mildnessof the condition, so rating scores are relatively low initially, leavinglittle scope for muchimpro~ement,and also a high placebo response and spontaneous remission rate. apparently mild cases do have biological features,as poor such sleep, indifferent appetite, muted libido, and vague aches and pains. These patients often show a gratifying response to an antidepressant. Similarly, antidepressants may be effective even where the depression appears be to an ~derstandable reaction to life-stresses (Paykel& Priest, 1992). One study comparedpro~lem-solving(essentially directive counselling), amitriptyline (150mg at night) and placebo (including general support, encourag~mentand detailed progress-monitoring) in patients with major depression in primary care (M~ors- alli is et al., 1995). By week 12, 60% of patients given problem-solving had recovered compared with 52% on amitriptyline and27% in the placebo group. The category of ’atypical depression’ hasbeen bandied about in antidepressant literature for many years, with particular reference to the MAOIs.
~ n f o r ~ n a titehas l ~ not been clearly defined until fairly recently. An early mention related to 'anxiety hysteria with secondary atypical depression' est & Dally, 1959). The usual features are irritability, somatic anxiety, h~ochondriasis and phobic symptoms, often with 'reverse' vegetative symptoms such as increase in sleeping, appetite and weight, and mood worsein the evenings. These patients do best with anMAO1 (
Increasing emphasisis being placed on long-term treatment. This is generally divided into two types, although the distinction is not alwayseasy to make in practice (Thase & Sullivan, 1995)(Fig. 9.1).Relapse is c o m o n in patients who have apparently responded to treatment. This is typically about 50% of
Progression into mood disorder
Normal mood
. ..
8 1
(I
li
Symptoms
t
-
Syndrome
I
Acute
I
Continuation Treatment phases
ig.9.1 A diagram of five possible outcomes (1, response; 2, remission; 3, relapse; 4, recovery; and 5, recurrence) during the three phases of treatment of depression (acute, continuation and maintenance).
responders and suggeststhat the clinical modeof action of antidepressants, in these patients at least, is to control or suppress symptoms until natural (10-20% of patients), remission takes place. If natural remission does not ensue then maintenance drug treatment may be needed indefinitely. The second type of long-term therapy is to prevent recurrenceof depression, prophylaxis. This assumes that full recovery has taken place, i.e. that the patient is euthymic without treatment but is constantly at risk of a further episode. The long-term antidepressant is then used to 'prevent' that recurowever, this prevention of clinical symptoms might not be true
prophylaxis at all but rather the controlof symptoms in the episode, which therefore remains subclinical. Prien and Kupfer (1986) reviewed the studies to that date which evaluated the effectiveness of antidepressants in preventing relapse. The overall conclusionwas that the relapserate on placeboduring a 2- to 8-month period of observation was50%as compared with around 25% on activedrug, mostly a TCA. Belsher and Costello (1988) point out the need to control for number of previous episodes,duration of normal mood before entering the comparative clinical trial and clear definitionsof recovery and relapse, with regular follow-up interviews,rather than relying on medical records. Similar methodological considerations occur in prophylactic trials.The need for a carefully defined period of recovery from theprevious period of depression is paramount. Relapse is commonest in the first 4 weeks after discontinuing an antidepressant but there is still some risk up to 4 months. Accordingly, a symptom-free period of 4-6 months is the minimum and removes almost all risk of labelling a relapse as a failure of prophylaxis. Discontinuation of the antidepressantin a placebo-controlledstudy should not be abrupt as a short-lived withdrawal with autonomic symptoms may confuse the clinical picture as well as jeopardize the double-blind procedures. Other considerationsare adequate criteria for illness recurrence and sufficient previous morbidity fora prophylactic effect to be manifestwithin 2-3 years (Montgomery& Montgomer~1992). I m i p r a ~ is e the best-studied drug with respect to prophylaxis. The larger studies have shown a significant superiority. One study involved 75 patients treated with imipramine, placebo orinterpersonal therapy and followed up for 3 years aftera 20-week symptom-free period (Frank et al., 1990). Only 22% of imipra~e-treatedpatients had a further episode compared with 78% on placebo and 62% given interpersonal therapy. Out-patients with major depressive disorder in the National ~ s t i t u t eof Mental Health Treatmentof Depression CollaborativeStudy received either cognitive behaviour therapy,interpersonal therapy, imipramineplus clinical management or placebo plus clinical management(Shea et al., 1992).After 16 weeks of treatment, follow-ups were conductedat 6,12 and 18 months. Response rates were equal in the four groups varying between 19% and 30%~ i.e. surprisingly low. Between a third and a half relapsed subsequently, emphasizing the need for sustained treatment. Evaluation of the SSNs long-term has involved largenumbers of patients. A rigorous study using fluoxetine in unipolar depression recruited 456 patients treated openly (Montgomery et al., 1988).Of these, 220 responded and remained well for 18 weeks. They were then allocated randomly and doubleblind either to continue on fluoxetine or to switch to placebo 1year. for Those on fluoxetinehad a 26% risk of developing a new episode, those on placebo
over twice this risk. Similar are available for paroxetine, sertraline and citalopram. The data from a combined relapse/recurrence prevention trial .Signifiover a year comparing paroxetine and placebo are shown in9.2 Fig.
0.8 c
.-
ti n.
$2
0.7
Log rank testp = 0.005
0.6
0.5
1.o
U
0.9
F
0.8
2 7
.......:............ ............... .............. Z'"
W
C
c
.-
9.4..
0.7
.................
ti
g
0.6
Log rank testp = 0.005
Q
0.5
0.1
0.0 0
250 50
100 200
150
300
Days from start of maintenance phase Placebo
Paroxetine
..............
Time to relapse or recurrence, paroxetine placebo-controlledstudy. (From Montgomery S.A. & Montgomery,D.B. (1992) Prophylactic treatmentin recurrent unipolar depression. In: Long-term ~ r e ~ ~ m e n ~ (eds o ~S.~Montgomery ~ r e s s ~ &~F. n Rouillon), pp. 53-79. Copyright John Wiley &:Sons Limited. Reproduced with permission.)
350
400
cant effects were found for both therapeutic effects ontgomery, 1992).
(
The dosageof antidepressant is often inadequate for full efficacy. be as low as a thir ticularly so in primary care where dosages of TCAs may of the recommendedlevel, particularly in women ( contrast, dosagesof SS S in primary care usually attain r e c o ~ e n d e levels d The reason for this is m a ~ toyavoid side-effects( ver, there are few good dose-effect studies with the antide1985). A further com~lication is that the dos nd to be higher in the USA than in Eu TCA are deemed adequate in the g and, rarely, 300 m for in~patientswho have In the USA, the acce ted therapeutic r failed to respond to the lower dose. starts at 225 mg and goes up to 300 mg and beyond. The recommended dosagedoes not depend on efficacy alone. atio is the relevant attri~uteat each dose level. Also, do§age§ are in a sequence, usual increase, occa~ionallya decrease in the unuly sensitive. Thus, compl o r i t ~ are s needed but fewif any of the Osage schema~aare suppo clear data. Accor schedules persist, usually obtained usin titration reflected t o l e r a ~ i lrather i~ than fluence is the ostensibledelay in onset of action whic even moreof a hit-and-miss affair. One dose-ranging study has ex fl~oxetine. Fluoxe~ne showed margin ut full efficacy was not reached until zomg daily ( It was generally accepted that when a patient ressive episode,a lower dose canbe used in maintenance andproph~laxis. herearefewdatatosupportthisandmostrecent -term studieshave to use full doses throughout.
clearer fora dose relations ore effective than low S to correlate with deg Although dose-effect relationships are difficu efficacy, clear r e l a ~ o n s h i ~ are s usually discer side-effects such as dry mouthand consti
had spent5 weeks in hospital. Marked differences emerged in the severity and rate of response of variouscomponentsofthessiveillness.Sadaffect, l thinkingand somatic lack of interest, hopelessness, helplessness, S d y s f ~ c t i o nrapidly recovered. An intermediate rate was observed for low self-esteem, sleep disturbances, cognitive slowing, impulsivit~,emotional blunting, self-criticism, anhedonia, and subjective memory complaints. Specific features like guilt, intropunitive anger,ac~uiescence,low expe~tancies, indecision, suspicion, hostility, dimi~shed libido and anxiety re~itted slowly. psycho~otor, immediate memo^ and pairedAfter 5weeks, performance on associate learning tasks reached a plateau. It was apparent that the response of some items was fairly prompt, raising ~uestionsabout the belief that response to antidepressants is delayed, the subject of the next topic. A study from the USA considered patterns of response in severel de( pressed in-patients categorized as full responders or non-responders al., 1987). Inresponders, improvement overall and in depressed mood, cognitive impairment, anxiety and physical s i p s was apparentas early as 1 week, whereas retardation and somatization were only significantly improved by 2 weeks. The differences between this and the previous study probably relate to differencesin criteria fori m ~ r o v e ~ e nthe t , former using clinical as well as statistical criteria.
From the earliest clinical experience with the TGAs and the MAOIS, it was parent that the depressed mood was slowlift toin the courseof the therautic response. This has led to oft-repeated statements concer~ng the delay in the onset of action of antidepres ts and the search, usually disappoint,for fast-acting agents (Norma A thorough search of the literature by Danjou et al. (1~94) identified 127 clinical trials which were appropriate for analysis with respect to onset of action definedas 'the first assessment with a statistical differencevs. placebo'. If no placebo difference ensued, then the onset of action was taken as the dura~on of the study. The mean onsetof action for the TCAs Is, 3.1 weeks; and for MAOIs, 6 weeks. By far the la studies (41) involved imipramine, anda weak positive relat was found between onset of action and dose, being lon This p r e s u ~ a breflects l~ the time taken to titrate the dose upwards. Conversely, data with venlafaxine showed a correlation (-0.52) in the opposite direction, believed due to the ability to ain high and effective doses idly. A re-working of such analyses usi only fixed dosage data would be very informative, asit would obviate titration delays.
Manifestly, one reason for delay in ofonset response reflects the time taken to attain therapeutic dosages. This is only a minor factor as the delays with fluoxetine (17 studies) and paroxetine(17 studies) were the same as the §§MS with the TCAs, and yet most patients start treatment on a standard dose which is not increased. A second factor concerns the detection of response. Various scales may be differentially sensitive to different symptomatic components. For example, the Hamilton Depression Scale has three items relating to sleep which would be sensitive to early actions of sedative antidepressants. A more important factor concerns the poolingof data from eventual responders and non-responders (Potter & Manji, 1994). Thus, patients who show no response will introduce a delay into the composite of the curve total group. Again, analysis of data from responders only would be useful and would probably show an onset of action around1-2 weeks forboth TCAs andSS&. After having taken all the above factors into account, it cannot be gainsaid that depressed patients take time to improve when treated with an antidepressant. This delay has been attributed to biochemical adaptive changes which take time to develop after the immediate changes in amine neurotransmitter disposition.
In his original account, Kuhn (1958) observed that the closer the patient approximated to theclassic 'endogenous depressive', the more likely was an adequate response. That remains the firm clinical impression. Nevertheless, up to 30% of such patients fail to show an adequate response and are termed 'refractory' (Guscott & Grof, 1991; Sokolov& Joffe, 1995). Antidepressants are effective across a wide rangeof severity, all but the mildest forms responding to atleast some extent (Paykelet al., 1988). However, recurrentbrief depressions appear tobe fairly unresponsive to antideof the continuum, pressant medication (Kasperet al., 1995). At the other end patients with psychotic (delusional) depression tend to need either the addition of an antipsychotic drug to the antidepressant or the of use ECT. A coruscating review by Joyce andPaykel(1989) concludes that the TCAs are b r o a ~ - s p e c ~ udrugs m 'that convey therapeutic benefit over a wide range of depressions and some anxiety disorders' (p. 95). The patient with a good premorbid personality,, an insidious onset of illness, psychomotor retardation, and intermediate in severity and endogeneity but without psychotic feawere regarded as an important tures is most likely to respond. TheM21.01~ option in previous responders, and as a second-line treatmentin refractory depression, and in patients with marked anxiety/panic features. Biological predictors were relatively uninformative.
e adventof the SSRIs has not changed this overall assessment. Efficacy is similar to thatof the TCAs, as is the type of patient who res times newer drugs have modified or additional indications in the data sheet, but this usually reflects proven efficacy in that subgroup of patients rather than differential efficacy. on-endogenous depression remains more difficult to treat than patients with classical 'endogenous' depression. Even then, more severe symptoms t loss were predictiveof better response (Parkeret al., 1985). Another important predictor of good response was the break-up of an intimate relationship in the12 months preceding treatment. Early signsof response predict later full response. Thus, improvement in overall depressedstate as well as on depressed mood, cognitive im~airment, anxiety and physical symptoms was apparent after a week in ultimate responders. Retardation and somatization improvements were also predictive but only after2 weeks (Katzet al., 1987).
The lengthof time needed for antidepressant treatment has become aoftopic debate recently. The tendency has been to recommend longer and longer courses of treatment as the recurrent/chronic na re of depression has become ~creasingly appreciated and the risk of relapse on premature ~i~drawal better quantified. The criteria for relapse and recurrence have been discussed el (1994) and by Kupfer and Frank (1992)~ and can be set against the d of the natural history of depressive illnesses( ~ g s t1992). , Thus, 15% commit: suicide. about 15-20% of depressives become chronic and about After the onsetof the illness, about20% of the subsequent lifetime is spent in t, 1992). Between episodes, many patients remain s~ptomatic, only one episode. AUK study of 70 patients showed that remission was quite rapid with70% doing so within 6 months (Ramanaet al,, 1995). However,40% relapsed over the subsequent 10 months. Greaterinitial ted longer time to remission and greater risk of relapse. Collaborative Program on the Psychobiology of ~epression concentrated on in-patients in special research centres ( eller et al., 1982a). o-thirds recovered within the first 4 months but26% ere still not recovered at 1 year; 21%, at 2 years; and 12%, at 5 years (Keller et al., 1992). After 3 months; but only remission, 12% relapsed within a month; a quarter, within 30%~by a year (Keller et al., 1982b). After relapse, recovery was less likely than initially. esidual symptoms after the initial response are predictive of relapse if medication is discontinued: about a third of patients fall into this category Spiker, 1981). Even mild symptoms are artial responders ( 00
predictive of relapse. ~ithdrawal of therapyis safe only after the patient has been freeof symptoms for16-20 weeks (Prien8r Kupfer, 1986). The definition of relapse is, itself, non-standard. Sometimes it refers to a recurrence of symptoms, sometimes to an operational change, such as admission to hospital. The borderline between relapse of the current episode and development of a new episode also involves an arbitrary decision. A symptom-free interval of z6-20 weeks is generally taken as indicating that the initial attack has resolved and that the risk is of a recurrence.
Although thereis a plethoraof studies investigating the initiation and maintenance of antidepressant treatment, very few address the issue of discontinuing medication. Greden(1993) recommends that the schedule should be gradual, over several weeks or even a few months. After long-term maintenance, the tapering period should be even longer. He opines 'that the brain requires months rather than days or weeks to adjust to the changes associated with stopping amedication'. TCAs with anticholinergic potency, particularly amitriptyline, require special care becauseof the problemof cholinergic rebound (Dilsaver& Greden, 1984). Symptoms include gastrointestinal distress, restless sleep with vivid nightmares, frequent awakenings, and anxiety and agitation. The withdrawal of an MAO1 should also be gradual. because the usual episode length in patients who respond be much longer than in more typical patients responding to a TCA or an SSN (Robinson et al., 1991). Furthermore, relapse maybe particularly slow AOI withdrawal. o planned discontinuationof treatment after successful response. Dropout refersto unplanned discontinuation, often early in treatment and most commonly related to treatment failure and/or intolerable sideeffects. Indeed, the dropout rate due to side-effects has been used as a crude indicator of tolerability (Lader, 1988). Dropout rates due to side-effects in SSNs (Song et al.,1993). In patients on TCAs tend to be higher than those on one meta-analysis of 42 randomized comparative controlled studies, the dropout rate due to side-effects was 19% in those on TCAs, 15% in those on SSNs (P~ 0 . 0 1 )(~ontgomeryet al., 1994). Another meta-analysis of 62 ies yielded dropout rates of 18.8% in patients on TCAs as compared with ~ ~ . L + O / Oin those given SSRIs(P
Depression is often encountered in conjunction with other psychiatric condi201
tions, particularly anxiety in its various forms (Keller &: Hanks, 1995). The unitary view considers the two disorders tobe a singleentity expressing both groups of symptoms concurrently or over time separately. Or the presenceof one condition may make the onset of the second, ostensibly independent, condition more likely. Third, comorbidity may just reflect the chance & Michelini, 1995).Angst concatenation of two common conditions (Cassano and colleagues (1990) in their detailed longitudinal studyin Zurich showed that primarily anxious individuals were more likely to develop comorbid depression as follow-up thanvice versa. Comorbidity occurs between depreset al.,1992), sion and panic disorder, and generalized anxiety disorder (Cassano as well as with social and simplephobia, obsessive-compulsi~edisorder, and probably post-tra~matic stress disorder. Concern about the use of benzodiazepines in the treatment of anxiety disorders, particularly in the long term, has led to a convergence of treatment stratagems in these comorbid conditions. c tide press ants of various types are widely used in these disorders, not only to treat the depressive component but increasingly as effective therapy for the panic, anxiety or obsessivecompulsive elements.In particular, drugs acting on serotonin seem effective (Burrows et al., 1994). However, more systematic evaluations are needed to establish the degree of efficacy in these cornorbid conditions. The use of antidepressants to treat 'pure' anxiety disorders lies outside the scope of this chapter.
Reservations have been expressed about the efficacy of TCAs in the treatment of childhood depression. A meta-analysis of $2 studies revealed no significant effect vs. placebo, either on mean response or number of responders azell et al., 1995). Data on SSRIs and MAOIs are insu~icientto allow any conclusions tobe drawn. everth he less, clinical experience still supports the &:Hussein, 1995). use of antidepressants in selected cases (Kaplan
The elderly The antidepressants in general are of proven efficacy in the treatment of the elderly depressive. Indeed, several regulatory authorities ask for a substantial minority of patients in the clinical trials of a new drug to be elderly in order to establish efficacy and tolerability in this age group. Dosages are generally low, at least initially, and dosage escalation should be gradual. This is because of complex pharmacokinetic changes in the elderly, particularly in those aged over85 years. 202
factor is the effect of cognition of both depression an an~depressants.However, cognitive function in general isnot dispropor~onately impaired in the elderly as compared with younger depressives, although the speedof performance may fall (Tarbuck Paykel, 1995). The use of with anticholinergic effects may further i ir performance although seem better tolerated in general (e.g. Fairweather et al., 1993). Forthis reason, the SSNs or other newer drugs (e.g. lofepramine, nefazodone) are increasingly preferred tostandard TCAs (Dufour et al., 1994). The ~hysicallyill
The TCAshave many effects on many bodily systems and some of these may compromise organ function. The best-known arethe cardiovascular effects, and special precautions are needed when treating depressed patients with cardiovascular disease (Glassman & Stage, 1994). The SSRIs are usually to be preferred. Renal and hepatic impairment are other conditions where care should be takenin choosing an antidepressant and its dosage (Hale,1993). Almost all antidepressants lower the convulsive threshold so epileptics may have their condition exacerbated. Many other physical conditions may affect the side-effect profileof individual drugs-for example, diabetes and parkinsonism. Finally,drug interactions mayoccur, and again patterns vary from drug to drug (Taylor & Lader, 1996).
These have been largely neglected until recently. Akiskal(1995) has reviewed the relationships betweentemperaments and depression and concluded that recognition of the importance of various personality factors reduced the heterogeneity of major depressive disorders as presently defined. Oneefficacy study suggested that patients who failed to respond to both a TCA and an MAO1 given at various times were more likely to an have abnormal personality than those who did respond (Shawcross& Tyrer, 1985).
Drug addicts
Some concernhas been expressedabout the abuse potential of antidepressants, particularly the more stimulant ones. In France, the antidepressant amineptine appeared to be associated with some risk (Castot, 1990). However, the risk with other antidepressants seems vanishingly small,and antidepressants should never bewithheld from depresseddrug addicts because
and deaths per million define daily doses for deaths .~ a l u ein s ~a~entheses are 9~~~con€i~ence Mean defined Deaths per daily Expected million doses, daily defined (mg) dose deaths, 1987-92 1987-92
Observed deaths, 1987-92
1346
1563
41 oxidase i n ~ i b ~ t o r s
12
Atypical 62drugs
26
Selective s e r o t o ~ re-uptake ~ i b i t o r s
All antidepressants 1.26 1606
5
34.98
20.90
1606
ifference fromall by
20.51
0.43 (0.19to 0.60)"
165
0.16 (0.34 to 0.74)"
66
0.08 (0.03to 0.23)
89
x* test).
1987-92 16.63 (1.36to 1.50)*
94
28
46.45 56
x2value
C ~ a ~9:t Aen ~ i d e ~ ~ e s s a n t ~ e d i ~ a t i
The adventof the SSRIs has opened up a massive, on-going debate concerning their merits vis-his the TCAs. Similarly, but on a much more muted note, the in~oductionof the selective and reversible inhibitor of MAO, moclobernide, has led to comparisons with the old MAOIs.This has coincided with aninsistence in many countries that cost-effectiveness mustbe a major factor in the choiceof medication. The SSRIs are more expknsive than the older TCAs, most of which are availableas generics. Many elements enter the comparancluding side-effect profile, tolerability and compliance, safety ry & Martin, 1987), and usefulness in special groups such as the elderly The one aspect which hardly enters into the equation is efficacy, the focus of this chapter. Some atypical antidepressants such as nornifensine (withdrawn some years ago), trazodone and perhaps mianserin have been found less effective than imipramine in meta-analyses (Muller et al., 1994). Another meta-analysis concluded that the SSRIs were slowerin action than TCAs but were equal in efficacy by 6 weeks, except on sleep items (Bech, 1990). A review some years ago concluded that TCAs arestill preferable in patients who have responded to them previously, and in patients where plasma concentration monitoring mightbe helpful (Rudorfer& Potter, 1989). More recently, Songet al. (1993) analysed 63 randomized controlledtrials comparing an SSRI with an older antidepressant. No difference in efficacy was found. A sirnilar exercise identified55 double-blind studies and found no differences in efficacy betweenSSRIs five and comparator TCAs (Anderson & Tomenson, 1994). There was some suggestion in the data that the TCAs had the edge over the SSNs in the more severely depressed patients and with inpatients. This was due to a clinically significant lower efficacy for paroxetine in those patients. Another meta-analysis indicated that trazodone, buproprion, fluoxetine and imipramine were all equally effective compared with placebo (Workman & Short, 1993). It is not surprising that efficacy differences among the antidepressants are marginal, if they exist at all. With about a third of patients non-responders and a third placebo responders, the efficacy ’window’ is quite narrow. Reduced efficacy would have to be substantial be detected, to as would enhanced efficacy. The debate continues with respect to the relative of merits various antidepressants. For example, Edwards(1995)~ after a detailed reviewof the topic,
t of existing knowledge, the decisionas to whether to reatment for depression and the opinion as to whether e cost must remain as value ju~gements’(pp. 156-7). 94) avers: ’But for the first line treatmentof patients seen every day in surgeries and psychiatric clinics,isnew better’ (p. 1281). In my experience,if one actually asks patients for their opinion, they generally prefer to takean SSRI ratherthan. a TGA.
It is not the purposeof this chapter to review the individual dru & Herrington, 1996; Recourse canbe made to standard textbooks (e.g. Lader Schat~berg& Nemeroff, 1995). The reader will also find useful the recent reviews and paperson individual drugs inTable 9.2.
Antidepressant drugs: reviews and key references. Drug Tricyclic a n t i ~ ~ r e s s a n t s Clomipramine Dothiepin Lofepramine Selective serutunin re-uptake inkibiturs Citalopram Fluoxetine Fluvoxamine Paroxetine
Sertraline Selective serutuninlnuradrenalinere-~ptakeinkibi~urs Venlafaxine Serutunin re-~ptakeinkibitur and antagonist Nefazodone Reversible inkibiturs of MAO Moclobemide
Brofaromine A ~ i nprecursurs e Tryptophan
Reference McTavish and Benfield (1990) Lancaster and Gonzalez (1989a) Lancaster and Gonzalez(198913) Milne & Goa (1991) Wernicke et al. (1988) Wilde et al. (1993) Dechant and Clissold (1991) Holliday and Plosker (1993) Murdoch and McTavish (1992) Artigas (1995) Cowen (1995) Holliday and Benfield (1995) Rickels et al. (1994) Fitton et al. (1992) Priest and Schmid-Burgk (1994) Moll et al. (1994) Volz et al. (1995) Kauhan and Philen (1993)
Depression is a common and distressing condition. It is associated with impairment of psychosocial f u n c t i o ~ n g ~ d u c a t i ooccupational, n, marital and
06
social-which tends to persist (Coryellet al., 1993). It imposes medical, social and economic burdens on the community: one estimate was that the annual USA total $43.7billion, comprising 28% direct costs costs of depression in the (medical care, etc.), 17% mortality costs including suicide, and55% indirect costs such as lossof work capacity (Greenberget al., 1993). Despite this, treatmentis often barely adequate. One survey showed that less than one-thirdof depressed out-patients used antidepressant medica~on (Wells et al., 1994). The need for more education among the public and also primary care practitioners, who see the bulk of depressed patients, is manifest (Paykel& Priest, 1992). However, in the harsh economic climate that now prevails, justification for the use of all but the cheapest drugs is demanded by our political masters. Increasingly, pharmacoeconomic considerations are entering not only into the use of newer antidepressants but also in their licensing (Wilde& Whittington, 1995). Let us strive to ensure that our choice of antidepressant medication remains a wide so onethat patients can receive the drug that suits thembest. And in the long run, let us hope that induceof novel agents ments remain to encourage the developmentintroduc~on and of enhanced efficacy.
tric Press, Washington,DC., pp. 123-137. Akiskal, H.S. (1995) Toward temperament-based a approach to depression: implications for neuro- Artigas, F. (1995) Selective serotonin/noradbiologic research. In: Depression and M a n ~ a ~ renaline reuptake inhibitors (SNRIs).PharmaFrom N e u r o ~ i o l to o ~~ r e a t ~ e(eds n t G. L.Gessa, cology and therapeuticpotentialin the treatment W. Fratta, L. Pani & G. Serra), pp. 99-112. of depressive disorders. CNS Drugs, 4,79-89. Raven Press,New York. Balant-Gorgia, A.E., Gex-Fabry, M. & Balant, L.P. (1991) Clinical pharmacokineticsof clomiAmerican Psychiatric Association (1994) Diagnostic criteria for DSM-IV. American Psychiatric 20,447-462. pramine. Clinical Pkar~aco~inetics, Association, Washington,DC. Beaumont, G., Baldwin, D. & Lader, M. (1996) A criticism of the practice of prescribing subAnderson, I.M. & Tomenson, B.M. (1994)The efficacy of selective serotonin re-uptake inhibitors therapeutic doses of antidepressants for the in depression: a meta-analysisof studies against a n treatment of depression. ~ u ~ Psyckopkarm a c o l o ~11,283-291. , tricyclic antidepressants. Journal of PsychoBech, P. (1990) Ameta-analysis of the antidepres~ ~ a ~ a c 8,238-249. o l o ~ , Anderson, I.M. & Tomenson, B.M. (1995) Treatsant properties of serotonin reuptake inhibitors. In~ernationalReview of P s y c ~ i a t r2,207-211. ~, ment discontinuation with selective serotonin reuptake inhibitorscompared with tricyclic anti- Belsher, G. & Costello, C.G. (1988) Relapse after depressants: a meta-analysis. ~ r i t i s kMedical recovery from unipolar depression: a critical Journal, 310,1433-1438. le~in, review. P s y c k o l o ~ ~ ~ u l104,8496. Angst, J. (1992)How recurrent and predictable is Brown, W.A. & Khan, A. (1994)Which depressed depressive illness? In: Long-term ~ r e a t ~ ofe n ~ patients should receive antidepressants? CNS ~ ~ r e s s i (eds o n S. Montgomery & F. Rouillon), Drugs, 1,341-347Buckley, N.A., Dawson, A.H., Whyte, I.M. & pp. 1-13. Wiley, Chichester. Henry, D.A. (1994)Greater toxicity in overdose Angst, J., Vollrath, M,,Merikangas, M.R. & Ernst, C. (1990) Comorbidity of anxiety and depresof dothiepin than of other tricyclic antidepressants. Lancet, 343,159-162. sion in the Zurich cohort study of young adults. In: C o m o r ~ i ~ ioft yMood and Anxiety Disorders Burrows, G.D., Judd, F.K.& Norman, T.R. (1994) (edsJ. Maser & R. Cloninger). American Psychia- Differential diagnosis and drug treatment of
panic disorder, anxiety and depression. CNS Drugs, I, 119-131. Cassano, G.B. & Michelini, S. (1,995)Pharmacological treatment of depression and comorbid anxiety disorders. In: De~ressionand Mania: From ~eurobiology to Treat~ent (eds G. L. Gessa, W. Fratta, L. Pani & G. Serra), pp. 113-125. Raven Press, New York. Cassano, G.B., Savino, M. & Perugi, G. (1992) Comorbidity of mood disorders and anxiety states: implications for long-term treatment. In: Long-term Treatment of D e ~ r e s s i o n(eds S. Montgomery & F. Rouillon), pp. 229-243. John Wiley, Chichester. Castot, A. (1990) Sur consommationd'amineptine. Analyse de 155 cas. Bilande l'enquete officielle cooperative des centres regionaux de pharmacovigilance. Therapie, 45,399-405. Correspondence (1994) Selective serotonin reuptake inhibitors. British Medical Journal, 309, 1082-1085. Coryell, W., Scheftner, W., Keller, M.,Endicott, J., Maser, J. & Klerman, J.L. (1993) The enduring psychosocial consequences of mania and depression. American Journal of Pssychiatrsy, 150, 720-727. Cowen, P. (1995) Venlafaxine:a new class of antidepressant drug. Prescriber, June,19-26. Paradoxical Damluji, N.F. & Ferguson, J.M. (1988) worsening of depressive symptomatology caused by antidepressants. Journal of Clinical P s yc h o p h a r ~ a c o l o8,347-349. ~, Danjou, P., Weiller,E. & Richardot, P. (1994)Onset of action of antidepressants: a literature survey. t Affective In: Critical Issues in the Tr e a t ~ e nof Disorders (eds S.Z. Langer, N. Brunello, G. Racagni & J. Mendlewicz), pp. 136-153. Karger, Basel. Dechant, K.L. & Clissold, S.P. (1991) Paroxetine: a review of its pharmacodynamic and pharmacokinetic properties, and therapeutic potential in depressive illness. Drugs, 41,225-253. Dilsaver, S.C.& Greden, J.F. (1984)Antidepressant 19, withdrawal phenomena. Biological Psychiat~, 237-256. Donoghue, J.M. (1995)A comparison of prescribing patterns of selective serotonin reuptake inhibitors in thetreatment of depression in pril mary care in the United Kingdom. J o ~ r n a of Serotonin Research, I,47-51. Dufour, H., Wertheimer, J., Baumann, P. & Bertschy, G. (1994) Efficacyand safety of antidepressants in the elderly. In: Critical Issues in the T r e a t ~ e nof t Affective Disorders (eds S. Z. h g e r , N. Brunello, G. Racagni & J. Mendlewicz), pp. 44-51. Karger, Basel.
208
Edwards, J.G. (1995) Drug choice in depression: selective serotonin reuptake inhibitors or tricyclic antidepressants?CNS Drugs,4,141-159. Fairweather, D.B., Ken, J.S., Harrison, D.A., Moon, A double blind comC.A. & Hindmarch,I. (1993) parison of the effects of fluoxetine and amitriptyline on cognitive function in elderly depressed patients. ~ u ~ Psa y cnh o p h a r ~a c o lo ~, 8,41-47. Fernando, M.L.D. & Kazarian, S.S. (1995) Patient education in the drug treatment of psychiatric disorders. CNS Drugs, 3,291-304. Fitton, A., Faulds, D. & Goa, K.L. (1992) Moclobemide: a review of its pharmacological properties and therapeutic use in depressive illness. Drugs, 43,561-596. Frank, E., Kupfer, D.J. & Perel, J.M. et al. (1990) Three-year outcomes for maintenance therapies in recurrent depression. Archives of General Pssychiatry, 47,1093-1099. Freemantle, N., House,A., Song, F., Mason, J.M. & Sheldon, T.A. (1994) Prescribing selective serotoninreuptake inhibitors as strategy for prevention of suicide. British Medical Journal, 309, 249-253. Glassman, A.H. & Stage, K.B. (1994) Depressed patients with cardiovascular disease: treatment considerations. CNS Drugs, I,435-440. Greden, J. (1993) Antidepressant maintenance medications: when to discontinue and how to stop. Journal of Clinical Pssychiatry,54 (Suppl. 8), 39-45. Greenberg, P.E., Stiglin, L.E., Finkelstein, S.N. & Berndt, E.R. (1993) The economic burden of depression in 1990.Journal of Clinical Ps y c h ia t ~, 543 405-418. Guscott, R. & Grof, P. (1991) The clinicalmeaning of refractory depression: a review for the clini~, cian. American Journal of ~ s y c h i a t148,695-704. Hale, A.S. (1993)New antidepressants:use in highrisk patients. Journal of Clinical Psychia~ry,54 (Suppl. S), 61-70. Hamilton, M. (1960)Arating scale for depression. ~ournalof Neurology, Neurosurgery and Psychiatry, 23,56-62. Harrison, G. (1994) New or old antidepressants? New is better. British Medical Journal, 309, 1280-1282. Hazell, P., OConne11, D.,Heathcote, D., Robertson, J. & Henry, D. (1995) Efficacyof tricyclic drugs in treating child and adolescent depression: a meta-analysis. British Medical Journal, 310,897901. Henry, J.A. & Martin, A.J. (1987) The risk-benefit assessment of antidepressant drugs. Medical Toxicology, 2,445-462.
~ ~ a ~9: A t en tr i ~ e ~ r e s s ~ n ~ American Journal of ~s y c h ia t y115/ , 459-463. Henry, J.A., Alexander, C.A. & Sener, E.K. (1995) Relative mortality from overdose of antidepres- Kupfer, D.J. & Frank, E. (1992) The minimum sants. British Medical ~ournal,310,221-224. length of treatment for recovery In: Long-term Treatment of Depression (eds S. Montgomery & F. Holliday, S.M. & Benfield, P. (1995) Venlafaxine: a review of its pharmacology and therapeutic Rouillon), pp.33-52. Wiley, Chichester. Kupfer,D.J. & Spiker,D.G. (1981) Refractory potential in depression.Drugs, 49,280-294. by cliniHolliday,S.M. & Plosker, G.L. (1993) Paroxetine: a depression: prediction of non-response cal indicators. ~ournalof Clinical Psychiatry, 42, review of its phar~acology,therapeutic use in depression and therapeutic potential in diabetic 307-312. Lader, M. (1988) Fluoxetine efficacy vs. comparaneuropathy. Drugs and Aging, 3,278-299. tive drugs: an overview. British ~ournalof PsyJick, SS., Dean, A.D. & Jick, H. (1995) Antidech~atry,153 (Suppl. 3),51-58. pressants and suicide. British Medical Journal, Lader, M.& Herrington, R.(1996) Biological Treat310, 215-218. 2nd edn. Oxford University ments in P s y c h ia t ~, Johnson, D.A.W. (1981) Depression: treatment compliance in general practice. Acta Psyc~ia~rica Press, Oxford. Lader, M., Lang, R.A. & Wilson,G.D. (1987) Scan~inavica,63 (Suppl. 290)~447-453. P a t t e ~ of s improve men^ in Depressed In-patients. Joyce, P.R. & Paykel, E.S. (1989) Predictors of drug response in depression.Archives of General PsyOxford University Press, Oxford. Lancaster, S.G. & Gonzalez, J.P. (1989a) Dothiepin: c h i a ~46,89-99. ~, a review ofitspharmacodynamicand Kaplan, C.A.& Hussein, S. (1995) Useof drugs in pharmacokinetic properties, and therapeutic child and adolescent psychiatry. B r i t ~Jouna~ s ~ of efficacy in depressive illness. Drugs,38,123-147. Psychiatry,166,291-298. Lancaster, S.G. & Gonzalez, J.P. (1989b) Kasper, S., Stamenkovic, M. & Fischer, G. (3995) of its pharmacodynamic Recurrent brief depression: diagnosis, epidemi- Lofepramine: a review and pharmacokinetic properties, and therapeuology and potential pharmacological options. tic efficacy in depressive illness. Drugs, 37, CNS Drugs, 4,222-229. 123-140. Katz, M.M., Koslow,S.H. & Maas, J.W. et al.(1987) D.& Benfield, P. (1990) Clomipramine: The timing, specificity and clinical predictionMcTavish, of an overview of its pharmacological properties tricyclic drug effects in depression.Psychologiand a reviewof its therapeutic use in obsessive cal Medicine, 17, 297-309. Drugs, Kaufman, L.D.& Philen, R.M. (1993) Tryptophan: compulsive disorder and panic disorder. current status and future trends for oral admin-39, 136-153. istration. Drug Safety, 8,8998. Milne, R.J. & Goa, K.L. (1991) Citalopram: a reKeller, M.B. & Hanks, D.L. (1995) Anxiety sympview of its pharmacodynamic and pharmatom relief in depression treatment outcomes. cokinetic properties, and therapeutic potential in depressive illness.Drugs, 41,450-477. ~ o u ~of aClinicaZ l Psychiat~, 56 (Suppl.6), 22-29. Keller, M.B., Shapiro, R.W., Lavori,P.W. & Wolfe, Moll, E.,Neumann, N., Schmid-Burgk,M., Stabl, N.(1982a) Recovery in major depressive disor- M. & Amrein, R. (1994) Safety and efficacy durder: analysis with the life table and regression ing long-term treatment with moclobemide. models. Archives of General P s y c ~39,905910. ~ ~ , ClinicalN ~ r o p h a ~ a c o17~ (Suppl. o ~ , I), S74-S87. Keller, M.B., Shapiro, R.W., Lavori, P.W. & Wolfe, Muller, H.-J., Fuger,J. & Kasper, S. (1994) Efficacy N.(1982b) Recovery in major depressive disof new generation antidepressants: meta-analyorder: analysis with the life table. Archives of sis of imipramine-controlled studies. P h a r ~ a General P s ~ c h i a t39, ~ , 911915. copsychiatry,27,215-223, Keller, M.B., Lavori, P.W. & Mueller, T.I.et al. (1.992) Montgomery, S.A.& Asberg, M. (1979)Anew deTime to recovery, chronicity, and levels of pression scale designed to be sensitive to change. psychopathology in major depression: a 5-year British ~ o u r ~ofaPsychiatry, l 134,382-389. prospective follow-up of 431 subjects.Archives Montgomery, S.A. & Montgomery, D.B. (1992) Prophylactic treatment in recurrent unipolar deof General P s y c h i a t~49,809-816. , Khan, A., Noonan, C.& Healey,W. (1991) Isa sinpression. In: ~ong-termTreatment of De~ression gle tricyclic antidepressant trial an active treat- (eds S. Montgomery & F. Rouillon), pp. 53-79. ment for psychotic depression? Progress in John Wiley, Chichester. Neuropsychopharmacology and Biological Psychia- Montgomery, S.A., Dufour, H. & Brion, S. et al. (1988) The prophylactic efficacyfluoxethe of in try, 15, 765-770. Kuhn, R. (1958) The treatment of depressive statesunipolar depression.brit is^ ~ o u ~ofaP l s y c ~ ~ ~ ~ , withG22355(imipraminehydrochloride). 3/69-76.
~e~ica
Montgomery, S.A., Henry, J. & McDonald, G. et al. (1994) Selective serotonin reuptake inhibitors: meta-analysis of discontinuation rates. ~nternational Clinical Psychopharmacolo~, g, 47-53. Munizza, C., Tibaldi, G., Bollini, E.,Pirfo, E., Punzo, F. & Gramaglia, F.(1995) Prescription pattern of antidepressants in out-patient psychiatric practice. PsychologicalMedicine, 25,771778. Murdoch, D. & McTavish, D. (1992) Sertraline: a review of its pharmacodynamic and pharmacokinetic properties, and therapeutic potential in depression and obsessive-compulsive disorder. Drugs, 44,604424. Mpors-Wallis, L.M., Gath, D.H., Lloyd-Thomas, A.R. & Tomlinson, D. (1995) Randomised controlled trial comparing problem solving treatment with amitriptyline and placebo for major depression in primary care. British MediculJournal, 310,441-445Norman, T.R. & Leonard, B.E.(1994) Fast-acting antidepressants: can the need be met? CNS Drugs, 2,120-131. ODonnell, I. & Farmer, R. (1995) The l ~ i t a t i o n s of officialsuicide statistics.BritishJournul ofPsychiatry, 166,458-461. Ormel, J., Oldehinkel, T., van Brilman, E. & den Brink, W. (1993) Outcome of depression and anxiety in primary care. Archives of General Psychi at^, 50, 759-766. Pagliaro, L.A. & Pagliaro, A.M. (1995) Abuse potential of antidepressants: does it exist? CNS Drugs, 4,247-252. Pare, C.M.B. (1985) The present status of monoamine oxidase inhibitors. British Journa~of Psychiatry, 146,576-584. Parker, G., Tennant, C. & Blignault, I. (1985) Predicting improvement in patients with nonendogenous depression. British Journal of Psychiatry, 146,132-139. Paykel, E.S. (1994) Relapse, recurrence and in the chronicity in depression. In:Crit~cff~l~ssues ~ e a t m e nof t A~ectiveDisorders (eds S. Z. Langer, N. Brunello, G. Racagni& J. Mendlewicz), pp. 9-20. Karger, Basel. and Paykel, E.S. & Priest, R.G. (1992) Recognition management of depression in general practice: h Journal, 305, consensusstatement. ~ r i t i sMedical 1198-1202. Paykel, E.S., Hollyman, J.A., Freeling, P. & Sedgwick, P. (1988) Predictors of therapeutic benefit from amitrip~linein mild depression: a general practice placebo-controlled trial. Journal of A~ectiveDisorders, 14,8395. Clinical onset of Potter, W.Z. & Manji, H.K. (1994) antidepressant action: implications for new drug 10
development. In: Critical Issues in the Treutment of Affective Disorders (eds S.Z. Langer, N. Bmnello, G. Racagni, J. Mendlewicz), pp. 118” 126. Karger, Basel. Continuation dmg Prien, R.F. & Kupfer, D.J. (1986) therapy for majordepressiveepisodes:how long Journal of Psyshould it be maintained?A m ~ i c a n chiatry, 143,18-23. Priest, R.G. & Schmid-Burgk, W. (1994) Moclobemide in the treatment of depression. Reviews in Contemporary Pharmacotherapy, 5, 35-43. Quality Assurance Project (1983) treatment A outh e for depressive disorders.Australian andNew ~ e a l a n Journal d of Ps y c k ia t ~/ 17,129-146. Quitkin, F.M. (1985)The importance of dosage in prescribing antidepressants. Brit~sh~ournalof p s y c h ia t ~,147, 593-597. Quitkin, EM., Rabkin, J.G., Ross,& Stewart, D. J.W. (1984) Identification of true drug response to antidepressants. Archives of General ~syckiatry, 41,782-786. Quitkin, EM., Stewart, J.W. & McGrath, P.J. et al. (1993) Columbia atypical depression. Brit~sh J o u r ~a l o fp s y c h ia163 t ~, (Suppl), 30-34. Ramana, R., Paykel, E.S., Cooper, Z.,Hayhurst, H., Saxty, M. & Surtees, P.G. (1995) Remission and relapse in major depression:a two-year prospec25, tive follow-up study. Psycho~og~ca~ Me~icine/ 1161-1170. E.,Clary, C.,Fox, I. & Weise, Rickels, K., Schweizer, C. (1994) Nefazodone and imipramine in major depression: a placebo-controlled trial. Br~tish 164,802-805. J~urnalof Psych~atry/ Robinson, D.S., Nies, A., Ravaris, L., Ives, J.O. & Bartlett, D. (1978) Clinical pharmacology of phenelzine. Archives of General Psychiatry, 35, 629-4535> Robinson, D.S., Lerfald, S.C. & Bennett, B. et al. (1991)Continuation and maintenance treatment of major depression with the monoamine a double-blind oxidase i ~ i b i t o phenelzine: r placebo-controlled discontinuation study. Psychopharmacolo~Bulletin, 27,31-39. Rouillon, F., Phillips, R. & Serrurier,D. et al. (1989) Rechutes de depression unipolaire et efficacite p 527-534. ~ ~ , de la maprotiline. ~ ’ ~ n c eXV, Rudorfer, M.V. & Potter, W.Z. (1989) Antidepressants: a comparative review of the clinical pharmacology and therapeutic use of the ’newer’ vs.the ’older’ drugs. Drugs,37,713-738. Schatzberg, A.F. & Nemeroff, C.B. (eds) (1995) The American Psychiatric Press text boo^ of P s y c h o p ~ r m a c o l oAmerican ~. Psychiatric Press, Wa~hington,DC. Shawcross, C.R.& Tyrer, P. (1985)Influenceof per-
~ ~ a ~9: tA en tri ~ e ~ r e s s a ~ t
~e~i~a
sonality on response to monoamine oxidase selective serotonin reuptake inhibitors.Clinical P ~ a r ~ a c o ~ i n e t24,203-220. ics, Journal inhibitors and tricyclic antidepressants. Volz, H.-P., Gleiter, C.H., Struck, M. 19, 557-562. & Miiller, of Psychiatric ~esearch, Shea, MT., Elkin, I. & Imber, S.D. et al. (1992) H.-J. (1995) Brofaromine: insight into the nature Course of depressive symptoms over follow-up: of drug development.CNS Drugs, 3, 1-8. findings from the National Instituteof Mental Wehr,T.A. & Goodwin, EK. (1987) Can antideHealth Treatment of Depression Collaborative pressantscausemaniaandworsenthe Research Program. Archives of General Psychicourse of affective illness? A m e r i c a ~ J o u r ~ a l of Psychiat~,144/1 4 0 3 - q ~ ~ 49, 782-787. Sokolov, S.T.H.& Joffe, R.T. (1995) Practical guide- Wells, K.B., Stewart, A. & Hays, R.D. et al. (1989) lines for combination drug therapy of treatmentpaThe functioning and well-being of depressed resistant depression.CNS Drugs, 4, 341-350. tients: results from the Medical Outcomes Study. Song, F., Freemantle,N. & Sheldon,A. et al. (1993) J o u ~ aofl the American Medical Association, 262, Selective serotonin reuptake inhibitors: meta914-919. analysis of efficacy and acceptability. British Wells, K.B., Burnam, M.A., Rogers, W., Hays, R.& Camp, P. (1992) The course of depression in Medical Journal, 306,683-687. Stewart, J.W., Quitkin, F.M., Liebowitz, M.R. adult outpatients: results from the Medical Outcomes Study. Archives of General Psychiat~,49, McGrath, P.J., Harrison, W.M. & Klein, D.F. (1983) Efficacy of desipramine in depressed out78S794. patients: response according to reserch diagnosWells, K.B., Katon,W., Rogers, B. &Camp, P. (1994) Use of minor tranquilizers and antidepressant tic criteria diagnoses and severity of illness. medications by depressed outpatients: results Archives of General Psychiatry,40,202-207. from the Medical Outcomes Study. American Sulzer, D.L. & Cummings, J.L. (1989) Drug~ournalof Psychiatry,151, 694-700. inducedmania-causativeagents,clinical charactristics and management a retrospective Wernicke, J.F., Dunlop, S.R. & Dornseif, B.E.et al. analysis of the literature. Medical ~ o x i c o land o~ (1988) Low-dose fluoxetine therapy for depresAdverse Drug Experience, 4, 127-143. sion. ~ s y c ~ o p h a r ~ a cBulletin, o l o ~ 24,183-188. Taylor, D. & Lader, M. (1996) Cytochromes and West, E.D.& Dally, PJ. (1959) Effectsof iproniazid psychotropic drug interactions. British Journal of in depressive syndromes. British Medical P s y c h i a t~168,529-532. , Journal, i, 149s-1492. Tarbuck,A.F. & Paykel, E.S. (1995) Effects of Wilde, M.I. & Whittington, R. (1995) Paroxetine: major depression on the cognitive function of a pharmacoeconomic evaluation of its use in younger and older subjects.Psychologic~lMedidepression. P~armacoecono~ics, 8,62-81. Wilde, MI., Plosker, G.L. & Benfield, P. (1993) cine, 25,285-296. of its pharmaTeicher, M.H., Glod, C.A. & Cole, J.O. (1993) Anti- Fluvoxamine: an updated review cology, and therapeutic use in depressive illness. depressant drugs and the emergence of suicidal tendencies. Drug Safety, 8,186-212. Drugs, 4 6 , 8 9 5 - y ~ Workman, E.A.& Short, D.D. (1993) Atypical antiThase, M.E. & Sullivan, L.R. (1995) Relapse and of recurrence of depression. A practical approach depressants vs. imipramine in the treatment major depression: a meta-analysis. Journal of for prevention.CNS Drugs, 4, 261-277. vanHarten, J. (1993)Clinicalpharmacokineticsof Clinical P s y c ~ i a t54,5-12. ~,
11
Antidepressants area chemically diverse group of drugs with a wide range of pharmacological actions. As a consequence, antidepressants are associated with a great manyadverse effects, both predictable and idios~cratic. The severity and frequency with which these adverse effects occur with individualdrugs are important considerations for three pressing reasons. First, antidepressants have broadly equivalent efficacythus drug choice is based largely on differences in theiradverse effects. Second, theadverse effects of these drugs have a major impact onpatient tolerability and therefore adherence to prescribed regimens. Third, some adverse effects arel i f e - t ~ e a t e ~ n g in overdose, making some agents potentialaids to suicide in those patients already relatively more likely to make a suicide attempt. This chapter reviews the adverse effects associated with the four major groups of antidepressants: monoamine oxidase inhibitors antidepressants (TGAs), selective serotonin re-uptake inh ical’ antidepressants. In each case,adverse effects common to all membersof a group will be stressed, but important differences betweenindividual drugs within a group will also beemphasized. Summary tables are included for ease of reference and to allow comparisons between groups to be made. Tables are also used to recommend drugs to be usedin conditions such as epilepsy and pregnancy, where some antidepressants may be strictly contra-indicated. Glinically relevant adverse drug interactions are outlinedand summarized in a table.
In many Western countries, four MAOIs are available: isocarboxazid, phenelzine and tranylc~romine,the ‘traditional’ or non-reversible MAOIs; and moclobemide, a reversible inhibitor of monoamine oxidase (RI Another RIMA, brofaromine, which has broadly similar properties to moclobemide, reached a late stage of development before beingdiscarded. Despite a shared mode of action, traditional MAOIs and RIMAs differ widely in the adverse effects theyproduce and so will be considered separately.
ation
le 10.1 Comparison of adverse effects ofMAOIs. Weight
Drug
+ +
Isocarboxazid Phenelzine Tranylc~prornine Moclobernide
_ .
1 t -
++ +
+ + +
I
_ .
+S
S
The most common adverse effects of traditional MAOIs are hypotension (often causing dizziness), and anticholinergic effects such as dry mouth, constipation, blurred vision and difficulty in micturition (Table 10.1). Both of these effects are dose dependent and tend to lessen in severity with time. The anticholinergic effects are usually less severe than those seen with TGAs (Bass & Kerwin, 1989). Mild sedation occurs with some MAOIsassuch phenelzine but rarely with~anylcypromine, which is morelikely to exhibit a stimulant action and may cause insomnia. This may be because tranylcypromine is structurally similar to amphetamine and is partly metabolized to amphetamine, atleast after an overdose (Dollery, 1991). Other common adverse effects include headache (Zisook, 1984) a & Turnb changes in weight. Isocarboxazid may cause weight loss (Davidson al., 1982). 1982); and substantial weight gain occurs with phenelzine et (Evans Other adverse effects of MAOIs occur much less frequently. ~eurological erse effects such as weakness, numbness and parasthesia may be seen cEvoy, 1995)~leading some authors to suggest that pyridoxine should be given to all patients taking MAOIs (Krishnan, 1995). Peripheral oedema has been associated with all MAOIs although it is thought to be less common with tranylcypromine (Krishnan, 1995); and it appears not to respond to diuretics. Iproniazid is well known tobe hepatotoxic and has now been withdrawn from the market in many countries. Other hydrazine compounds such as phenelzine and isocarboxazid may occasionally- be associated with altered liver function testsbut these changes are usually benign (Zisook, 1984). Progressive necrotizing hepatocellular damage may occur, though this is rare (McEvoy, 1995); regular checksof liver function are therefore essential. Sexual dysfunctionis sometimes observed with MAO1 therapy. symptoms suchas low libido, impotence and delayed or absent orgasm occur as a resultof depression itself, thus making any direct association between a drug and sexual dysfunction difficult to establish. Nevertheless, the number et al., 1982; Zisook, 1984; Jacobson, of reports of sexual dysfunction (e.g. Evans
213
1987) makethe association very likely. The true incidence of MAOI-induced sexual dysfunction virtually is impossible to estimate because reports of such adverse effects are rarely volunteered by patients. An increase in central serotonergic function may be responsible for these adverse effects and the serotonergic antagonist cyproheptadinehas been suggested (Krishnan, 1995) as a suitable remedial treatment. MAOIs may,in rare instances, be implicated as a causeof speech defects. Phenelzine was reported to have caused 'speech blockage' in a 45-year-old woman who had not experienced this problem while takin tranylcy~romine (Goldstein & Goldberg, 1986),and tranylcypromine has been associatedwith severe stuttering in a 14-year-old girl (Duffy, 1994).
In general, any psychotropic drug given regularly for morethan 4-6 weeks should be withdrawn slowly-usually over 2.weeks or more. Thisis particularly important with MAOIs, especially tranylcypromine,with which withdrawal reactions can occur following abrupt discontinua~on(Tyrer, 1984). ~ymptomsinclude headache, nightmares and paraesthesia; occasionally more severe symptoms such as psychosis and delirium result from abrupt withdrawal of an MAO1 (Lejoyeux et al.,1996). Cautious, slow dosetapering is therefore advised.
Toxicity in overdose Overdose with MAOIs may have serious consequences and is sometimes fatal. Symptoms vary but usually include agitation, tachycardia, muscular rigidity and coma (Williamset al.,1995).Hypotension occurswith hydrazine compounds, whereas tranylcypromine causeshypertension (Dollery, 1991). Tranylcypromine overdose may also present with unusual signs such as delirium and thrombocytopenia (Chatterjee& Tosyali, 1995). Signs of overdose may take several hours to developand may persist for up to 2weeks (McEvoy, 1995).Death maybe due to pulmonary and cerebral oedema( ~ i c h t e n w a ~ e r et al.,1995).
Food interactio~s The interactions between MAOISand foodstuffs containing tyramine has been comprehensively reviewed (e.g. Blackwell et al., 1967). Tyramine is produced by degradation of protein and is present in many foods and drinks. Ingested tyramine is normally metabolized by monoamine oxidase and causes few problems. In those taking MAOIs, ingested tyramine isnot deactivated and
may cause hypertension, headache, subarachnoid haemorrhage and death 1976). (the severity of the reaction is related to the amount consumed) (Anon., Debate now centres on the exact frequency and severityclinical or importance of this interaction,Pare (1985),reviewing the conflicting evidence relating to this issue, noted that there had only been of food interactions 17reports involving phenelzine and none of these were fatal. With tranylcypromine, seven deaths had been reported but in only two of these could a definite relationship with dietary indiscretion be established. The author went on to calculate the risk of death due to dietary interaction as one per 14000 patient-years; a small incidence compared to the risk of suicide in untreated depressives. Bass and Kerwin(1989) concurred, statingthat 'fears about the interaction have been grossly exaggerated'. The rangeof foodstuffs forbidden to patients taking MAOIs may also have (1984) averred that only four foods been exaggerated. Sullivan and Shulman should be absolutely ~ r o h i b i t mature ~ ~ : cheese, pickled fish, concentrated yeast extracts, and (the rarely eaten) broad beanpods; other 'forbidden' foods could be taken in true moderation. This conclusion was based on a survey and literature review. Using measurements of tyramine content of over loo foodstuffs, Shulman et al. (1989) came to very similar conclusions, adding only sauerkraut and (possibly) aged meats to the realistic prohibited list. Drug i ~ f e r a c f i o ~ s
MAOIs are involved in a large number of clinically important drug interactions. Serious interactions (usually hypertensive crises) may occur with pethidine, laevodopa, many antidepressants and a variety of sympathomimetic agents such as amphetamine and ephedrine. A comprehensivelist of known drug interactions can be found in the relevant manufacturer's data sheet or package insert, and in Bazire (1996). More recently reported interactions include that between isocarboxazid and venlafaxine (Klysner et al., 1995) and between phenelzine and venlafaxine (Heisler et al., 1996).Serotonin syndrome (diaphoresis, restlessness, rigidity and changes in mental state) appeared to have occurred in both instances. Use in specialpafienfgroups
There are few publisheddata on the useof MAOIs in special patient groups. This is undoubtedly becauseof their relatively early introduction. ~ p ~ l e pMAOIs s ~ . are thought not to reduce seizure threshold (Skowron & Stimmel, 1992; Stimmel & Dopheide, 1996).One study found that no seizures
occurred in 132 patients treated with MAQIs (Rabkin et al., 1984). Seizures may occur as a result of hypertensive crisis and myoclonic movements are sometimes seen in patients on MAOIs. Neitherof these effects represents a lowering of the seizure threshold. Overall, MAQIs appearbeto safe to use in epilepsy although there are few conclusive data to confirm this. ~ r e g ~ aThe ~ c danger ~ . of hypertensive crisis makes the of use nancy inadvisable. In addition, MAQIs have been suggested as human teratogens (Miller,1.991) and phenelzineis teratogenic in animals (Poulson& Robson, 1964). ~ r e a s ~ e ePublished ~ ~ ~ g . data onthe use of MAQIs in lactation arevery limited. It is thought that small amounts of MAO1 may be excretedinto breast milk (Rowan, 1976) but little more is known. AQIs should not, therefore, be given to breastfeeding mothers.
Adverse efects
oclobernide’s pharmacological action is limited to the reversible~ i b i t i o n of monoamine oxidase-A. Becauseof this, adverse effects are few and infrequent: in clinical trials dropouts due to side-effects were uncommon and only nausea occurred with a higher frequency (10%) than placebo (Fitton et al., 1992). Other adverse effects observed in clinical practice include insomnia, dizziness and headache; activation (e.g. agitation and restlessness) also seems to occur in some patients taking moclobemide, although the incidence of all of these events is around 10% or less (Baldwin & Rudge, 1994). Confusion and asymptomatic rises in liver enzymesbemay associated with moclobemide but these are rare (Freeman, 1993). Unlike many other antidepressants, moclobemide appears not to impair psychomotor performance (Tiller, 1990) or to cause sexual dysfunction (Philipp et al., 1993). ~oclobemide,perhaps because of its short duration of action, does not suppress REM sleep, while it lengthens overal leep time in depressed patients (Monti, 1989).This is in contrastto standard QIs, which are known to abolish sleep and cause profound rebound on withdrawal. Brofaromine is also short-acting butit does suppress REM sleep (Steigeret al., 1987). Clinical experience with moclobemide indicates that tyramine reactions (including hypertensive crisis) may occur. Coutler and Pillans (1995) have reported six cases of hypertension related to moclobemide use and a controlled risein blood pressure was induced by the combinationof Bovril (containet al., 1995). ing tyramine) and moclobemide (Taylor
C ~ a ~TO: ~ Adverse e r e ~ e c of~ ~s e d i c a ~ i o ~
~ i t h d r a w areactions l have not been reported followingabrupt withdrawal of moclobemide. Slowwithdrawal is, however,advised. Overdose
If taken alone, moclobemideappears to be safein overdose (Freeman, 1993). Symptoms include drowsiness and disorientation (Fittonet al., 1992). Deaths have occurred from serotonin syndrome after overdoseof moclobemide and an inhibitor of serotonin re-uptake (Neuvonen et al., 1993). Drug i ~ t e ~ a c t i o ~ s
Moclobemide is involved in few clinically important drug interactions. et al., 1990) and so Cimetidine decreases moclobemide metabolism (Zimmer the doseof moclobemide should be halved. Moclobemide should not be given with SSHs because of the danger of serotonin syndrome and because SSHs and moclobemide are poorly tolerated when co-administered. Theoretical interactions with opiates such as pethidine (meperidine), dextromethorphan and dextropropoxyphene should be noted. This caution is based on experience with other, standard MAOIs but may have relevance since there has been one case report of serotonin syndrome occurring with pethidine and moclobemide (Gillman, 1995). Moclobemide may safely be given with many sympathomimetics and TCAs (Fitton et al., 1992). S~ecial~ a t i e ~ t
grou~s
~ ~ ~ Moclobemide l ~ s y , appears to be safein epilepsy: no seizures have been reported, even following overdoses of up to 7"G (Myrenfors et al., 1993). ~ r e g ~ aData ~ c ~on. the use of moclobemide in pregnancy are limited. Because of this and the small risk of hypertension, moclobemide should not be used in pregnancy. ~ r e a s ~ e e ~Moclobemide i~g. is excreted only in small amounts into breast milk (Pons et al., 1990) but is not considered to be the drug of first choice in postpartum depression (Duncan&:Taylor, 1995b).
C ~ a ~ tIO: e rAd~erseeffects of ~ e d i ~ a t i o ~ Table 10.2 Comparison of adverse effectsof tricyclic drugs.
olinergic Cardiotoxicity Sedation
Drug Amitriptyline Amoxapine Clomipramine Desipramine Dothiepin Doxepin Imipramine Lofepramine ++ Nortriptyline Protriptyline Trimipramine
+++
++
++ ++ + +++ +++ ++ + + ++++
+++ + +++ ++ +++
++ +++
+ ++ +++ +++
+++ ++ ++ ++ ++ +++ +
++ ++
Tdcyclic antide~ressants
The adverse effects of TCAs are well known and have been widely described over thepast 30 years orso (Table 10.2). The majorityof these adverse effects can be predicted from the diverse pharmacology demonstrated by this group of drugs: they are antagonists at cholinergic, histaminic and alpha-adrenergic receptors and have a quinidine-like effect on cardiac tissue. Thus, constipation, sedation, hypotension and cardiac conduction changes, respectively, are commonly found with tricyclic drugs. Differences between drugs are largely due to minor differences in chemical structure. Sedation is caused by all tricyclic drugs, with the notable and sole exception of protriptyline, which has stimulant properties. In general, tertiary amine drugs suchas amitriptyline, dothiepin, doxepin and trimipramine are the mos sedative. Secondary amines such as desipramine and nortriptyline are the least sedative. Lofepramine is metabolized to desipramine and is also only mildly sedative. Sedation is most profound when starting therapy and tends to wear off over a period of several weeks. This property may be considered advantageous in patients with prominent anxiety and insomnia. However, because all tricyclics have both long plasma half-lives and active metabolites, the sedation caused by a single night-time dose is often prolonged. Many patients experience difficulty in awakening and may be drowsy throughout the morning and even in the afternoon. This obviously has important implications for those who need to drive, operate machinery or work on skilled tasks. In the longer term, subjective feelingsof sedation resolve but reaction times may continueto be extended. Anticholinergic effects are common and include constipation, dry mouth, blurred vision, tachycardia and urinary hesitancy.No tricyclic drug is free from these adverse effects although secondary amines are relatively less likely to give rise to profound symptoms. Because of these effects, tricyclics are I
218
C ~ ~ ~TO:t Adverse e r effects of ~ e d i c ~ t i o ~ contraindicated in narrow-angle glaucomaand in patients with a historyof urinary retention. Anticholinergic side-effects are said to wear off in time,but many patients report problems with constipation and dry mouth after months or yearsof t r e a ~ e n t . Cardiovascular effects are also common with tricyclics. Blockade of aladrenoreceptors inducespostural hypotension which is often severe enough to engender dizziness and even falls (Ray, 1992). Secondary amines are again less likely to cause,severe problems and there is some evidence to suggest that nortriptyline rarely causes hypotension even in the elderly (KraghSorensen, 1982). Tricyclic drugs have complex effects on cardiac tissue: their quinidine-like membrane-stabili~ingactivity provokes a slowing in impulse conduction which may lead to a benign lengthening ofPR, QRS and QTC intervals, amongst others.Tricyclics maysuppress premature contractions and have a beneficial effect in those predisposed to certain cardiac dysrhythmias. However, their quinidine-like effect representsthe most serious aspect of their overall toxicity: tricyclics may precipitate, especially in overdose, bundle branch block, completeheart block and a bewilderingarray of other serious dysrhythmias. The cardiac effect of tricyclics has been comprehensively reviewed by ~ a r r i n g t o net al. (1989). There are few, cogent data comparing the pro-arrhythmic potential of tricyclic antidepressants and the relative danger of sudden death with their use. Perhapsthe best indication of relative cardiac toxicity is provided by overdose data. This is discussed later. A reduction in cardiac contractility has been reported to occur with tricyclic drugs. However, evidence suggesting a deleterious effect (e.g. precipitating or worsening cardiac failure) is equivocal. It is now generally that TCAs held may safely be used in cardiac failure. However, since the introduction of newer antidepressants such as the SSRIs, the use of tricyclics in heart failure has become unnecessary. Tricyclics are also associated with a wide range of other adverse effects. In commonwith other antidepressants, they can precipitate mania in susceptible individuals. They can also cause hyponatraemia (Committee on Safetyof Medicines, 1994)and sexual dysfunction, which is probably most likely to occur when a potent inhibitor of serotonin re-uptake such as clomipramine is used (Hawley & Smith, 1994). Tricyclics are well known to cause weight gain and increased appetite, possibly by increasing insulin activity (Winston & McCann, 1972). Tremor is sometimes reported,although frank dyskinesias are much less common. Rare, idiosyncratic adverse effects include hepatitis, rashes and various blood dyscrasias. ~ i t ~ ~Anxiety, ~ ~ ~headache a l . and tremor are seen after the discontinuation
of a short course of TCAs and accompanied by increased production of MHPG (Charney et al., 1984). Much more serious withdrawal symptoms are seen after several yearsof TCA treatment: symptoms include confusion, incontinence, nausea and convulsions may be due to supersensitive muscarinic receptors (Dilsaver& Greden, 1984). In addition, there are a few case reports (e.g. van Sweden, 1988) describing the emergence of cardiac arrhythmias after TCA discontinuation.
~~te~action. Most interactions with which tricyclic drugs are involved are read ily predicted from their pharmacology. Thus, tricyclics show additive pharma codynamic interactions with sedatives (including alcohol), anticholinergics, hypotensives and antiarrhythmics. With MAOIs, hypertension and central nervous system (CNS) excitation may occur. The metabolism of tricyclics is inhibitedby disulfiram, calcium channel blockers, cimetidine, phenothiazines, and most SSRIs. An increased frequency and severity of adverse effects may result. Carbamazepine and phenytoin may increase the rate of tricyclic metabolism and reduce efficacy.
Toxicity in overdose. Tricyclic antidepressants, with the exception of lofepramine, may cause deathif taken in large doses such as those takenin suicide bids. Their toxicity is probably a consequence of their arrhythmogenic potential, but anticholinergic effects and profound GNS depression may also play a part along with rarer complications such as respiratory difficulties and acid-base disturbances (Pinder, 1990). Convulsions sometimes occur, especially with amoxapine and maprotiline. Relating the number of deaths by overdose to the number of prescriptions issued for a drug gives some idea of relative toxicities, although other factors (variability in use of particular drugs in patient types; inherent of ability drugs to promote or prevent suicide attempts) are also important and difficult to evaluate. It is now accepted that some tricyclics, such as desipramine, dothie and amitriptyline, have greater inherent toxicity and are relatively more dan gerous in overdose (Cassidy& Henry, 1987; Poweret al., 1995). Lofepramine, in stark contrast, seems be to safe in overdose despite desipramine being the drug’s major metabolite. The reasons for this are not clear. The profound and unarguable toxicity of tricyclic compoundsis seen by some as sufficient reason to prevent their use now that safer alternatives are available. After all, why should a patient at risk of suicide be prescribed the means with whichto accomplish the task? Others argue that any patient intent on suicide will find the means do sotowhether or not they are prescribed a tricyclic, and that tricyclics are well knownbetocheap and effective. Individual prescribers need to come to their own conclusions at this dilemma. Lofepramine, a safe tricyclic drug, stands out as a suitable compromise.
S
~ ~ ~ ~Tricyclics e ~ s have y . been associated with seizures both in clinical and use following overdose, thus indicating a t~eshold-loweringeffect. Firm, conclusive data on the relative incidence of seizures are lacking (Duncan & Taylor, )but amoxapine and maprotiline (essentially a tricyclic) probably have eatest effect on seizure ~ r e s h o l d(Wedin et al., 1986). Where possible, tricyclic drugs should not be used in patients with seizure disorders. ~ ~ e ~Tricyclics ~ ~ ~have c been y . widely used in pregnancy and, despite a number of case reports suggestinghuman teratogenicity, are considered the safest antidepressant drugs to use in pregnancy: extensive epidemiological ies have failed to show an association with birth defects (RobinsonG.E. aldson, 1995).Nevertheless, theyshould not be used in gnancy unless absolutely necessary. Moreover, tricyclics ithdrawn before delivery so as to avoid withdrawal
ecause of the high incidence of postpartum depression and the long historyof TCA use, thesedrugs have been relatively well studied in breastfeeding mothers.As expected, tricyclics are excreted in breast milkand infants are commonly exposed tothe drug taken by the mother. However, can be detected in the infant's serum and where detectable, & Taylor, 1995b).Adverse rations are extremely low (Duncan events are infre~uentand development doesnot seem to be impaired (Buist Janson, 1995). Tricyclics are therefore recommended in ~reastfeeding mothers. Doxepin, however, should be avoided because of its association with infant respiratory depression (Matheson et al., 1985).
Since the introduction of fluvoxamine in the mid-lg8os, the group of drugs S has gradually grown in size. TheSSRIs available in most 1996are: citalopram, fluoxetine, fluvoxamine, paroxetine growing popularity can be seen to be a resultof their perceived low toxicity, relatively good tolerability and generally slight adverse effects. Comparative efficacy and tolerability data on SSRIs and TCAs ewed inChapter 9. most common and clinically important adverse effects of SSRIs are nausea, i n s o ~ i and a a collectionof 'activation' symptoms usually described as 'ne~ousness','a itation' or 'anxiety'(Table 10.3). The precise incidence of these effects is difficult to establish. Variation in reported incidence can be
Table 10.3 Comparison of adverse effects of SSKIs. Nausea
Drug Citalopram Fluoxetine Fluvoxamine Paroxetine Sertraline
Anti~~olinergic Sedation effects
++ ++ +++ ++ ++
-
+
_ .
-
EPSE
-
+ +
-
+/+/+/-
+
+/-
accounted for by the way in which information was gathered (for example, by self-reporting or from direct questioning) and by different dosing schedules used in different studies and reviews. In addition, adverse effects are generally more common at the beginningof treatment, so making the timeof reporting most important. This imprecision makes difficult any meaningful comparison of the incidenceof common adverse effects with individual drugs. Data on the use of citalopram in short-term trials (Lundbeck, 1995) indiin 21.4% of patients and insomniain 18.8%, although cate that nausea occurs in the latter case this figure did not differ significantly from placebo. An earlier meta-analysis (Dencker& Hopfner Petersen, 1989) indicated that nausea and/or vomiting occurred20% inof patients taking citalopram. Fluoxetine has been reported to cause nausea in 22.9% of patients, insornnia in 15.5% and nervousness in 14.5% (Pande & Sayler, 1993). In contrast, an early estimate of the frequency of adverse effects occurring with fluvoxamine (Benfield & Ward, 1986) listed nauseaas occurring in 37% of patients, agitation in 16Y0 and insomniain 15%. A later review (Wagneret d., 1992)~ which included data on more than 24000 patients, found the incidence of nausea with fluvoxamine to be 15.7%, and insomnia and agitation to occur in less than 5%. Many clinicians have noted that fluvoxamine frequently causes an unacceptable degree of nausea and its use has declined substantiallyin the UK. With paroxetine, the incidence of nausea hasbeen quoted tobe 23% (Caley & Weber, 1993)~ although insomnia and nervousness appear be to uncommon (less than 15%). With sertraline, nausea occurs in 27% of patients, insomniain 14% and anxiety in8% (Grimsley & Jam, 1992) The incidence figures quoted here, given the caveats outlined, give the clinician little guidance in drug choice. Nevertheless,it is important to note that individual experience and tolerability vary widely and that nausea, ins o d a and nervousness all tend to decline in severity as treatment progresses. The short-term(1-3 weeks) useof benzodiazepinesis recommended for those who suffer severe activation. A drug holiday or the use of cisapride (Bergeron & Blier, 1994)~ but not withP4503A-4 inhibitor fluoxetine, may help reduce the toactivating andto severity of nausea. WhilstSSNs are accepted generally be cause insomnia in some, others may experience symptoms of sedation such 222
C ~ ~10:~Adverse ~ e efects r of ~ e d i c ~ ~ i u as somnolence or drowsiness. This adverse effect is most commonly seen with paroxetine which has been reported to cause sedation, particularlywith regard to driving ability. SSRIs are dispensed in many countries with a label cautioning against driving if affected. However, studies appear to indicate that SSRIs, including paroxetine,do not adversely affect driving performance (North West Regional Drug Information Service, 1995). All SSRIs can cause sexual dysfunction. Men may suffer a lowering of libido, and may have difficulty maintaining an erection. Orgasm may be delayed or be absent.Women may also suffer lack of libido and anorgasmia. Again, an accurate assessmentof comparative incidence cannot be made, but all SSRIs have been reported to cause sexual dysfunction. This effecta may be result of enhanced serotonergic activity: SSRI-induced anorgasmiaappears to be ameliorated by the serotonin antagonist cyproheptadine(Arnott & Nutt, 1994; Lauerma, 1996) SSRIs are sometimes, albeit infrequently, associated with dry mouth and constipation, perhaps reflecting their relative selectivity for the serotonergic system compared with tricyclic antidepressants. Paroxetine is most comrnonly associated with these adverse effects (Grimsley & Jann, 1992) and, indeed, has the greatest anticholinergic activity (Hyttel, 1994). shouldItbe notedthat dry mouth and constipation appear to occur with all SSRIs, despite very poor anticholinergic activity with some drugs. This may indicate that alterationsin serotonin function are at least in part responsible. Ex~apyramidaladverse effects such as dystonia and parkinsonism are rarely associated with the useof SSRIs (Arya, 1994). Thereasons for theoccurrence of these effects are unclear(although the serotonin and dopamine system are known to be linked) and the true incidence difficult to establish. Paroxetine is possibly the most likely of the SSRIs to cause extrapyramidal effects (Consumers' Association, 1993) but there are few comparativedata to confirm this. Hyponatraemia, with concomitant inappropriate secretion of antidiuretic hormone, isa rare but well recognizedadverse effect of many antidepressant drugs including theSSRIs (Thomas & Verbalis, 1995)."his adverse effect seems only to occur in elderly patients and is usually benign (e.g. dizziness and weakness). Occasionally, disorientation and delirium have been reported (Thornton & Resch, 1995). Sodium levelsreturn to normal afterwithdrawal of the offendingdrug. In recent years, debate has raged over the association between the of use SSRIs and the occurrence of suicide ideation and suicide attempts. A metaanalysis (Beasley et al., 1991) has shown that fluoxetine was not associated with an increase in suicide and it is now widely acceptedthat SSIUs do not provoke suicidal ideation or suicide (Montgomery 1992). Unlike other groups of antidepressants, SSRIs do not cause weight gain.
Fluoxethe sometirnes causes weight loss, especially in those who are over Other adverse effects linked to the use of SS S include transient headache, diarrhoea and sweating, and rare events such asrash (especially with fluoxetine) and vasculitis. Inaddition, hair losshas been associatedwith the use of fluoxetine (Seifritzet al., 1995) and sertraline (~ourgeois,1996). ~ i t ~ ~ r aAbrupt ~ a l .withdrawal of an SSRI may bring about a withdrawal syndrome. Symptoms include delirium, malaise, incoordination, nausea, fatigue and dizziness (Lazowick& Levin, 1995). Skindisorders may occur on abrupt withdrawal of sertraline (Leiter et al., 1995)and extrapyramidalsymptoms have been reported following withdrawal of fluoxetine (Stou~des,1991). A flu-likewithdrawal syndrome has been described in patients withdrawing from paroxetine (Lejoyeux et al., 1996),in some cases despite slow withdrawal over 7-10 days. Some debate surrounds the issueof which SSRT is most likely to cause a withdrawal syndrome. Fluoxetine, becauseof its long plasma half-life,was originally thought not to induce withdrawal symptoms but several (sometimes delayed) withdrawal reactions have been described (e.g. Einbinder, 1995). Paroxetineis widely believedto be particularly liable to causewitha drawal reaction. All SSRIs should therefore be withdrawn cautiously and slowly; usually over a period of 2 weeks or more.
Overdose. It is wellknown that SSRIs are safe in overdose.T h i s is probably because of their specific mode of action and their lack of effect on cardiac tissue. Remarkably fewdeaths have beenreported following overdoseof an SSH alone. Symptomsof overdose include vomiting, agitation, seizures and, less often, loss of consciousness (Anon.,1993;Grimsley 6rJam, 1992; Baldwin & Johnson, 1995). SSRIs may be dangerous if taken in overdose with other drugs which potentiate the serotonergic system since life-threatening seroto syndrome may result (Neuvonen et al., 1993). More recently,the safety of citalopram in overdose has been brought into question. ijstrom et al. (1996)have reported six deaths following citalopram overdose where,in each case, high plasma levels of citalopram appeared to be the cause of death. A possible effect on cardiac interval was suggested as the primary toxic mechanism. ~ ~ t e ~ a c t Knowledge io~s. of the range of drug interactions in which SSHs are involved has grown substantially since theirintroduction and it is now well recognized that they havethe potential to cause serious adverse interactions. SSRls are inhibitorsof a variety of hepatic cytochromeP450 (CYP) enzymes: most SSRIs inhibit GYP2D6, fluvoxamine inhibits CY PIA^ and fluoxetine inhibits CYP34. As a consequenceof this activity, allSSHs (with the possi-
ble exception of citalopram) may increase plasmalevels of co-administered tricyclic antidepressants and engender serious adverse effects (Taylor, 1995). Fluvoxamine may raise levels of caffeine and theophylline, and fluoxetine has been reported to increase levels of clozapine (and possibly other antipsychotics), cyclosporin and terfenadine (Taylor & Lader, 1996). Fluvoxamine, sertraline and paroxetine also increase clozapine plasma concentrations (~enturrinoet al., 1996)~but the author’s experience with citalopram suggests no important altering of clozapine plasma levels. SS increase levels of some anticonvulsants, propranolol a Pharmacodynamic interactions may occur with tryptophan. Interactions may also occur following discontinuationof an §SRI, especially fluoxetine, whichhas a long plasma half-life andan active metabolite with an even longer half-life.If stopping treatment with anSSRI, manufacturer’s information should be consulted before starting any drug known to olved in serious pharmacod~amicinteractions with . ,1995). Fatal serotonin syndrome has been reported when 01were given together, and so MAOIs andSSRIs should never be CO rescribed. ~teractionsmay occur after discontinuationof either ne has a longhalf-life; MAOIs have a long durationof e advised to consult manufacturers’ literature before changing from anSSRI to an MAOI, orvice versa. S
s probably do not alter seizure threshold to a clinically important extent,alth~ughthis is the subjectof some debate. They are thoughtbeto
safe to use in epilepsy (Duncan & Taylor, 1995a). Interactions with some been reported anticonvulsants (notably arb amaze pine and phenytoin) have and so cautian is advised.
~ ~ e g ~ Available a ~ c ~ data , on the use of SSNs in pregnancy are limited, but
the drugs seem not to be teratogenic. However, on account of the lackof data and the absence of large prospective studies, and because fluoxetine has been linked withan increased rateof miscarriage (Pastuszak, 1993), SSRIs should usually be avoided in pregnancy. ~ ~ e a s ~ eSSRIs e ~ ~are ~ li~id-soluble g . drugs which would be expected to be excreted in breast milk, an expectation confirmed for fluoxetine (Isenberg, 1990). Neither the acute effects on the neonate nor the long-term effects on SSRIs should thereforebe avoided in breastfeeding development are known. mothers (Duncan
~ ~ a ~10:t Adverse e r efjects of ~edication
Antidepressant drugs which do not fall readily into the classes so far described are usually referredto as 'atypical'. They mayhave different chemicalstructures (e.g. non-tricyclic) or different modes of action (e.g. dopamine reuptake biti ion) or both, and have little in common with each other. Sice this group of drugs isso heterogeneous, each compound will be consideRd separately.
Bupropion isan inhibitor of dopamine re-uptake which is widely used in the USA. It has few, if any, sedative, anticholinergic, hypotensive or cardiotoxic properties (Rudorfer & Potter, 1989). Most of bupropion's adverse effects are related to overstimulation of dopaminergic function: insomnia, agitation, nausea, weight loss and psychosis can occur (Reynolds, 1996). Bupropion should be withdrawn slowly to avoid precipitating a withdrawal reaction. Acute toxicityin overdose is less profound than that seen with tricyclics (Hayes & Kristoff, 1986). Bupropion should not be given with MAOIs, dopamine precursors or agonists (e.g. those used in Parkinson's disease) or to those at risk of seizures sinceit decreases seizure threshold in a dosedependent fashion (Reynolds, 1996). There are few safety data relating to the useof bupropion in pregnancyand so its use should be avoided. Bupropion accumulates in breast milk but infant exposure is minimal (Briggs et al., 1993). Further studies are needed before bupropion can be recommended for breastfeeding mothers.
Viloxazine is an antidepressant with an unknown mode of action which is marketed in theUK and Europe. It is not widely used, probably because of doubts over its efficacy (Bazire, 1996). Naysea, vomitingand headache (sometimes severe or migraine-like) are commonly reported adverse effects (Reynolds, 1996).
Mianserin isa tetracyclic compound which is an antagonistat alphaz,5-HTIA, 5-HT2and 5-HT3receptors. It is not currently available in USA. the Adverse effects of mianserin are usuallymild-only sedation is frequentlyreported, albeit with impairment of psychomotor function (Peet & Behagel, 1978). Blood dyscrasias have also been described (someof them fatal, particularly in the elderly) and so monthly blood monitoring is recommended when mianserin 226
is started( C o ~ i t t eon e Safetyof Medicines, 1985). Mianserin is safe in overdose (Chandet al., 1981) and appears to lack cardiotoxicity. Convulsions have been reported to occur in patients taking mianserin, but a true association has not been shown and its effect on seizure threshold is uncertain (Edwards Glen-Bott, 1983).
Venlafaxine is a relatively new antidepressant which inhibits the re-uptake of serotonin and noradrenalinebut, unlike tricyclics, haslittle or no affinity for alpha adrenoceptors, cholinergic or histaminic receptors. Nevertheless, its adverse effect profile includes nausea, dry mouth, headache, sedation and constipation. Nausea has been reported to occur in 44% of patients (Lecableet al., 1995), although combining data from several studies indicates that the true incidence is nearer30% and thatit occurs mainly in thefirst 2. weeks of treatment (Danjou& Hackett, 1995). Cisapride appears to ameliorate nausea caused by venlafaxine (Russel, 1996). Dry mouth, constipation, sedation, insomnia, headache and nervousness all occur in more than 15% of patients & Perry, 1994) and are very probably treatmentgiven venlafaxine (Ellingrod related. Perhaps as a result, the rate of withdrawal from treatment because et al., of adverse effects is similar to that seen with imipramine (Shrivastava 1994). Elevation of blood pressure may occur at higher doses of venlafaxine (>aoomg daily) (Danjou & Hackett, 1995). Close monitoring of blood pressure is essential for those on higher doses and venlafaxine should not be given to those withh~ertension.A s with other antidepressants, abrupt cesA syndrome of sation of venlafaxine may precipitate a withdrawal reaction. headache, nausea,a b d o ~ adistension l and fatigue has been described (Farah & Lauer, 1996). Venlafaxine does not appear to inhibit strongly any of the & PerT, 1994). cytochrome enzymes and is not highly protein-bound (Ellingrod Few adverse drug interactions are therefore predicted. However, venlafaxine is metabolized by cytochrome P45oIID6, an enzyme which is strongly inhibited by a numberof drugs including fluoxetine. Venlafaxine appears to have little, if any, cardiotoxicity. Limited data indicate that venlafaxine is safe in overdose, although seizures have been reported (Woo et al., 1995). Venlafaxine should not be used in epilepsy, pre~nancyor breastfeeding because of the absenceof convincing safety data. Milnacipran has a similar action to venlafaxine but little is reported of its adverse effect profile in English language literature. It is marketed in Portugal and France.
ine is a noveltidep press ant with unusual it is an antagonist at presynaptic alpha~receptorsa receptors. These actions result in enhanced seroto Tl,receptors) and this is though activity. The adverse effects of zapine have been avies & Wilde, 1996 ain occurq ~ tfreq~ently e ( 1 ~ ~ of 0 % patients) but th to be better tolerated than a ~ ~ i p t y l i n e . P r edata ~ i ~S a r y is safe in overdose (Hoes & Zeijpveld, 1996) and seems This latter observationneedsmoreconclusive effect on seizurethId. examination since ceptor antagonists are nown to reduce seizure threshold.
azodone is a weak inhibitor of serotonin re-uptake and an an T andalphal adrenoreceptors.It has no antichol y predicts adverse effects suchas sedation and a low .Indeed, sedation and dizziness(due to frequently reported adverse effects ( rogden et al., 1981). S cularly c o ~ o nespecially , at higher doses. Trazodone appears to lack the profoundcardiotoxicit~of but d y s r h y t ~ ahave s occasion~lly been reported (Schatzbe is not usually fatal. verdose with trazodone alone Some sexual problems are associated with theofuse trazodone. Priapism is a rare but frequently reported proble ompson et al., 1990) and increased libidoandspontaneousorgasm(Purcurye,1995)havebeendescribed. Abrupt or rapidly tapered withdrawal of trazodone may give rise to withdrawal symptoms which include malaise, myalgia, nausea and restless legs drome (Otaniet al., 1994). Very slow withdrawal is therefore reco .over 1 month or more, where possible. Trazodone exhibits additive pharmacod~amicinteractions with other sedatives inch ents, Combination with SS because of the riskof serotonin s ~ d r o m e , Trazodone probably does not lower seizure threshold ( ~ e Convulsions have been reported in overdose in one patient,but were perhaps et al., 1992). Data on the use of ~azodone related toh ~ o n a t r a e (~alestrieri ~a re~ancy and lactation are limited.
~ o ~ ~ a rof iadverse s o ~ effects of a
ical a ~ t i d e ~ r e s s a ~ t s .
overdose Sedation Hypotension Anticholiner in Toxicity Drug Bupropion Mianserin Mirtazapine Nefazodone Trazodone Venlafaxine Viloxazine
+ _ .
_ .
+
_ I
_ I
-
"+ ++ + +++ ++ +
-
+ ++
-
Anon. (1976)MAOI's and foodfact and fiction. zodone: adverse drug reaction Adverse Drug~ e a c t i BulZeti~, o~ 5 ianAdverseDrug ~eac~ion Anon.(1993)Selectiveserotoninre-uptake inhibitors for depression? Drug and~ ~ e r u ~ e u ~Arnott, i c s S. CSS Nutt, D.(1994) Successful treat~ent of fluvoxamine- induce^ anorgasmia b Bulletin, 31,57-59. Anon. (1995) Can SSRI's affect driving ability? c~roheptadine. ~ r i t i~s ~o ~ ro ~ Pa ls ~ c x~ i a ~ ~ , Materia Medica,Nov. 1995. NorthWest Regional 838-839. Drug Infor~ationService, Manchester,UK. Arya, D.K. (1994) Extrapyrami~als y ~ p t owith ~s
ER. et al. (1996) Serum levels of clozapine and norclozapine in patients treated with selective American Journal serotonin reuptake inhibitors. Baldwin, D. &Johnson, EN. (1995) Tolerability and of P ~ s c h i a t153,820-822. ~, safety of citalopram. Review in Conte~porary Chand, S., Crome, P. & Dawling, S. (1981) One Pharmacotherap~,6,315-325. hundred cases of acute intoxication with Baldwin, D. & Rudge, S. (1994) Tolerability of mianserin hydrochloride. Pharmacopsychiatria, in Contemporary moclobemide. Reviews Phar~acotherapy,5,57-65. 14115-17. Balestrieri, G., Cerudelli,B., Ciaccio, S. & Rizzoni, Charney, D.S., Heninger, G.R. & Sternberg, D.E. (1984) The effect of mianserin on alphaz D.(1992) Hyponatraemia and seizure due to adrenergic receptor Eunction in depressed paoverdose of trazodone. gritish Medical Journal, tients. British Journal of Psychiatry,144,407-416. 304,686. M.C. (1995) Chatterjee, A. & Tosyali, Bass, C. & Kerwin, R. (1989) Rediscovering Thrombocytopenia and delirium associated monoamine oxidase inhibitors. Bri~ishMedical with tranylcypromine overdose. J o u ~ aofl Clini~ournal,298,345-346. cal Psychopharmacolo~, 15,143-q4. Bazire, S. (1996) The Psychotropic Drug Directory, Ciraulo, D.A., Shader, R.I., Greenblatt, D.J. & Quay Books, Salisbury,UK. Creehan, W. (eds)(1995)Drug ~nteractions in PsyBeasley, C.M., Jr, Dornseif, B.E. & Bosomworth, chiatry, 2nd edn. Williams & Wilkins, Baltimore. J.C. et al. (1991) Fluoxetine and suicide:a metaanalysis of controlled trials of treatment for de- Committee on Safety of Medicines (1985)Adverse British MedicalJourreactions to antidepressants. pression. British Medical Journal, 303,685492. nal, 291,1638. Benfield, P.B. & Ward, A. (1986) Fluvoxamine: a of Medicines (1994)Antidereview of its pharmacodynamic and pharma- Committee on Safety CurrentProbcokinetic properties and therapeutic efficacy pressant-induced hyponatraemia. lems, 20,5-6. in depressiveillness. Drugs, 32,313-334. Bergeron, R. & Blier, P. (1994) Cisapride for the Consumers’ Association (1993) Selective serotonin treatment of nausea produced by selective re-uptake inhibitors for depression? Drug and TherapeuticsBulletin, 31,57-59. AmericanJournal serotonin re-uptake inhibitors. Coulter, D.M. & Pillaus, P.I. (1995)Hypertension of Psychiatry, 151,1084-1086. J. &Taylor, D. (1967) with moclobemide. Lancet, 346,1032. Blackwell, B., Marley, E., Price, Hypertensive interactions between monoamine Danjou, P. & Hackett, D. (1995) Safeljr and tolerance profile of ven1aFax:ine. ~nternationalJouroxidase inhibitors and foodstuffs. Bri~ishJournal of P s ~ c h o p h a r ~ a c o1 l o0~ (Suppl, , 2), 15-20. nal of Psychiatry,113,349-365. Bourgeois, J.A. (1996) Two cases of hair loss after Davidson, J. & Turnbull, C.(1982) Lossof appetite and weight associated with the monoamine sertraline use.~ournalof Clinical Psychophar~aoxidase inhibitor isocarboxazid. J o u ~ aofl CliniCOlOgY, 16, 91-92. 2,263-266. cal Psyc~ophar~acologY, Briggs, G.G., Samsck, J.H., Ambrose, P.J. & Schroeder, D.H. (1993) Excretionof bupropion Davies, R. & Wilde, M. (1996) Mirtazapine: a review of its pharmacology and therapeutic in breast milk. Annals of Phar~acotherapy,27, potential in the managementof major depres431-433. sion. CNS Drugs, 5,389-402. Brogden, RN., Heel, R.C., Speight,T.M. & Avery, (1989)SideG.S. (1981) Trazodone: a review of its pharma- Dencker, S.J.& Hopfner Petersen, H.E. cological properties and therapeutic use in de- efjcect profile of citalopra~and referencea n t id ~r e s 22nd sants in depression. Paper presented at the pression andanxiety. Drugs, 21,401-429. Buist, A.& Janson, H. (1995) Effect of exposure to Norwegian Congress on Psychiatry, Reykjavik, dothiepin and nordothiaden in breast milk on 10-13 August 1988,31-42. of P s y c h i a ~ ~ ,Dilsaver,S.C. & Greden, J.F. (1984)Antidepressantchild development. Bri~ish ~ourna2 induced activation (hypomania and mania): 167,370-373. mechanism and theoretical significance. Brain Caley, C.F. & Weber, S.S. (1993) Paroxetine: a seResearch, 319,29-40. lective serotoninreuptake inhibiting antidepresDollery, C. (1991) Therapeut~cDrugs. Churchill 27, 1212sant. The Annals of P~ar~acotherapy, 1221. Livingstone, Edinburgh. Duffy, J.D. (1994) Neurogenic stuttering and Cassidy, S.L. & Henry, J.A. (1987) Fatal toxicity of lateralized motor deficits induced by tranylsh antidepressant drugs in overdose. B r ~ t ~Medi7, 171-174. cypromine. Behavio~ral ~eurologY, cal ~ournal,295,1021-1024. Centurrino, F., Baldessarini, R.J. & Frankenberg, Duncan, D. & Taylor, D. (1995a) Which is the selective serotonin re-uptake inhibitors. British
Journal of Psychiatry,165,728-733.
safest antidepressant to use in epilepsy? Psychiatric Bulletin, 19,355-356. Duncan, D. & Taylor, D. (1995b)Which antidepres-
Jacobson, J.N. (1987) Anorgasmia caused by an MAOI. American Journal of Psychiatry, 144,527. Klysner, R., Larsen, J.K., Smensen, P. et al. (1995) Toxic interaction of venlafaxine and isosants aresafe to use in breast-feeding mothers? Psychiatric Bulletin, 19, 549-550. carboxazid. Lancet, 346,1298-1299. AmeriEinbinder, E. (1995)Fluoxetine withdrawal? Kragh-Sorensen, P, (1982)Antidepressants: cardio, can Jo~rnalo f p s y c h i a t ~152,1235. vascular effects in elderly patients.Acta Phara macologica et Toxicologica, 53,138. Ellingrod, V.L. & Perry, P.J. (1994) Ven1aFax:ine: heterocyclic antidepressant.American Jo~rnaZof Krishnan, K.R.-R. (1995) Monoamine oxidase inhibitors. In: The American Psychiatric Press Text~ o s p i t a l P h a r m51,3033-3046. a~, book of Psychopharmacolo~(eds A. F.Schatzberg Edwards, J.G. & Glen-Bott, M. (1983) Mianserin Br~tishJournal of Clini& C. B. Nemeroff). American Psychiatric Press, and convulsive seizures. Inc., Washington,DC. cal P h a r ~ c o l o15, ~ ,~ ~ ~ ! 5 ” y 1 S . Evans, D.L., Davidson, J, & Raft, D. (1982) Early Lader, M.H. (1996) Tolerability and safety: essenand late side effects of phenelzine. J o ~ r ~ofa l tials in antidepressant pharmacotherapy.Journal ofclinical P s y c h i a ~57 ~ ,(Suppl. 2), 39-44. Clinical P s y c h o p h a r ~ a c o l o2,208-210. ~, Possible ven1aFax:ine Laird, L.K. & Benfield, W.H. (1995) Mood disorFarah, A.& Lauer, T.E. (1996) ders. I Major depressive disorders.In: Applie~ withdrawal syndrome.American Journal of PsyTherapeutics (eds L.L. Young & M.A. Kodachiatry, 153,576. Kimble), 6th edn. Applied TherapeuticsInc., Freeman, H. (1993) Moclobemide. Lancet, 342, Vancouver, W/A. The Clinical Use of Drugs. 1528-1532, Fitton, A., Faulds, D. & Goa,K.L.(1992) Lauerma,H. (1996) Successfultreatment Moclobemide: a reviewof its pharmacological of citalopram-induced anorgasmia by cyproheptadine. Acta Psychiafrica Scan~inavica, properties and therapeutic use in depressive illness. Drugs, 43,561-596. 93/69-70. Gillman, P.K. (1995)Possible serotonin syndrome Lazowick, A.L. & Levin, G.M. (1995) Potential withdrawal syndrome associated with SSRI with moclobemide and pethidine.Me~icalJoury, discontinuation.Annals of’ P h a r ~ a c o t h ~ a p29, nal of Australia, 162,554. 1284-1285. Goldstein, D.M. & Goldberg, R.L. (1986) E. et al. (1995) Monoamine oxidase inhibitor-induced speech Lecable, P., Letzelter, J.-M., Lichtblau, 47,604. An open-label study of the clinical acceptability blockage. J o u ~ aofl Clinical Psych~try, of ven1aFax:ine for depression. Primary Care Grimsley, S.R. & Jam, M.W. (1992) Paroxetine, Psychiatry, I, 119-125. sertraline and fluvoxamine: new selective Clinical Pharmacy, Lee, A. & Donaldson, S. (1996) Drug use inpregserotonin re-uptake inhibitors. nancy (4) Psychiatric and neurological disorders: 1% 930-957. Hawley, C.& Smith, V. (1994) Sexual dysfunction Part I. PharmaceuticalJournal, 254, 87-91. and clomipramine: reviewof ’Effects of antide- Leiter, F.L., Nierenberg, A.A., Sanders, K.M. & Stem T.A. (1995) Discontinuation reactions pressants on sexual function’. Me~icalDialogue, no. 421. Biological Psychiat~, 38,694following sertraline. Hayes, PE. & Kristoff, C.A. (1986)Adverse reac695. tions to five new antidepressants.ClinicaZ Phar- Lejoyeux, M., Ades, J., Mourad, I., Solomon, J. & Dilsaver, S. (1996) Antidepressant withdrawal macy, 5,471-480syndrome. CNS Drugs, 5,278-292. Heisler, M.A., Guidry, J.R. & Arnecke, B. (1996) Lichtenwalner, M.R., Tully, R.G. & Cohn, R.D. & Serotonin syndrome induced by administration Pinder, R.D. (1995) Two fatalities involving of ven1aFax:ine and phenelzine. Annals of phenelzine. Journal of Analytical To x ic o lo ~,19, Pharmacotherapy,30, 84. 265-266. Hoes, M.J.A.J.M. & Zeijpveld, J.H.B. (1996) First report of mirtazapine overdoses. 1nternatio~aZ Lundbeck (1995) CitaZopram~Summary of clinical Clinical P s y c h o p h a ~ a c o l o11,147. ~, tolera~ility.Data onfile. Report 001. Copenhagen: Hyttel, J. (1994) Pharmacological characterization Lundbeck Ltd. & Alertsen, A.R. of selective serotonin re-uptake inhibitors Matheson,I.,Pande,H. (SSRIs).l n t ~ a t i o n aClinical l P s y c h o p ~ r ~ c o l o ~ ,(1985) Respiratory depression caused by Ng (Suppl. I), 19-26. desmethyldoxepin in breast milk. Lancet,2,11q. Isenberg, K.E. (1990) Excretion of fluoxetine in McEvoy, G.K.(ed.) (1995) AHFS 95 Drug l n f o ~ a l Psych~a~ry,tion. American Societyof Health System Pharhuman breast milk.J o u ~ofCl~nica1 macists, Bethesda,Maryland. 51,169.
~
~IO: ~ ad ~ eeffec~s ~~ s of e ~~ e d iec a ~ i~o ~ phenelzine and some related compounds in Miller, L.J. (1991) Clinical strategies for the use of 30,205-215. pregnancy. Journul of Endocrinolo~f psychotropic drugs during pregnancy. PsychiPower, B.M., Hackett L.P., Dusci L.J. & Ilett, K.F. atric Medicine,g, 275-297. (1995) Antidepressant toxicity and theneed for Montgomery, S.A. (1992) Suicide and antideidentification and concentration monitoring in pressants. Drugs, 43 (Suppl. 2), 24-31. Pharmacokinetics, 29,154-171. overdose. Clinica~ Monti, J.M. (1989) Effect of a reversiblem o n o a ~ e Preskorn, S.H. (1995)Comparisonof the tolerabiloxidase-A inhibitor (moclobemide) on sleepof depressed patients.British Journalof P s y c h i a f ~ , ity of bupropion, fluoxetine, imipramine, nefazodone, paroxetine, sertraline and ven1aFax:ine. 155 (Suppl. 6), 61-65. ~ o u r n aof~ C ~ i n i c aPsychiatry, ~ 56 (Suppl. 6), Myrenfors, P.G., Eriksson, T., Sandsteat, CS. & 12-21. Sjoberg, F. (1993) Moclobemide overdose.JourPurcell, P. & Ghurye, R. (1995) Trazodone and nal of Internal Medicine,233, 113-115. spontaneous orgasmsin an elderlypostmenoNeuvonen, P.J., Pojhola-Sintonen, S., Tacke, U. & pausal woman: a case report. Journal of Clinical Vuori, E. (1993) Five fatal cases of serotoninsynP s y c h ~ p h a r ~ a c o15, l o ~293-295. , drome after moclobemide-citalopram or moclobemid~lomipramin~ overdoses.Lancet, Rabkin, J., Quitkin, F., Harrison, W. et al. (1984) Adverse reactions to monoamine oxidase inhibi342, 1419. tors. Part I: a comparativestudy. Journal of CliniNorth West Regional Drug Information Service cal P s y c h o p h f f r ~ a c o l4, o ~270-278. , (1995) Can SSRIs affect driving ability?Materia Medica, November. Ray W.A. (1992) Psychotropic drugs and injuries among the elderly: a review. Journul of Clinical Nefazadone (1996) Nefazodone: adverse drug P s y c h o p h a r ~ a c o l o22,386-396. ~, reaction profile.Canadian Adverse. The Extra Otani, K., Tanaka, O., Kaneko, S. et al. (1994) Reynolds, J.E.F. (ed.) (1996) Mar~indale~ Pharmucopo~ia,31st edn. Royal Pharmaceutical Mechanisms of the developmentof trazodone Clinical Society of Great Britain,London. withdrawal symptoms. Inte~nationffl & Flak, K. (1996)The Robinson, G.E., Stewart, D.E. Psychopharmacology,9, 131-133. rational use of psychotropic drugs inpregnancy Ostrom, M.,Eriksson, A., Thorson, J. & Spigset, and postpartum.Canadian Journalof Psychiatr~, 0. (1996) Fatal overdose with citalopram. Lancet, 348,1345-1346. 31,183-190. Adverse events Robinson, D.S., Roberts, D.L., Smith, J.M. et al. Pande, A.C. & Sayler, M.E. (1993) (1996) The safety profile of nefazodone.Journa~ and treatment discontinuations in fluoxetine clinical trials.Internafional ClinicalP s y c h o ~ h a ~ a - of Clinical Psychiat~,57 (Suppl., 2), 31-38. Rowan, J.J. (1976) Excretionof drugs inmilk. PharCOlOgy, 8,267-269. ~aceutical Jour~al, Pare, C.M.B. (1985) The present status of 217,184. ~ o u r nof a ~ Rudorfer, M.V. & Potter, W.Z. (1989)Antidepressmonoamine oxidase inhibitors.Br~tish P s y c h i a ~146,576-584. ~, ants: a comparative review of the clinical pharPastuszak, A., Schick-Boschetto,B.,Zuber, C.et al. macology and therapeutic use of the 'newer'vs. (1993) Pregnancy outcome following first-trithe 'older' drugs.Drugs, 37, 713-738. mester exposure tofluoxetine (Prozac).Journal Russel, J.L. (1996) Relatively low of doses cisapride of the A ~ e r i c a nMedical Association, 269, 2246in the treatment of nausea in patients treated 2248. with venlaFax:ine for treatment refractory dePeet, M.& Behagel, H. (1978)Mianserin: a decade pression. Journal of Clinical P s y c h o p ~ r m a c o l o ~ , of Cliniof scientific development.British Jou~n~Z 16, 35-57. cal P h a r ~ a c o ~5,o55-95. ~f Seifritz, E., Hatzinger, M.,Miller, M.J.,Hemeter, U. & Holsboer-Traschler, E. (1995) Hair loss Philipp, M.,Kohnen, R. & Benkert, 0. (1993) A comparison study of moclobemide and doxepin associated with fluoxetine but not withcitalo~ ~962. f in major depression with special reference to pram. ~ a n a ~ i Ju no u o~~ P~ s ~y c h i a40, effects on sexual function.International C~~nical Shrivastava,RK., Cohn, C., Crowder,J. et al. (1994) 7, qy153. Psychophar~acolo~, Long term safety and clinical acceptability of ven1aFax:ine and imipramine in outpatients Pinder, R.M. (1990) The toxicity of antidepressants in overdose: incidence and mechanism. Interwith major depression.Jour~1of Clinical Psychonational Reviewof Psychiatry, 2,213-227. p h a ~ a c o l o14, ~ ,322-329. Pons, G., Schoerlin, M. & Tam, Y.K. (1990) Shulman, K.I., Walker, S.E., MacKenzie, S. & Moclobemide excretionin human breastmilk. Knowles, S. (1989)Dietary restriction,tyramine British Journalof Clinical P~zarmacolo~~, 29, 27-31. and the use of monoamine oxidase inhibitors. Journal of C l i n i ~ l P s y c h o p ~ ~ a g, c o397-402. lo~, Poulson, E.A. & Robson, J.M. (1964) Effect of
Skowron, D.M. & Stimmel, G.L. (1992) h t i d e pressants and the risk of seizures. Ph a r ~ a c o therapy,zz, 18-22. Steiger, A., Holsboer, F., Gerken, A. et al. (1987) Results of an open clinical trial of brofaromine (CGP 11305,A),a competitive, selective and short-acting inhibitor of MAO-A in major endogenous depression. Phar~acopsychiatry,20, 262-269. (1996) Stimmel, G.L. & Dopheide, J.A. Psychotropic-induced reductions in seizure threshold. CNS Drugs, I,37-50. Staukides,J.A. & Stoukides,C.A. (1991) Extrapyramidal symptoms upon discontinuation of fluoxerine. A ~ e r i c a nJournal of Psychiatry, 148,1263(letter). Sullivan, E.A. & Shulman, K.I. (1984) Diet and monoamine oxidase inhibitors: a re-examination. Canadian~ournalof Psychiatry,29,707-711. Taylor, D. (1995) Selective serotonin reuptake inhibitors and tricyclic antidepressants in combination:interactions and therapeutic uses. British ~ournalof Psychia~ry,167,575-580. Taylor, D. & Lader, M. (1996) Editorial: the l cytochrome enzymes. British ~ o u r ~ofa Psychiatry, 168,529-532. A. & Faivre, F. (1995)Clozapine Taylor, D., Reveley, induced hypotension treated with moclobemide and Bovril. British ~ournalof Psychiatry, 167, 409-410. Hyponatraemia Thomas,A. & Verbalis, J.G. (1995) and the syndrome of inappropriate secretion of antidiuretic hormone associated with drug therapy in psychiatric patients. CNS Drugs, 4, 357-369. Thompson, J.W., Jr, Ware, M.R.& Blashfield, R.K. (1990) Psychotropic medication and priapism: a comprehensive review. ~ o u ~of aClinical l Psychiatry, 51,430-433.
Thornton, S.L. & Resch, D.S. (1995) SIADHasso~ o u ~ a l ciated with sertraline therapy. A ~ e r i c a n of Psychiatry,152,809. Tiller, J.W.G. (1990)Antidepressants, alcohol and psychomotor performance. Acta Psychiatrica Scandinavica,(Suppl. 360), 13-17. Tyrer, P. (1984) Clinical effects of abrupt withdrawal from tricyclic antidepressants and monoamine oxidase inhibitors after long-term treatment. Journal of Affective Disorders, 6, 1-7. Wagner, W., Plekkenpol, B., Gray, T.E. & Vlaskamph, Esser, H. (1992) Review of fluvoxamine safety data base. Drugs, 43 (Suppl. 2), 4s54. Warrington, S.J., Padgham, C. & Lader, M. (1989) The CardiovascularEffects of Antidepressants. Psychological Medicine Monograph Supplement 16. Cambridge University Press, Cambridge. Wedin, G.P.,Oderda, G.M., Klein-Schwartz,W. & Gorman, R.L. (1986) Relative toxicity of cyclic antidepressants. Annals of Emergency ~ e d i c i n e , 151 797-804. Williams, D.R.,Peel, W.J. & Coe, P. (1995) Fibreoptic tracheal intubations: severe trismus due to monoamine oxidase inhibitor. British ~ o u r ~ofa~ntensive l Care, September, 256-261. Winston, F. & McCam, M.L. (1972) Antidepressant drugs and excessive weight gain. British ~ournalof Psychiat~,120,693-739. Woo, O.F., Vredenburg,M., Freitas,P. & Olson, K.R. (1995) Seizures after ven1aFax:ine overdose: a Toxicase report. ~ournalof ~ox~co~ogy-ClinicaZ cology, 331349-550, Zimmer, R.,Gieschke, R., Fischbach, R.& Gasic, S. (1990) Interaction studies with moclobemide. Acta Psychiatrica Scandinavica, (Suppl. 360)~ 84-86. Zisook, S. (1984) Side effects of isocarboxazid. ~ournalof ClinicaE Psychiatry, 4553-58.
33
P H I L I P J. COVVEN
About 20"30% of patients with major depression fail to respond to treatment an appropriate with a single antidepressant drug given in adequate dosage for period of time. At this point there are many possible ways to pursue pharmacological treatment,but few controlled trials to help choose between the various options. In addition, with some exceptions, there are few clinical predictors to help match patients to an appropriate treatment. Reports of different pharmacological treatments in resistant depression continue to abound (Table XLX), showing the need for improved management of these disorders. Below, a number of the better studied of these interventions are outlined, together with the current evidence for their efficacy, particularl Table 11.1 Some treatments for resistant depression. Treatment
Evidence from Evidence from controlled trials studies open
Lithium addition
++ ++ ++
++ ++ ++
Triiodothyronine
++ 0
++ +
TCA* + SSRIt
0
+
Psychostimulants
0
Pindolol + SSHs
0
+ +
ECTS
0
++
MAOI (non-selective) Antipsychotic drugs
Tryptophan (with MAOI or TCA)
++, Compelling; +,quite good; 0, absent.
*Tricyclic antidepressant.
t Selective serotonin re-uptake inhibitor. $Electroconvulsive therapy.
Comment Requires adequate dose For depressive psychosis Augments wide range of antidepressant drugs. Caution with SSWs Potential cardiovascular problems Monitoring for Eosinophilia Myalgia Syndrome needed Monitoring of plasma TCA levels advisable Tolerance and dependencepossible Irritability can occur early in treatment High relapse rates after successful treatment
from blind and placebo-controlled studies. It should be noted that many studies report response rates over ashort period of follow-up (6-8 weeks). From the clinical point of view, however, information about progress over several months is more useful. Another factorworth consideration is the way in which studies assess the clinical response to treatment. For example, a 50% decline in score on the Hamilton Rating Scale for Depression (HRSD) will often be regarded a positive outcome, but patients with high pretreatment HRSD scores can meet this criterion while still retaining significant symptomatology.
itc
If a patient does not respond to oneantidepressant drug, it is comparatively simple totry a preparation with different pharmacological properties. Most of the publishedstudies of this manoeuvre treat patients an open, in sequential way with a newer classof antidepressant; clearly this cannot control for the placebo effect or the possibility of spontaneous remission. With the exception of monoamine oxidase inhibitors, there is, in fact, suprisingly little published evidence that this procedure is really effective, although clinically it appears to be widely used. For example, instudy a by Beasleyet al. (1990)~ 35 patients who had shown minimal response to adequate tricyclic antidepressant (TCA.)treatment in placebo-controlled trials were treated openly with the selective serotonin rehalf uptake inhibitor (SSRI), fluoxetine. After6 weeks of treatment, just over the patients had responded as judged by a decrease HRSD in scores of at least 50%. However, using the more exacting criteria of a finalHRSD score of less than 7, only 17% of the subjects achieved full remission. The study of Nolen et al. (1988a), however,provide less support for the authors studied 68 depressed role of SSRIs in TCA-resistant depression. These patients who had failed to respond to a minimum dose of 150'mg imipramine for 4 weeks. Patients werethen randomly allocated to treatmentwith either the SRI, fluvoxamine or the selective noradrenaline re-uptake inhibitor, oxprotiline. While about 25% of the patients randomized to oxprotiline responded (as judged by a50% decreasein HRSD score), noneof 35 patients allocated to fluvoxamine met this response criteria. There appear to be few useful clinical predictors of response toSSRIs in TCA non-responders. It is fairly well established, however, that where a depressive disorder occurs in the setting of an obsessional illness, treatment with an SSRI or clomipramine is likely to be more effective than other antidepressants (Goodman et al., 1992).
double-blind study in out-patients suffering from this syndrome, Hi~elhoch et al. (1991)found 'moderate' or 'marked improvement' the on CGI in 21 of subjects given tranylcypromine but in only10of 28 treated with imipramine. Recently, reversible, selective type-A MAOIs have become available. These drugs have the great advantage relative to conventional MAOIs in that they do not cause hypertensive reactions with tyramine-containing foods. In a comparative trial of 93 depressed in-patients who had failed to respond to two adequate courses of antidepressant drug treatment, Volz et al. (1994)found that the reversible type A-MAO inhibitor brofaromine(150 mgdaily) produced a similar rateof clinical remission (74%) tothat seen with tranylcypromine (72%) defined by a 50% decline HRSD in ratings. In this study, however, the dose of tranylcypromine (up to 30mg daily) was less than that usually given to depressed in-patients; for example, Nolen et al. (1988b)used up to loo mg daily. The reversible type-A MAO1 moclobemide is more widely available than brofaromine. However, there are no studies indicating its usefulness in resistant depression and a little evidence to the contraryet al., (Nolen 1994). For example, the author studied eight out-patients with major depression who had failed to respond to at least one course of a TCA and one course of a SSRI (at therapeutic doses) with lithium augmentation. Subsequently they failed to respond to moclobemide, 600mg daily for at least 6 weeks. Moclobemide was discontinued and treatment with conventional MAOIs started (tranylcypromine, 3 subjects; isocarboxazid, 3 subjects; phenelzine, 2 subjects). Over the next 8 weeks, fiveof the eight patients were clinically improved as determined by a fall in HRSD scores of at least 50%. Treatment was well tolerated, although two subjects required theaddition of trazodone (5omg) for insomnia. Three patients had histories of previous response to conventional MAOIs; two of these were responders.
A problemin switching antidepressantpreparations is that withdrawal of the first compound may not straightforward. be Patients may have gained some small benefit from the treatment, for example in terms of improved sleep or reduced tension,and this will be lost.In addition, if the first medication is stopped quickly, withdrawal symptoms may result. On the other hand, gradual tapering of the dose makes the changeover in medication rather protracted, which may not be easily tolerated by a depressed and despairing patient. For this reason,in patients unresponsive to first-line medication it may be more appropriate to add a second compoundto the primary antidepressant, in the hope of producing an additive or even synergistic effect. "he major
disadvantage of this procedure isthat that the riskof adverse effects throu drug interaction is increased.
Naturalistic studies have shown that patients with depressive psychosis have low responserates to treatmentwith TCAs alone but may respond well when antipsychotic drugs are combined with a TCA. In a double-blind, random et al. (1985) allocation study of 51in-patients with depressive psychosis, Spiker found that the responserate (finalHRSD <7) to a~triptylinealone was 41%, while that to perphenazine alone was 19%. However, patients receiving combined treatment with these drugs had a sig~ficantly higher responserate (78%)than either of the other two groups. Most of the reports of the drug treatment of depressive psychosis have involved TCAs. However, inan open study Rothschild et al. (1993)found that the combination of fluoxetine (up to 40mg daily) and perphenazine (up to 35mg daily) produced a 73% response rate (reduction in HRSD rat in^ of at least 50%)in 30 patients with DS~-III-R depressive psychosis. From this, it appears that a combination of SSNs and antipsychotic drugs may be effective in depressive psychosis. However, the response criterion w less exactingthan that employed by Spikeret al. (1985)and the majorityof the patients (87%) required 40 mg of fluoxetine. Inaddition, 14 patients developed tremor and two experienced akathisia. SSRIs can themselves cause extrapyramidal movement disorders and may potentiate the extrapyramidal effects of antipsychotic drugs through both pharmacod~amicand pharmaco~neticinteractions (seeYoung & Cowen, 1994).
Lithium given alonehas modest antidepressant properties in patients with bipolar disorder but other depressedpatients show little response (see Price, 1989). There is now, however, good evidence from uncontrolled and controlled trials that lithium added to ineffectiveantidepressant t r e a ~ e ncan t produce useful clinical improvement in patients with major depression. Whetherthis effect of lithium is a true potentiation (augmentation) of the primary antide~ressantcompound or represents simply an additive antidepressant effect of its own is debatable. However, naturalistic studies suggest equivalent rates of response to lithium addition in unipolar and bipolar patients (Price et al., 1989; Nierenberg et al., 1990). Uncontrolled trials have reported that lithium addition is followed by a rapid onsetof tidep press ant effect (within48 h) in a high proportion of subjects (60-70%) (de Montigny et al., 1983). Double-blind, placebo-controlled trials
Table 11.2 Placebo-controlled trials"of lithium addition in patients unresponsive to initial antidepressantdrug treatment. Authors Heninger et al. (1983)
RDC unipolar major depression
Double-blind placebo cross-over (12 days)
Lithium Placebo
5/8 017'
Zusky et al. (1988)
DSM-111 unipolar major depression
Double-blind placebo control, parallel group (14 days)
Lithium Placebo
3/8 2/8
Joffe et al. (1993a)
RDC unipolar major depression
Double-blind placebo control, parallel group (14 days)
Lithium Placebo
9/17 3/16
Stein & Bernadt (1993)
RDC major depression
Double-blind placebo/ Lithium Placebo low doset control, parallel group(6 weeks)
15/34 6/34
Katona et al. (1995)
DSM-111-R major depression
Double-blind placebo control, parallel group (6 weeks)
Lithium Placebo
15/29 8/32
Lithium Placebo
62/130 (48%) 25/131(19%)
Overall response rate
"At least 12 days' duration. t250 mg lithium daily.
confirm that lithium is effectivebut show a more gradual onset of action over 2-3 weeks in about 40-50%of depressed patients (Table 11.2). For example, Katona et al. (1995) studied 62 patients with major depression who failed to respond to 6 weeks' treatment with lofepramine or fluoxetine. Subjects were randomly allocated to either lithium or placebo for further a 6 weeks. Taking the response criterion as a HRSD final score
Patients
response to lithium addition; that is, if a patient had showed a good initial response they were likely to remain well over the follow-up period. Interestingly, this positive outcome was independent of the length of time for which lithium treatment was continued. Demographic and clinical factors such as age, sex, the presence of melancholic depression, dysthmia, hospitalization and suicide attempts were not useful response predictors. There have been few comparisons of lithium augmentation with other treatments of resistant depression. However, Dinan and (1989) Barry randomly allocated 30 severely depressedpatients who had failedto respond to 150rng TCA t r e a ~ e nfor t 4 weeks to either bilateral electroconvulsive therapy (ECT; 6 treatments over3 weeks) or lithium addition. Of the 15patients who received lithium, 10responded (final HRSD rating of<11). The response rate of the ECT-treated patients (11/15) was not different. The lithium addition group included two psychotically depressed patients who did not respond. In this context, it is worth noting that Price et al. (1983) reported useful clinical improvement following lithium addition in five out of six patients with depressive psychosiswho had failed to respond to combined treatment with desipramine and anantipsychotic drug.
Several openstudies have indicatedthat the addition of triiodothyronine to ineffective TCA treatment can bringabout useful clinical response,and this Inthe most has beensupported in threeof four controlledstudies (Table 11.3). recent investigation, Joffe et al. (1993a) studied 50 out-patients with unipolar, non-psychotic major depression. They had failed to respond to 5 weeks' treatment with a TCA (daily dose2.5 mg kg") after which they randomly were allocated to double-blind addition of lithium carbonate, triiodothyronine Table 11.3 Controlled trialsof triiodothyronine (T,) in TCA-resistant depression. Authors Goodwin et al. (1982) In-patients; unipolar depression,4; bipolar depression, 8
Double-blind addition
8/12 responded
Gitlin et al. (1987)
Double-blind placebo controlled cross-over
T3= placebo
Double-bhd, comparison ofT3and
T, response (9/17)>T,
Joffe & Singer (1991) out-patients, 38
Joffe et al. (1993a)
16 out-patients, unipolar major depression
non-psychotic unipolar major depression
T*
50 out-patients, non-psychotic unipolar major depression
Double-blind, placebo-controlled lithium comparison
response (4/21)
T, response (10/17)=
lithium (91'7) >placebo (3/16)
(37.5 pg daily) or placebo for 2 weeks. At the end of treatment, 10 of 17 patients treated with ~iiodothyroninehad responded (50% reductionin HRS (nine of 17patients) wa final HRSD score
Studies in both experimental animals and humans suggest that it is possible to increase brain 5-HT function by administering precursors such as 5h y d r o x y ~ t o p h a n(5-HTP)and tryptophan(TRP) (see Cowen, 1994). There is evidence from controlled trials that the additionof TRP can improve the therapeutic effectof MAOI treatment.It is also possible that TRP can enhance the efficacyof c l o ~ p r a m i n (see e Chalmers & Cowen, 1990). However, there are no controlled trials to indicate that TRP can reliably produce therapeutic benefit in patients who have failed to respond to MAOI or TCAs. Nevertheless, the useof TRP has been recornended to supplement the 5-HTP-potentiating effects of l i t h i u m - ~ O 1and lithium~lomipramine combinations (Barkeret TRP has been associated with the development of the Eosinophilia Myalgia S ~ d r o m e(EMS), a severe connective tissue disease that can have a fatal outcome. Subsequent studies have shown that EMS is almost certainly caused by a contaminant that occurred in the production of TRP from a single m a n u f a c ~ r i nsource ~ (Slutsker et al., 1990). It is of interest that when TRP was withdrawn in the UK, several patients whose severe refractory depression had responded to a combination of lithium, phenelzine and TRP relapsed on TRP discontinuation, despite being maintained on lithium-MAO1 treatment (Ferrier et al., 1990). This suggests that in some patients TRP was exerting an antidepressant effect.
In the UK it remains possible to prescribe TW, in combination with other antidepressant drugs, for patients with chronic treatment-resist~t depression. It should be noted, however, that the combination of TRP with MAOIs can lead to 5-HT neurotoxicity,so caution is needed. In addition, T W given with SSRIs can also result in 5-HT toxicity; so this combinationis not recommended (Sternbach, 1991; Young & Cowen, 1994). y”TP given as a sole treatmentdoes not appear to havea useful role in resistant depression (Nolenet al., 1988b). On thebasis of an open case series, (1980)reported that the addition of 5-HTP to MAOI treatment ine and Sachs produced benefit in 11 of 25 non-responders. owever, this report does not appear tohave been followed up.
he combination of TCAs and MAOIs has beenin use since the1960s when the efficacy of this regime was strongly advocated by leading authorities, Although the combinationof MAOI and TCAsis reported tobe hazardous, the risksof significant interaction can be minimized if reasonable precautions are taken. These include avoiding imipramine and clomipramine, and startin the drugs togetherat low dose or adding the A01 cautiously to established TCA treatment(see Chalmers Csr Cowen, 1990). In patients not selected for treatment resistance the combination of MAOIs and TCAsdoes not appear to confer additional therapeutic benefit over either & Cowen, 1990).Inone controlled study (Davidson drug used alone (Chalrners et al., 1978) 17 in-patients who had not responded to TCA treatment were randomly allocated to ECTor treatment with phenelzine and amitriptyline. The outcome judgedby decrease in HRSD ratings favouredECT. However, about half this group of patients suffered from depressive psychosis and in such patients ECT (or antidepressants and antipsychotic drugs in combinatio would usuallybe preferred. An open studyby Sethna (1974)of 1 2 depressed patients who had failed to respond to either TCAs or MAOIs given separately (or ECT in 10 cases) suggested useful clinical effects of ~ 0 1 - T C Acornbinatio~.At follow-up periods of 7-24 months, nine subjects were reported be to without significant depressive sympto~atology.Most of these subjects had chronic nonmelancholic depression with prominent anxiety symptoms. In addition to these series, case reports continue to appear where it seems well documented that a patient has failed to respond to either a TCA or a I given alone but achievesa useful clinical response when both drugs are used together (Tyrer & Murphy, 1990).Therefore, although controlled evidence is lacking, it seems likely that individual patients with refractory depression are helpedby MAOI-TCA combinations.In general, the adverse
effects of the combination are no worse than with either drug alone, although weight gainand postural hypotension may be more ~oublesome.Conversely, I is given with a TCA such as amitriptyline or trimipramine, induced insomnia may be prevented. There is less information about the combination of other AOIs. Trazodone is fairly commonly used to treat a and is generally well tolerated (Nierenberg C3r Keck, also case series in which patients wi ctory depression have been treated with a combination of the reversibleI-A,moclobemide, and SS as fluvoxamine. Thus far, this com~inationhas been reported safe and, in some patients, therapeutically effective (seeet al., Ebert 1995). The combination of fluvoxamine and moclobemide has also been reported to be safe in healthy volunteers (Dingemanse, 1993). Despite these reports, it is well es hed that conventional contraindicated in combination with because of the risk o toxicity (Sternbach,1991; YoungC3r Cowen, 1994). Furthermore, there are case reports of fatal adverse reactions following moclobemide-SSN overd Young & Cowen, 1995). At present, therefore,the combination of S moclobemide should be usedwith extreme caution,if it is usedat all.
Some open case series have suggested that c o m b i ~ TCA g and SSN treatment may be helpful in refractory depression. For example, in a retrospective chart review, Weilburg et al. (1989) reported a positive response (defined as improvement noted byboth patient and clinician) in 22 of 25 depressed outpatients when fluoxetine was added to ongoing TCA treatment. Of further five patients taking trazodone, four also improved. Thisstudy does not, of course, clarify whether or not the improvemenlt was due to the fluoxetine alone or the co~binationof fluoxetine with the TCA. However,in eight patients the therapeutic responsewas lost when the TCA was withdrawn and restored when it was recommenced. Subse~uently,Weilburg et al. (1991)in a similar study found resolution of depression in13of 20 out-patients in whom nortriptyline or desipramine were added to ineffective fluoxetine treatment. One patient worsened, with severe agitation. Finally, Seth et al. (1992)reported remarkable improvement eight in elderly patients with chronic refractory depression who received SS combination (usually sertraline and nortriptyline). Some patients received concomitant lithium treatment. The only prospective study of TCA-SSN treatment was carried out by Fava et al. (1994) who treated 41 depressed out-patients who had failed to achieve a 50% reduction in HRSD scores in response to a standard dose of
3
fluoxetine (2omg daily). Subjects were randomly allocated to three different groups: (a) high-dose fluoxetine (40-6omg daily), (b) the addition of desipramine (25-5omg daily), or (c) lithium. Taking response as a final HWD score (at 5 weeks) of <7, the most effective t r e a ~ e nwas t highfluoxetine (53% response rate) while lithium and desipramine augmentation appeared of similar efficacy (response rates of 29% and 25%, respectively). However, the plasma levels of lithium obtained in this study were low and probably subtherapeutic. Taken together, the evidence for the efficacy of TGA-SS not compelling. In addition, there are numerous case reports of adverse reactions with agitation and, rarely, seizures. These reactions are generally associated with marked elevationsin plasma TGA levels because SSRIs are potent inhibitors of the cytochrome P450 system by which TGAs are metabolized. M i l e some SSRIs-for example, citalopram and sertraline-may be less likely to produce this effect, plasma TGA monitoring is advisable if combination treatmentis used (see Taylor, 1995). Another option maybe to use treatment witha single drug suchas c l o ~ p r a m i n or e venlafaxine which produces potent 5-HT and noradrenaline re-uptake ~ibition.
Treatments listed under this heading have some limited evidence for efficacy Inaddition, a number of them-for example, cortisol in refractory depression. suppression or pindolol-have an interesting experimental rationale for their use.
The main indication for carbamazepineinisthe acute management of mania and the prophylaxisof bipolar disorder. In general, the acute antidepressant effects of carbamazepine are not striking. However, Ballenger and(19 Post described an antidepressant effect of carbamazepine in 4 of 13 bip depressed patients, 3 of whom hadbeen unresponsive to lithium. ~ubsequently,in a double-blind, cross-overstudy, Post e f al. (1986) found a good clinical response (decrease of at least 2 points on the Bu~ey-Hambur scale) to a 4-week courseof carbamazepine in 12 of 35 depressed patients. In this study, good response was seen in 10 of 24 bipolar depressives, and two of 11 unipolar depressives. However, most of the patients did not relapse upon subsequent placebo-con~olleddis~ontinuationof carbamazepine, questionin the specificityof the treatment effect. In. addition, how far the patients studied were, in fact, refractory to conventional treatment is not clear from the autho description.
C ~ a ~ t11: e r~ r e a t ~ for e ~ r~sistant ts ~~ressio~ In a further blind study, Kramlinger and Post (1989) found that the combination of lithium and carbamazepine produced good clinical improvement (decreaseof at least 2 points on the Bu~ey-Hamburgscale) in eight of 15 patients (13bipolars) who had failed to respond to carbamazepine alone. Taken together, thesedata suggest that carbamazepine may beworth trying in bipolar patients with resistant depression,perhaps in combination with lithium. If the latter combination is used, it is worth noting that coad~inistrationof lithium and carbamazepine has been reported to cause neurotoxicity despite the presence of normal plasma concentrations.
The widespread use of psychostimulants as the sole treatment for depression has been superseded by antidepressants that are not associated with the development of tolerance or abuse. It is possible, however, that psychostimulants such as amphetamine and methylphenidate may retain a place in the treatment of resistant depression, usually in combination with other antidepressants.For example, Wagerand Klein (1988)reported benefit inpartial responders to TCAs following the addition of amphetaminein daily doses of 5-2omg daily. Potential problemsof this approach include troublesome hypertensive reactions as well as tolerance to the therapeutic effect. There is also a small clinical literature on combination treatment with psychostimulant and MAOIs in refractory depression. This combination is normally contraindicated because of the risk of hypertensive crisis. However, in some patients its use appears safe and apparently useful. For example, Fawcett et al. (1991) reported that 10 of 32 severely depressed patients,who had failed multiple treatment trials, were 'improved' or 'much improved' for at least 6 months on the Combination 01with psychostimulant ordexamphetamine).Similarly,Fei et al. (1985) found simila improvement in nine of 16 severely depressed and t r e a ~ e n t - ~ s i s t a who received dexamphetamine or methylphenidate in combinat A01 (and usually a TCA as well). Improvements of this nature in such severely ill patients are worth noting. it must be emphasized that the combination of psychostimulants Is is potentially hazardous and should be used with very close monitoring. Serious adverse reactions were not notedthe incase series above. however, somepatients discontinued treatment for adverse effects includin impotence, hypertension, orthostatic hypotension and shakiness.
There is growing interest in the possible role of excessive glucocorticoid
apte er ss: ~ r e a t ~ e n t s r~sistant for ~epression
secretion in the pathophysiology of depression (Dinan, 1994). Ghadirian et al. (1995) studied 20 depressed in-patients (nine with psychotic symptoms) who had failed to respond to at least two trials of antidepressant drug treatment. They were withdrawn from other psychotropic medication and then treated with one or more steroid suppressive drugs (aminoglutethamide, metyrapone and ketaconazole) for period a of 8 weeks. Of the 20 patients, 11 were responders (>50% decrease on HRSD scores) and two were partial responders (20-50% HRSD decrease). Patients with psychotic depression faired less well. In responders, improvement was maintained for several months even though steroid suppressive treatment had been discontinued. Side-effects were mild and included li~ht-headedness, headache, nausea and fatigue. The of use steroid-suppressing drugs offers an interesting new avenue for the treatmentof resistant depressionbut clearly controlled trials are needed.
In laboratory experiments, many different kinds of antidepressant treatment, including TCAs and MAOIs, produce a down-regulation of P-adrenoceptors or the associated noradrenaline-sensitive adenylate cyclase (Vetulani et al., 1976). The speed of the TCA-induced down-regulation canbe increased by a,-adrenoceptor blockade, perhaps because reduction in presynaptic a,adrenoceptor function increases noradrenaline outflow (Crews et al., 1981). These laboratory findingshave led to suggestions that combining ana,adrenoceptor antagonist with a TCA might speed theofonset antidepressant action or, more speculatively, leadto increased therapeutic benefit in TCAtwo open studies have failed to confirm the latter resistant patients. However, possibili~(Charney et al., 1986; Schrnauss et al., 1988). There is, however, a small caseseries in which thea,-adrenoceptor antagonist,idazoxan, givenas a sole treatment, produced an effective antidepressant response of several A third patient months durationin two resistant bipolar depressed patients. with resistant unipolar depression showedan initial response but relapsed after a numberof weeks (Osmanet al., 1990).
Repeated a d ~ s t r a t i o nof a number of antidepressant drugs, particularly SSRIs and MAOIs, desensitizes inhibitory5-HTI, autoreceptors on 5-HTcell bodies. It has been suggested that this effect contributes to the antidepressant effect of such drugs by freeing 5-HT cell bodies from. feedback control and thereby facilitating the release of 5-HT from nerve terminals (Blier & de Montigpy, 1994).
Based on this idea, Artigaset al. (1994) proposed that the addition of a 5HTIAreceptor antagonist to the medication of patients who had not responded to conventional antidepressants, particularly SSRIs, might produce a therapeutic effect. Because there are at present no selective 5-HT1, receptor antagonists available for clinical use, these authors employed pindolol, a P-adrenoceptor antagonist with 5-HTIA receptor antagonist properties. Pindolol (2.5mg three timesdaily) was added to the drug treatment of seven patients with major depression who had been resistant to multiple medication trials. Five subjects were currently taking an paroxetine SSRI (four and one fluvoxamine) while one received imipramine and one phenelzine. All subjects showeda decrease inHRSD score of at least 50% after a week of pindolol addition and in five, HRSD scores were lessthan 8, indicating full remission (Artigaset al., 1994). Subsequently, Blier and Bergeron (1995) reported on the addition of 'pindolol(2.5mg three timesdaily) to 18 patients with major depressionwho had failed to respond to treatment at least with antidepressant together with the addition of another treatment, usually lithium. Pindolol(2.5 mg three times daily) was added to paroxetine (eight subjects), sertraline (five subjects), fluoxetine(threesubjects), and moclobemide (two subjects). Overall, pindolol produced a significant improvementin depressive symptoms after1week. By 2weeks, all the patients except those on sertraline had a HRSD score of 10 or less showinga very good clinical improvement. Interestingly, none of the sertraline-treated patients responded. It is not clear whether or not this may represent a chance finding. Pindolol addition in the latter study was generally well tolerated. However, two patients had to stop taking pindolol in3 days the because first of irritability. Furthermore, one patient who responded to the addition of pindolol to paroxetine developed mania5 weeks later.
Among the indications for ECT is that of failure to respond to adequate antidepressant drug treatment. Trials of ECT typically report high response rates for ECT (about 80%), but patients unresponsive to drug treatment are not usually consideredas a separate group. Prudic et al. (1990) studied the effect of previous antidepressant drug treatment on the response of53 patients who received bilateral ECT. They found that amongst those who had received adequate pharmacotherapy (a TCA in a dose of at least aoomg daily for at least 4 weeks), the response rate to ECT (defined as a 60% reduction inHRSD score) was50%. In contrast, the response rateof patients who had not receivedadequate drug treatment was significantly greater (86%). The presence of medication resistance therefore
"47
~ ~ a ~11:t ~er re a t ~ e nfor t s resist~nt de~ression decreases the likelihood that a patient will respond to ECT. Nevertheless, about half of such subjects arelikely to experience significant improvement. Another pointthat needs tobe considered is the outcome after ECT. Some limited data suggest that even where ECT is has been therapeutically successful, pretreatment medication resistance can be associated witha high subsequent relapse rate. Sackheim et al. (1990) followed 58 patients who responded to ECT and found that1 year post-treatment, 50% had relapsed, meeting the research diagnostic criteria (RDC) criteria for major depression. The relapse ratein patients who had received adequate drug treatment prior (64% compared to 32%). to ECT was significantly higher in those who had not The relapse rate after ECT was only weakly influenced by whether or not patients received adequate antidepressant drug treatment post-ECT.
Chronic, severe and intractable depression is regarded by some as an indication for psychosurgery (see Gelder et al., 1996). There are no controlled trials of this procedure and they would, of course, be extremely difficult to carry out. Poynton et al. (1995) described a cohort of 23 severely ill, treatment-resistant patients, 16 of whom had major depression and five of whom had bipolar disorder. They underwent stereotactic subcaudate tractotomy and were followed up prospectively. At 6 months postoperatively the mean HRSD score had fallen from 17.9 to 9.7. After 1 year of follow-up, 13 of the patients had shown either a very good or good response, with the remainder experiencing intermediate benefit no orchange. The HRSD scores6 months at correlated with global outcome, suggestingthat at 6 months the pattern of response maybe established. No serious physical sequelaeof the operation were reported. However, detailed neuropsychological testing showed changes in the patternof cognitive function. For example, there were improvements in some tests requiring speed and attention but deterioration on others, such as the Stroop task, which depend on intact frontal lobe function.
Artigas, F. Perez, V.& Alvarez, E. (1994)Pindolol induces a rapid improvement of depressed patients treated with serotoninreuptake inhibitors. Archives of General Psychiatry,51f 248-251. Ballenger, J.C.& Post, R.M. (1980)Carbamazepine in manic depressive illness: a new treatment. A~erican ~ournulof Psychiat~,137~ 782-790. Barker, W.A., Scott, J. & Eccelston, D.(1987) The Newcastle Chronic Depression Study: results of a treatment regime. ~n~ernational Clinical
Psychophar~acolo~, 2,261-272, Beasley, C.M., Sayler, M.E. & Cunningham, G.E. et al. (1990) Fluoxetine in tricyclic refractory Z major depressive disorder. ~ o u r ~of ~Afective Disorders, ~0,19y”oo. Blier, P. & Bergeron, R. (1995) Effectiveness of pindolol with selected antidepressant drugs in the treatment of major depression. ~ o u r n aof~ ClinicaZ Psychophar~acolo~, 15,217-222. Current advances Blier, P.& de Montigny, C. (1994)
and trends the in treatmentof depression. Trends Fontaine, R., Ontiveros, A., Elie, R. & Vezina, M. (1991) Lithium carbonate augmentation of in Phar~acological Sciences, 15,220-226. desipramine and fluoxetinein refractory depresDrug treatment Chalmers, J.S. & Cowen, P.J. (1990) sion. Biological Psychiat~,29,9464948. of tricyclic resistant depression. ~nternational Gelder, M.G., Gath, D.H., Mayou,R.M. & Cowen, Review of Psychiatry,2,239-248. P.J. (1996) Oxford Textbook of Psychiatry. Oxford Charney, D.S., Price, L.H.& Heninger, G.R. (1986) D e s i p r a ~ ~ y o h ~ bcombination ine for refrac- University Press, Oxford. tory depression.Archives of General Psych~atry, Ghadirian, A.M., Engelsmann, F. & Dhar, V. et ai. (1995) The psychotropic effects of inhibitors of 43,1155-1161. steroid biosynthesis in depressed patients reCowen, P.J.(1994) The effect of tryptophan on fractory to treatment. Biological Psychiatry, 37, a n brain 5-HT function: a review. ~ u ~ Psycho~ h a r ~ a c o 9,371-376. lo~, 369-375. Crews, F.T., Steven, M.P. & Goodwin, F.K. (1981) Gitlin, M.J., Weiner,H. &Fairbanks,L. (1987) Failure of T3 to potentiate tricyclic antidepressant Acceleration of beta-receptor desensitisation in response. Journal of Afjcective Disorders, 23, 267combined administrationof antidepressant and 272. phenoxybenzamine. Nature, 290,787789. Davidson, J., McLeod, M,, Law-Yone, B. & Goodman, W.K., McDougle, C.J. & Price, L.H. (1992)Pharmacotherapyof Obsessive CompulLinnoila, M. (1978)Acomparison of electroconsive Disorder. Journal of Clinical Psychiatry, 53 vulsive therapy and combined phenelzineamitriptyline inrefractory depression.Archives (4 Supp1.), 29-37. Goodwin, F.K., Prange, A.J. & Post, R.M. (1982) of General Psychiatry, 35,639-642. Potentiation of antidepressant effects by LDinan, T.G. (1994) Glucocorticoids and the gentriiodothyronine in tricyclic non-responders. esis of depressive illness: a psychobiological A~erican Journal of Psychiatry, 239,34-38. , 365model. British Journal of P s y c h i a t ~164, Hale, A.S., Procter, A.W. & Bridges, P.K. (1987) 371. Clomipramine, tryptophan and lithium comin Dinan, T. & Barry, S. (1989)A comparison of elecbination for resistant endogenous depression: 7 troconvulsive therapywith a lithim and tricyclic case studies. British Journal of Psychiafry, 151, combination among depressed tricyclic non213-217. responders. Acta Psychiatrica Scandinavica, 80, Heninger, G.R., Charney, D.S. & Sterriberg, D.E. 97-100. (1983)Lithium carbonate augmentation of antiDingemanse, J. (1993) An update of recent Archives ofGeneralPsychidepressant treatment. moclobemide interaction data. ~nternational Clinical P s y c h o p h ~ r ~ a c o l7,167-180. o~, 40,1335"1342. Ebert, D., Albert, R., May A., Stosiek, I. & Kaschka, Himmelhoch, J.M., Thase, M.E., Mallinger, A.G. & Houck, P. (1991) Tranylcypromine vs. W. (1995) Combined SSRI-RIMA treatment in imipramine in anergic bipolar depression. refractory depression: safety data and efficacy. A~erican Journal of Psychiatry,148,910916. P s y c h o ~ h a r ~ a c o219,342-344. lo~, Rosenbaum, J.F., McGrath, P.J., Stewart, Joffe, R.T. & Singer, W. (1991) Thyroid hormone Fava, M., potentiation of antidepressants. In: Advances in J.W., Amsterdam, J.D. & Quitkin, EM. (1994) Neuro~sychiatryand P s y c ~ o ~ h a r ~ a c Vol. olo~ 2: . Lithium and tricyclicaugmentationof fluoxetine ~ e ~ a cDepression t o ~ (ed. J. D. Amsterdam), pp. treatment for majordepression:a double blind 185-189. Raven Press, New York. controlled study.A ~ e r i c a nJo~rnalof Psychiatry, Joffe, R.T., Singer, W., Levitt, A.J.& MacDonald, 151,1372-1374. C. (1993a)A placebo-controlled comparison of Fawcett, J., Kravitz, H.M., Zajecka, J.M. & Schaff, lithium and triiodothyronine augmentationof M.R. (1991) CNS stimulant potentiation of tricyclic antidepressants in unipolarrefractory monoamine oxidase inhibitors in treatmentdepression. Archives of General Psychiafry, 50, refractory depression.Journal of Clinical Psycho~ h a r ~ a c o XI, l o ~127-132. , 387-393. Feighner, J.P., Herbstein,J. & Dadouji, N. (1985) Joffe, R.T., Levitt, A.J., Bagby R.M., MacDonald, C. & Singer, W, (199313) Predictors of response Combined MAO, TCA, and direct stimulant to lithium and triiodothyronine: augmentation Jourtherapy of treatment-resistant depression. of antidepressants intricyclic non-responders. nal of Clinical Psychiatry,46,206-209. British Journal of Psychiatry, 163,574-578. Ferrier, I.N., Eccleston, D., Moore, P.G. & Wood, K.A. (1990) Relapse in chronic depressives on Johnson, F.N. (1991) Lithium augmentation therapy for depression.Reviews in C o n t e ~ ~ o r a r y withdrawal of L-tryptophan. Lancet, 336,380P h a ~ a c o t h e r a ~2,3-52. y, 381.
249
C ~ ~ tTI:e ~r r e a t ~ e for n ~ resistant s d~ression Katona, C.L.E., Abou-Saleh, M.T. & Harrison, D.A. Nolen, J. Zohar, S.P. Roose & J.D. Amsterdam), pp. 59-68. John Wiley & Sons Ltd, Chichester. et al. (1995) Placebo-controlled trial of lithium augmentation of fluoxetine and lofepramine. Osman, O.T., Rudorfer, M.V. & Potter, W.Z. (1990) Idazoxan: a selective a2-antagonist and effecBritish ~ournulof Psychiutry, 116,80"86. tive sustained antidepressant two in bipolar deKline, N. & Sachs, W. (1980) Treatmentof deprespressed patients. Archives of General Psychiatry, sion with anMAOinhibitor followed by 5-HW: an unfinished research project.Acta Psychiu~ricu 46,95&-959. Post, R.M., Uhde, T.W., Roy-Byrne,P.P.& Joffe, R.T. Scundinavicu,280,233-241. Kramlinger, K.G. & Post, R.M. (1989) The addi(1986)Antidepressant effects of carbamazepine, A ~e r ic u ~ournul n of Psychiutry, ~43~2cy-34. tion of lithium to carbamazepine: antidepressant efficacy in treatment resistant depression. Ar- Poynton, A.M., Kartsounis, L.D. & Bridges, P.K. chives of General Psychiatry,46,794-800. (1995)Aprospective clinical study of stereotactic McGrath, P.J., Stewart, J.W. & Nunes, E.V. (1993) subcaudate tractotomy. Psychologicul Medicine, A double blind cross-over trial of imipramine 25,763-770. augmentationin tricyclic and phenelzine for outpatients withtreatment- Price, L.H. (1989) Lithium resistant depression. In: Treutment of Tricyclicrefractory depression. Americun ~ournalof Psyresistant Depression (ed. I. L. Extin), pp. 5179. chiatry, 150, 118-123. American PsychiatricPress, Washington, DC. de Montigny, C., Cournoyer, G. & Morissette, R. et al. (1983)Lithium carbonate additionin tricyclic Price, L.H., Conwell, Y. & Nelson,J.C. antidepressant-resistant unipolar depression: (1989) Lithium augmentation and combined correlationsof the neurobiologicalactions of trineuroleptic-tricyclic treatment in delusional l 140, cyclic antidepressant drugs and lithium ion depression. Americun ~ o u ~ofuPsychiut~, 318-322. on the serotonin system. Archives of General Prudic, J., Sackeim, H.A.& Devanand, D.P. (1990) P s yc h ia t ~40, , 1327-1334. Medication resistance and clinical response to Nierenberg, A.A. & Keck, P.E. (1989)Management of monoamine oxidase inhibitor-associated inelectroconvulsive therapy. Psychiutry Reseu~ch, ~ l Psycho31,287-296. somnia with trazodone.~ o uof Clinicul Quitkin, F.M., McGrath, P.J. & Stewart, J.W. et al. phurmacolo~,9, 45-54. (1989) Phenelzine and imipramine inmood reNierenberg, A.A., Price, L.H., Charney, D.S. & active depressives. Archives of Generul PsychiHeninger, G.R. (1990) After lithium augmentation: a retrospective follow-up of patients with atry, 46,787793antidepressant-refractory depression. ~ o u ~ofu l Rothschild, A. J.,Samson, J.A., Bessette, M.P. & Carter-Campbell, J.T. (1993) Efficacy of the comAfective Disorders, 18,167-175. bination of fluoxetine and perphenazinein the Nierenberg, A.A., Feighner, J.P., Rudolph, R., Cole, treatment of psychotic depression. ~ournulof J.O. & Sullivan, J. (1994) Venlafaxine for treatClinicul Psychiat~,54,338-342. ment-resistant unipolar depression. ~ournulof Sackheim, H.A.,Prudic, J. & Devanand, D.P. et. al. Clinical P s y c h o p h u ~ u c o l o14, ~ , 41cy-423. (1990) Theimpact of medication resistance and Nolen, W.A., Van de Putte, J.J. & Dijken, W.A. et continuation of pharmacotherapy on relapse al. (1988a) Treatmentstrategy indepression. I. following response to electroconvulsive therapy Non-tricyclic and selective reuptake inhibitors in major depression. ~ournulof Clinical Psychoin resistant depression: a double bllnd partial p h u ~ a c o l oio, ~ ,96-104. cross-over study onthe effects of oxprotilineand fluvoxamine. Acta Psychiutricu Scun~inuvicu,78, Schmauss, M., Laakman, G. & Dieterle, D. (1988) Effects of az-receptor blockade in addition to 668-675. tricyclic antidepressantsin therapy resistant deNolen, W.A., Van de Putte, J.J. & Dijken, W.A. l Psychophu~ucolo~, pression. ~ o u ~ofuClinicul et al. (1988b) Treatmentstrategy indepression. 8,108-111. 2. MAO inhibitors in depression resistant to cyclic antidepressants: two controlled cross- Seth, R., Jennings, A.L., Bindman, J., Phillips, J. & Bergmann, K. (1992) Combination treatment over studies with tranylcypromine vs. L-5with noradrenaline andserotonin reuptake inhydroxytryptophan and nomifensine. Acta hibitors in resistant depression. British ~ o u ~ u l Ps~chiutricuScandinuvica, 78,676-683. of Psychiutry, 161,562-565. Nolen, W.A., Hoencamp, E., Haffmans, P.M.J. & Sethna, E.R. (1974) A studyof refractory cases in Bouvy, P.F. (1994) Classical and selective depressive illness and their response to commonoamine oxidase inhibitors in refractory bined antidepressant treatment. British ~ o u ~ u l major depression. In: Refiuctory Depression; of Psychiutry, x q , 265-272. Current Strategiesand Future Directions(eds W.A.
250
Slutsker, L., Hoesly, F.C., Miller, L., Williams, P., Volz, H.-P., Faltus, F., Magyar, I. & Moller, H.-J. (1994) Brofaromine in treatment-resistant de& Flemming, D.W. (1990) Watson,J.C. Eosinop~ia-myalgiaSyndrome associated with pressed patients: a comparative trial vs. ~ Disorders, exposure to tryptophan from a single manufac- tranylcypromine. J o u ~ oaf Affective 30, 209-217. turer. ~ournalof the American Medical Association, Wager, S.G. & Klein, D.F. (1988) Treatment refrac264, 213-217. tory patients: affective disorders. PsychopharmaSpiker, D.G., Weiss, J.C. & Dealy, R.S. (1985) The pharmacological treatment of delusional depres-cology Bulletin, 24,6974. Weilburg, J.B., Rosenbaum, J.F., Biderman, J., sion. American ~ournalof Psychiatry, 142, 430Sachs, G.S., Pollack, M.H. & Kelly, K. (1989) 436. Stein, G. & Bernadt, M. (1993) Lithium augmenta- Fluoxetine added to non-MAO1 antidepressants converts non-responders to responders: a pretion therapy in tricyclic resistant depression: a liminary report. Journalof Clinical P s y c h ~ a t50, ~, controlled trial using lithium in lower than normal doses. British ~ournalof Psychiatry, 162, 447-449. S. Weilburg, J.B., Rosenbaum, J.F., Meltzer-Brody, 634-4540. & Shushtari, J. (1991) Tricyclic augmentation of Sternbach, H. (1991) The serotonin syndrome. fluoxetine. Annals of Clinical Ps y c h ia t ~/ 3, 2 0 9 American Journal of P s yc h i a t ~148,705y13. , Taylor, D. (1995) Selective serotonin reuptake 213. inhibitors and tricyclic antidepressants in com-Young, A.H. & Cowen, P.J. (1994) Antidepressant drugs. In:Side-effectsof Drugs Annua~17 (eds J.K. bination: interactions and therapeutic uses. Aronson & C.J. van Boxtel), pp. British ~ournalof P s yc h i a t ~167, / 575-580. 16-25. Elsevier, Amsterdam. Tyrer, P. & Murphy, S. (1990) Efficacyof combined antidepressant therapy in resistant neurotic disYoung, A.H. & Cowen, P.J. (1995) Antidepressant order. British ~ournalof Psychiatry, 156, 115drugs. In:Side-effectsof Drugs Annual18 (eds J.K. 118. Aronson & C.J. van Boxtel), pp. 14-24. Elsevier, Amsterdam. Vetulani, J., Stawarz, R.J., Dingell, J.V. & Sulser, P. (1976) A possible common mechanism of action Zusky, P.M., Biderman, J.& Rosenbaum, J.F. (1988) of antidepressant treatments: reduction in sen- Adjunct low dose lithium carbonate in treatsitivity of noradrenergic cyclic ANP generating ment-resistant depression: a placebo controlled study. ~ournalof Clinical Psychopharmacolo~,8, system in the rat limbic forebrain. Archives of Phar~acology,293, 109-114. 120-124.
A L L A N I. F. SCOTT
The focus of this chapter is research evidence for the efficacyof electroconvulsive therapy (ECT) in the treatment of depressive illness,but it must be acknowledged at the outsetthat the clinical useof ECT is heavily influenced by more than pharmacology. Pippard (1992) found a l=?-folddifference in the rate of ECT usage between health districts in the South of England. Hermam et al. (1995) found a variation of 0.4 to 81.2 patients per IOOOO population among 317 metropolitan areas of the USA; in approximately one-third of these areas no useof ECT was reportedin ~$38-89.ECT may simply notbe available in some partsof the world. Other non-pharmacological factors that affect usage include psychiatric education and training, individual/ c o ~ ~prejudice, t y and legal restraints. Some critics of ECT have suggested that it causes actual brain damage, but recent prospective brain-imaging studies have found no evidence of any structuralbrain changes after single or repeated courses of ECT (Devanand et al., 1994; Scott, 1995). The conditions necessary to produce the death of neurones after continuous cerebral seizure et al., 1994). activity arein no way approached during modified ECT (Devanand
After the introduction of ECT by Cerletti in Rome in 1938, a substantial number of studies of theoretical interest were conducted (see Fink, 1979; Crow & Johnstone, 1986),but uncertainty over diagnosis, heterogeneity of diagnosis, non-random allocation and non-blind assessment of clinical outcome meant that the vast majority were of uncertain relevance to contemporary practice and of no more than historical interest.
of the tricyclic antidepressants and m o n o a ~ noxidase e It was the introduction inhibitors (MAOIs) that ledto a more thorough and methodologically sound assessment of the effectivenessof ECT in the treatmentof depressive illness.
le 12.1 shows the findings fromtwo large-scale, multicentre,random allocation comparisonsof ECT with tricyclic and MAO1antidepressant drug treatment and placebo. Thestudy by Greenblattet al. (1964)was conducted among S at three state hospitals around Boston in the USA and the study dical Research Council (MRC; 1965) was conducted among inpatients at 30 hospitals in three regionsof England. There were differences between the two studies in entry criteria and the period of assessment, but the results were broadly similar. ECT was associated with the greatest likelihood of marked improvement or syrnptom resolution. MRC The trial also concluded that ECT had a faster onsetof antidepressant effect than imipramine, but the statistical assessmentin support of this statement was not provided. 0th studies were carried out more than 30 years ago. Greenblatt and coles (1964) reported that depression occurred in the context of manicdepression, involutional melancholia, or schizoaffectivedisorder in 65% of ients r a n d o ~ z e d18% ; suffered a psychoneurotic depressive reaction; 6% the depressive diagnoses were mixed.MRC Thestudy randomized patients whose depressed mood was observable an to examiner and accompanied by one of the following features: guilt, sleep disturbance, hypochondriasis, psychomotorretardation or agitation;43% of patients were rated erely ill by clinicians at the outset of treatment. A major criticismof the study is that some imipramine-treatedpatients might have responded better to a higher dose or longer period of treatment; a similar criticism has been made for the dose of phenelzine. Both studies were largely randomallocation, although the placebo arm of the study by Greenblattand colleagues was added during the second year. Neitherpatients nor raters could distinguish who was treatedwith active antidepressant drug or placebo,but both ~atients and raters h e w who was treated by ECT. Crow and Johnstone (1986) have argued that neither study eliminated the possibility that some aspectof treatment other than the induction of the convulsion was responsible for the e other two studies in Table 12.1 have been included because they assessed the antidepressant effect attributable to the convulsion itself by the inclusion of a simulated ECT condition; they also measured the change in depressive symptoms among ECT-treatedpatients and compared this tothe change among imipramine-treated patients. Robin and Harris (1962)reported that two bilateral ECTs combined with placebo tablets led to a greater reduction in Hamilton RatingScale Depression (HRSD) score than imipramine plus simulated ECT after only1 week. Unfortunately, neitherthe depression rats nor significance test was reported. Theof dose imipramine was not stated and the very poor clinical outcome among patients treated drugby plus simulated ECT suggested that this may have been too low. The trial reported was only in s u ~ a r yand , this left considerable methodological uncertainty about
its significance. Thestudy by ~angadharet al. (1982) was the only methodologically sound study in the ECT l i t e r a ~ r ethat compared the rates of the antidepressant effects of ECT and a tricyclic antidepressant, although there might now be some concern that the dose of irniprarnine was increased too slowly. The reduction in HRSD score was clearly and significantly greater among patients treated by six bilateral ECTs over 2 weeks, although no difference was seen from 4 weeks on; by 4 weeks, ECT-treated patients had received eight bilateral treatments and drug-treated patientshad been prescribed75mg 3 weeks. imipramine for the first week, then 150mg imipramine for another In conclusion, the clinical efficacy of ECT in the treatment of depressive illness that requires in-patient treatment was clearly establishedin two large random-allocation, multicentre trials. These also suggested that the clinical efficacy of ECT was greaterthan that of imipramine over4-8 weeks, but this finding must now be treated with caution because more drug-treated patients might have recovered if higher doseshad been allowed. One of the suggested indications for the use of ECT is that it works more quickly than antidepressant drug treatment, and it is disappointing to note that this topic has not received the research attention that it merits; there is preliminary evidencethat the rate of onset of antidepressant effect is greater for bilateralECT given three times a week than moderate dosesof irnipramine during the first 2 weeks of treatment.
Table 12.2 shows the six trialsthat have been designed specifically to assess the role of electrically induced cerebral seizure activity in the therapeutic efficacy of EGT in depressive illness.Each involved the random allocation of at least 25 depressed patients and included reliable outcome measures that were made by raters reported to beunaware of which patients had been treated with real as opposed to simulatedEGT. In the three largest studies, almost all the patients had been ~ i t ~ d r a w from n concomitant antidepressant drug treatment, Every study found that real ECT was superior to simulatedECT in at least one major outcome measure, although in onestudy the difference was not statistically sig~ficantambo bourn SE Gill, 1978). The overall conclusion from thesestudies was that they disprove the suggestionthat some non-convulsive aspectof ECT such as the therapeutic attention is entirely responsible for the observed antidepressant effect, and they confirmthe findings of s e d na1 studies of the modeof action of ECT that the induction of cerebral seizure activity contributes the to therapeutic effect. Nevertheless, some of these studies did arouse considerable controversyat the time (see Crow& Johnstone, 1986). Thepresent author would contrast the debate about the finding that depressed patients improved during a courseof simulated ECT with the tacit
acceptance of the observation that 39% of depressed patients in the MRC trial recovered on placebo tablets alone (so-called spontaneous recovery). Clearly, factors unrelated to the induction of cerebral seizure activity contribute to the clinical improvement seen in depressed patients treated with ECT in the same way that non-pharmacological factors contributethe toclinical improvement seen in depressed patients admitted to hospital and treated with antidepressant drugs. The first fivestudies in Table 12.2 have been subjected to detailed methodological criticism by Crowand Johnstone (1986). Eight patients dropped out of the study by Freemanand colleagues for reasons other than rapid clinical improvement,and how thesedropouts were treated in the statistical analysis had a considerable bearing on the findings. The study also included depression rating scales that showed as i ~ i f i c a nadvantage t in favour of real ECT after two treatments, but Crow and Johnstone were sceptical that a significant clinical difference could appear so early in a course of cluded thatthe result of the study was not nearly as firm as suggested. It was shown su~se~uently that substantial improvement can occur inearly a course of ECT (Rodgeret al., 1994), whichwould lend credibility the to findin on depressionrating scales. The clinicalim~rovementobserved by Lambourn and Gill amongpatients treated with real ECT was the least of the studies listed in Table 12.2; it was the onlystudy to use a low-dose, brief-pulse electrical stimulation in combie f al. (1987)later showed nation with unilateral electrode placement; Sackeim that a low-dose techniquewith a higher electrical dose was~ e r a p e ~ t i c a l l y inferior to low-dose treatment with bilateral ECT and associated with a poor clinical response bythe end of a courseof treatment (Table 12.3). The clinical description of patients for inclusion in the Northwick Park ECT trial was rigorous and the vast majorityof patients (89%) completed the ree of clinical improvement observed amongpatients treated with simulated ECT was the greatestof the studies in Table 12.2, and Crow and Johnstone suggestedthat the relatively small and well-staffed research ward in which the patients were nursed may have contributed tothe nonspecific therapeutic effects through increased medicaland nursing attention. No information was provided by West (1981) about the procedures for r ~ d o ~ z a t i and o n clinical assessment without knowledge of treatment ing in what was reported as a single-author study. Crow and Johnstone believed that this diminished the weight that could be attributed to the findings. The Leicester trial (Brandon et al., 1984)was the largest ever random-allocation comparisonof simulated and real ECT in the treatment of depressive illness. Crow and Johnstone have rightly notedthat the initial sample was diagnostically widerthan that of the Northwick Park trial, although the meth-
57
~ h ~ p12: t ~lect~oco~vulsive e ~ ther~py
*
odology was sound. The clinical improvement observed among those treated by real bilateral ECT was similar to that observed in the Northwick Park trial, but the difference in clinical improvement between those treated with six simulated andsix bilateral ECT was three times greater in the Leicester trial than that observed in the Northwick Park trial; as noted above, greater clinical improvement was seen among patients treated with simulated ECT at Northwick Park for reasons speculated upon above. The Nottingham ECT study (Gregoryet al., 1985) used inclusion criteria similar to the MRC study, although the age range was extended to include patients younger than 40 years andan additional entry criterion was weeks. stipulated thatthe index episodeof depression had lasted for at 4least Unfortunatel~the results cannot be directly compared with the earlier study, which did not report change in total rating scale scores during the course of ECT. Twenty-five patients(36% of the total sample) were withdrawn before they had receivedsix treatments. Sufficient data from earlier in the courseof treatment were reported to confirm that clinical improvement in those treate with ECT was significantly greater than among those treated with simulated ECT after only four treatments. The effect of size the difference between clinical improvement in bilateral and simulated ECT was much closer to that observed in the Leicester study than that observed in the Northwick Park study. The clinical improvement observed after six right unilateralECTs with a high-output ECT machine was greater than that observed by Lambourn a low-output ECT machine; and Gill (19$3), when treatment was given with the importance of electrical dose in determining the clinical effectiveness of unilateral ECT willbe discussed in detail below. In conclusion, the role of the electrically induced convulsion in the therapeutic efficacy of ECT was established by these six random-allocation comparisons of simulated and real ECT in which the clinical improvement observed among patients treated with real ECT was always greater than that observed among patients treatment with simulated ECT; the only study in which the greater improvement was not statistically significant used a technique of unilateral ECT that has subsequently been shown to be inadequate. The first five of these studies have been subject to detailed methodological scrutiny by Crow and Johnstone (1986) and, writing 10 years later, the criticisms of greatest pertinence remain the uncertainty to what extent the assessments of outcome were made without knowledgeof treatment in the study by West, and the importance in the statistical handling of early withdrawals from treatment and its impact upon the effect size.
Early reports of the use of ECT in depressive illness noted that relapse was
sometimes a problem and this was later quantifiedstudies in conducted with the then newly introduced tricyclic antidepressants. Seager and Bird (1962) observed that 11of 16 patients (69%) who had been successfully treated for depression byECT relapsed within 6 months while prescribed placebo tablets in contrast to patients prescribed imipramine,of whom onlytwo out of 12 (17%) relapsed. An important study by Bartonet al. (1973) monitored the progress of 50 patients who had suffered a primary depressive illness that had been successfully treated with ECT; no treatment for depression was offered during follow-up, although hypnotics were prescribed. Twenty-three patients (46% of the sample) relapsed without treatment within 12 weeks, and 16 of those who relapsed (70%) did so within 2 weeks of the end of the course of ECT. The Northwick Park, Leicester and Nottingham ECT studies all included assessment inthe 6 months after ECT, although treatment was then open for all patients. No significant advantage was found for ECT-treated patients. Crow and Johnstone (1986)concluded that the effects of the induced convulsion wereof limited duration. A distinction is now made between symptomatic remission and actual recovery fromthe index episodeof depressive illness. Amongpatients who are symptom-free after treatmentwith ECT, as many will be in potentially temporary remission as fully recovered and c o n ~ u a t i o antidepressant n drug treatment ought to bestandard (Royal Collegeof Psychiatrists Special Committee onECT 1995). ECTis just like any otier pharmacological treatment for depressive illnessin this regard.
There have been manyattempts to correlate clinical features, physiological measures, or personality characteristicswith successful outcome after ECT (Fink, 1979). Early studies often includedpatients suffering from schizophrenia and neurotic conditions as well as patients suffering from depressive illness and many of the putative correlateswith recovery were simply characteristics of depressed patients as opposed to schizophrenic or neurotic patients. None of the studies has ever included a comparisongroup of patients who received no treatment at all to investigate the possibility that a putative correlate of a good outcome was actually a correlate of the likelihood of spontaneous recovery. Crow and Johnstone (1986) have argued that only the inclusion of a group of patients treated by simulated ECT will answerthe potential objectionthat any correlate of a good outcome is simply a correlate of a good outcome to any physical treatment for depressive illness. A combined analysisof the Leicesterand Northwick Park trialsfound that depressed patients with either retardation or delusions accounted for mostof
the difference in clinical improvement between the patients treatedby real and simulatedECT. The difference in clinical improvement between real and simulated ECT in depressed patients without retardation or delusions was not statistically significant uchan et al., 1992). This finding stimulated further analysis of the results from the Nottingham trial (O’Leary et al., 1995)~with different results. There was a marked and statistically significant differencein clinical i~provementamong depressed patients without psychomotor retardation or delusions between real and simulated ECT. The paper also contained a useful comparisonof percentage imrovement in different categories of depressed patients among the Leicester, orthwick Park and Nottingham trials. The major difference among the three trials was the extent of the clinical improvement in patients treated with simulated ECT at Northwick Park; possible reasons have already been discussed above. The average percentage change in HDRS score was 57 was 30% in the Leicester trial and 38% in the Nottinghamtrial colleagues suggested that the different p~tternsof improvement with simulated and real ECT among the trials mightbealso explained by differences in the study samples in factors suchas the lengthof stay in hospital before ECT, previous failureof the index episode to respond to antidepressant drug treatment, and a past history of a good response to ECT. They noted that about three-quarters of the patients treated with simulated ECT in the Leicester and Nottingha~trials had failed to improve with antidepressant drug treatment; a comparison figure for the Northwick Park trial was not available. This was an important observation, andit is unfortunatethat no definition of antidepressant drug failure was provided in either the Leicester or trials. In conclusion, there was general agreement among the trialsinthat moderate to severe depressive illness that required in-patient treatment, patients with delusions, agitation or retardation improved substantially after real bilateral ECT and this improvement was greater than that seen after simulated ECT. This is part of the evidencein support of the r e c o ~ e n d a t i o nthat the best predictor of recovery after ECT is the presencenand u ~ b eof r the typical features of depressive illness itself (Scott, 1989). It has not been established that any one symptom or sign of depressive illness identifies depressed atients who will improve substantially with real as opposed to simulated ECT; it may be that other clinical factors such as the length of the index episode, earlier treatment resistance during the index illness, and a past histor of good ECT response also influence the likelihood of recovery. The dexam~thosonesuppression test renewed interest in potential biological predictorsof recovery. There were early promising reports of an association between non-suppression beforetr~atmentand early normalization during a courseof ECT with a good clinical outcome, but later studies found
no association between non-suppression and either immediate outcome or outcome 6 months after ECT (Scott, 1989). In summary, there are no established physiological measures or tests that are superior to clinical criteria in the selection of depressed patients for whom ECT wouldbe an effective treatment.
In the author’s experience, 30% of contemporary depressed patients treated by ECT are older than 65 years, and it is disappointing to note the lack of research data pertaining to the elderly depressed patient. It is clear that the vast m a j o r i ~ of depressed patients included in the studies of real and simulated ECT were under the age of 70 years (see Table 12.2). reviewed the useof ECT in the treatment of depressive illness in old age and concluded that the elderly respond to at ECT least as well as younger people. She also noted that the small number of studies that had considered which clinical features might correlate with a good response to ECT found that it was the classical cluster of so-called endogenous features that were associated with a good outcome; however, psychic anxiety was not necessarily the unfavourable feature as suggested in the early literature on the prediction of ECT response. The largest retrospective review of elderly depressed patients treated by ECT was not included in her review; Kramer (19$7) described the 22 of whom were at least 75 clinical outcomeof 50elderly depressed patients, amer himself noted that retrospective case reports were somewhat limited, but concluded that ECT was safe and very effective in the elderly patients he described. Physical illness is more prevalent in elderly depressed patients and the safe practice of ECT is described elsewhere (Royal Collegeof Psychiatrists Special Committee on ECT, 1995).
There are no randomized comparative studies of the useof ECT in depressed adolescents or children, and this sectionis included because of a recent debate about the useof ECT in depressed young people. The literature consists of case reports and descriptive studies that began to appear in the early years after the~troductionof ECT, and these havebeen reviewed by Tomb (1991). The efficacy and potential roleof ECT in the ma~agementof depressive illness in young people cannot be established from these reports; EGT had been given safely in most cases, but there were also a few reports of prolonged seizures. It has been suggested that ECT should only be considered for young people when other treatments have failed, and specific guidance has been given about consent and the second opinions in those depressed people
under the age of 16 years (Royal Collegeof Psychiatrists Special Committee on ECT, 1995). Thepresent author would draw attention to the warning that young people may have low seizure thresholds and that young people ought to be treated with the most up-to-date ECT machines that allow treatment with low electrical doses.
The electrical stimulationproduced by contemporaryECT machines differs from that of traditional ECT machines, both inthe waveform of the stimulation and in the intensity of electrical energy Moreover, theamount of electrical energysupplied by contemporaryECT machines is preset by the treating doctor (hence the term 'constant current'), whereas it is the voltage applied to the head that was standard in traditional ECT machines (hence the term 'constant voltage'); the electrical charge delivered by traditional machines therefore varies inversely with the impedance between the electrodes. Constant-voltage ECT machines usually delivered a sine wave or modified sine wave stimulation and contemporary constant-current ECT machines deliver electrical stimulation in the form of a train of brief square-wave pulses that induce cerebral seizure activity more efficiently (see Royal College of Psychiatrists Special Committee ECT, on 1995). The ECT literature does contain reportsthat early constant-currentECT machines were not as effective as traditional constant-voltage ECT machines and also that some depressedpatients who failed to recoverwith brief-pulse stimulation subsequentlywent on to recover after treatmentwith constantvoltage ECT, although treatments were often complicated by a switch from unilateral to bilateral electrode placement (see Scott et al., 1992). A few prospective studies have found no difference in therapeutic efficacy between sine-wave and brief-pulse stimulation (see Scott et al., 1992), although Andrade et al. (1988)found that more patients recovered when treated with sine-wave stimulation. The treatments also differed, however, in another important aspect in that the electrical dose deliveredwith sine-wave stimulation was substantially higherthan that kith brief-pulse stimulation. All of the studies listed in Table 12.2 used a constant-voltage EGT machine, thus the actual doseof electrical energy deliveredto patients was not known. Oneof the few earlystudies that included any attempt to quantify the actual doses of electrical energy delivered was conducted by Robin and DeTissera (1982).lAlthough the average lengthof the convulsions were almost identical Charge = current x effective duration of current flow, In brief-pulse current, charge (milliCoulombs) = current (amperes) x pulse width (milliseconds)x frequency (Hz)x stimulus duration (seconds).
in each of the three treatment groups, the patients treated with low-energy pulses had the slowest rate of improvement and the poorest final outcome (see Table 12.3). The authors suggested that the amount of electrical energy influenced the therapeutic outcome of ECT; they speculated that the induction of a convulsion was necessary to produce a therapeutic effect with that ECT, but the convulsionitself need not necessarily produce the therapeutic itself. effect Robin and DeTissera also suggested that the convulsive threshold may have a bearing on the outcome of ECT. Sackeim et al. (1987) developed an empirical titration method to establish the minimum amount of electrical energy to induce a convulsion of defined minimum length at the outset of a course of treatment. Electrical stimulation consisted of a standard currentof 800mAmp delivered in 1.5 ms square-wave pulses, and the dose was adjusted firstlyby frequency and thenby length of stimulation. Their study was novelin that the electrical charge delivered was adjusted for each patient to just exceed the threshold, defined empirically at of the a course start of treatment. When the minimum electrical dose necessary to produce generalized cerebral seizure activity was used (so-calledthreshold dosing), bilateral ECT led to a greater rateof clinical improvement aftersix treatments and significantly greater likelihoodof recovery by the endof a courseof treatment. The treatment groups did not differ in the average length of convulsion or cerebral seizure activity measured by single-channel electroencephalogram, and the authors suggested this challenged the claim that the induction of generalized cerebral seizure activity was sufficient in itself for the antidepressant effect of ECT. Their next study tested the hypothesis that for both unilateral andbilateral electrode placements, a higher electrical dose leads to more rapid clinical improvement (Sackeimet al., 1993). The hypothesis was supported for both electrode placements (see Table 12.3), and confirmed the weak antidepressant effect of threshold right unilateralECT. The likelihood of recovery after 2.5 times threshold was greater than right unilateral ECT with a stimulation that seenin right unilateral ECT with threshold stimulation. The likelihood of recovery after high-dose right unilateral ECT was not significantly different to that after either of the bilateral treatments, but the final HRSD score 1 week after completionof treatment was significantly higher than after either of the bilateral treatments. This suggested that even high-dose right unilateral EGT may not be as efficacious as bilateral ECT. Detailed cognitive assessments were also carried out that showed that patients treated with the higher dose of unilateral treatment tooknearly twice as long to become fully orientated than did patients treated with threshold unilateral treatment. Patients treated with either of the bilateral treatments took twice as long again to become fully orientated.In general, electrode placement had a more pronounced and persistent effect on cognition than dosage; these findings were important
Chapter 12: ~lectroco~vulsive thera~y illness. Applying their criteria for response as described in Table 12.3, they found that 86% of patients who had not received an adequate course of an antidepressant drug for the index episodewent on to recover, whereas only 50% of the patients who had received an adequate trial before bilateralECT went on to recover. 'There were, however, other differences between the two groups of patients: the group who had failed to respond to an antidepressant drug was on average younger, had a lower prevalence of psychosis, and had been depressed twice as long; the authors reported that none of these was found to be statistically significant predictor of improvement in depression rating scores in regression analyses. In contrast, Kindler et al. (1991) found no statistically significant difference inthe likelihood of recovery after a course of bilateral ECT between patients who had or had not received an adequate trial of an antidepressant drug. Those who did not recover after a course of ECT were more depressed at the outset of the courseof ECT and had been depressed longer before ECT. Although thesetwo studies used the same definitionof recovery after ECT, they were not directly comparable because they used a different scale to rate the intensity of pharmacological treatment beforeECT. In conclusion, the available literature confirms the clinical impression that patients who have already failedto recover with antidepressant drug treatment can subsequently recover when treated with bilateral ECT, but that the likelihood of recovery may be lessthan among the generality of depressed patients treated with ECT. It has not been shown conclusively that this association is the resultof treatment resistance itselfrather than another related factor: for example, that patients who have already received another form of treatment beforeECT have been depressed longerthan patients who are offered ECT as a first treatment.
se In the UK, bilateral EGT is usually given twice weekly, whereas inthe USA it is usually given three times a week. Until recently, the factors that influence the onset and rate of the antidepressant effect of ECT have been little studied, although the available evidence has been reviewed (Scott elsewhere & Mihalley, 1993). It was once claimed that there is a delay in the of onset any antidepressant effect, although it has now been recognized that marked clinical improvement can occur after only three bilateral treatments (Rodger et al., 1994). Gangadhar et al. (1993)compared the clinical improvement in depressed patients of melancholic subtype who were randomly allocated to thrice-weekly bilateral ECT or twice-weekly bilateral ECT plus one simulated ECT each week. Patients treated with real ECT twice weekly improved just as quickly as
patients treated thrice weekly, although it must be noted that stimulation was given witha constant-voltage ECT machine that presumably delivereda high electrical dose. In contrast, Lerere f al. (1995) found that thrice-weeklybilateral ECT with brief-pulse stimulation about 1.5 times the seizure threshold led to a more rapid antidepressant effect after 2 weeks. Adverse cognitive effects were more markedin the group treated thrice weekly, and there was also a cumulative impairmentin some aspectsof cognition later in the course of treatment. The authors concluded that twice-weekly bilateral ECT was preferred for routine clinical practice unless the need for rapid clinical improvement was pressing.
This topic arises in two ways in clinic practice. The first concerns at what stage to abandon a course of ECT if no clinical improvement occurs. The second concerns treatment-resistant depressive illness, when an earlier course of ECT was recorded as producing no improvement. Only a few guiding principles can be given because of the lack of appropriate research. First, a depressive illness ought notto be considered as resistant to ECT untilpathe tient has received bilateralECT. Price and McAllister (1986) showed clearly that depressed patients who failed to improve with brief-pulse ~ i l a t e r aECT l may subsequently recover when treated by bilateral ECT; it is regrettable that t r e a ~ e nwas t complicated froma switch to brief-pulse Stimulation to sinewave stimulation. Neverthless, this finding was compatible with later researc that suggested that even high-dose ~ i l a t e r aECT l may notbe as efficacious as bilateral ECT. Second, ECT needs to be properly administered; that is, it must be confirmed thata generalized tonic clonic convulsion was observed after each stimulation. The use of the cuff technique and electroencephalogram have definite advantages in this respect. Third, is the there actual number of properly administered treatments. The Royal College of Psychiatrists noted that in the past it had been suggested that no more than eight properly administered treatments be given inthe absence of any clinical improvement, although the American Psychiatric Association (1990) advised that the need for more treatments be re-assessed after6-10 treatments. Devanandet al. (1991) found that most patients who do not improve with typicalof bi~ateral courses ECT have previously failed to improve with antidepressant drug treatment; s a dose of electric it^ they suggested that15 or morebilateral t r e a ~ e n t with 150% above the seizure threshold may be necessary.h contrast, Abrams (1992) believed that if no substantial antidepressant effect hadbeen observed in a case of typical depressive illness after 12 treatments, then it was prudent to withhold further treatment for several days in order to observe the patient.
The following practical issues arise from the studies which have been reviewed in this chapter. 1 The best indicators of the likelihood of recovery from depressive illness after ECT are the number and severity of symptoms and s i p s of depressive illness itself. 2 Whether ornot ECT can bring about clinical improvement in certain types of depressive illnessthat could not have been achieved by other means is yet to be clearly established.If one symptom or signof depressive illness alone can be expected to discriminate betweenpatients who improve with real ECT and not with simulated ECT, then the presence of depressive delusions is the most promisingcandidate. Patients with visible psychomotor change also improve si~ificantlymore with real asopposed to simulatedECT, but there is no consensus about whether these impairments can identify thosepatients who will improve specificallywith ECT. The lack of consensus may be because clinical features otherthan s i p s and symptoms of depressive illness are related to likelihood of recovery-for example, the length of the index episode, an earlier failure to improve with antidepressant drug treatment, and a past history of a good responseto ECT. 3 All the published comparisonsof antidepressant drug treatment and ECT in the treatment of depressive illness can be criticized because upperthe doses allowed were less than recommended today, leading to the criticism that more drug-treated patients might have recoveredif higher doses had been used. There is preliminary evidence that thrice-weekly ECT bilateral leads to a greater clinical improvementin the first 2 weeks of treatment than a moderate dose of imipramine. A common contemporary indication forEGT in depressive illness isan earlier failure to improve with antidepressant drug treatment, and it isdisappointing to note the lack of methodologically sound research on this topic. There is clear evidence that patients who have not recovered after 2oomg of impramine over 4 weeks can recover if treated subsequently with bilateral ECT. The extent to which a failure to improve with antidepressant drug treatment influencesthe likelihood of recovery afterECT is unresolved,but it seems likely that the rate of recovery is lessthan that seen amongpatients who receive ECT as a first choice treatment. has It not been clearly establishedthat treatment resistance itself reducesthe likelihood of recovery rather than the length of the index episode,and there is alsodebate about the actual definition of treatment resistance. 5 The importance of induced cerebral seizure activity in contributing to the therapeutic efficacy has been clearly established in studies that have compared the antidepressant effect of real vs. simulatedECT. It is also now
generally accepted that although cerebral seizure activity itself a necessary is prerequisite for maximum therapeuticefficacy it is not sufficient in itself to guarantee this. The electrical dose used to induce cerebral seizure activity influences the rate of antidepressant effect in both unilateral and bilateral ECT and is critical in determining the final outcome after a course of unilateral ECT. There is still an unresolved debate about how this research evidence should influence the routine clinical practice of ECT. Most critics advocate stimulus dosing where the electrical dosage is adjusted for each patient to take account of that individual patient’s seizure threshold. Fixed high-dose stimulationhas also been advocated. The debatebewill resolved if it is shown that stimulus dosing leads to fewer or less marked adverse cognitive effects than fixed high-dose stimulation. Techniques of empirical dose titration have been illustrated elsewhere (Royal College of Psychiatrists Special Committee on ECT 1995). While unilateral electrode placement has a clearly established advantage over bilateral placement in that it causes less cognitive impairment,it is yet to be established that a technique of unilateral ECT is as effective as bilateral ECT in termsof clinical improvementby the end of a course of treatment. If the minimization of adverse cognitive effects is an important consideration for an individual patient, high-dose unilateral ECT may be prescribed initially; a switch to bilateral ECT may be indicated later in the course of treatment if clinical improvement is slight orslow. Bilateral ECT with brief-pulse stimulationbecan given twice weekly withut any cumulative cognitive impairment and can be recommended as an appropriate frequencyin routine clinicalpractice. If a rapid onset of antidepressant effect is of paramount importance, the frequency of ECT can be increased to thrice weeklyat least in theearly stage of treatment. About 30% of contemporary depressed patients treated by ECT areelderly andit is disappointing thatlittle specific research has been conducted in this group. The minimization of adverse cognitive effects is clearly a pertinent issue and the advantages and disadvantages of twice- and thrice-weekly bilateral ECT have not been investigated in the elderly. Con~uation antidepressant drug treatment is essential after recovery with ECT to lessen the substantial likelihood of relapse in the early weeks and months after successful treatment. A common contemporary indication for ECT is an earlier failure to improve with antidepressant drug treatment and the appropriate pharmacological management of the continuation phase in patients who subsequently respond to ECT has not been adequately researched.
S Abrams, R., Swartz, C.M.& Vedak, C. (1991) Anti- Archives of General Psychiatry,26,57433. depressant effects of high-dose right unilateral Fink, M. (1979) Convulsive Therapy: Theory and electroconvulsive therapy. Archives of General Practice. Raven Press, NewYork. Psychiatry, 48,746-748. Freeman,C.P.L., Basson, J.V.& Crighton, A. (1978) Abrams, R. (1992) ~lectroconvulsiveTherapy, 2nd Double-blind controlled trial of electroconvule h . Oxford University Press, Oxford. sive therapy and simulated ECT in depressive Abrams, R., Taylor, M.A. & Hayman,N. et al. (1979) illness. Lancet, i, 738-740. Unipolar mania revisted. JournaZ of Affective Gangadhar,B.N.,Kapur,R.L. & KalyanaDisorders, I,59-68. sundaram, S. (1982) Comparisonof electroconAmerican Psychiatric Association (1990) Clinical vulsive therapy with imipramine in endogenous Guidelinesfor the Practiceof ECT. American Psydepression: a double-blind study. British Jourchiatric Association, Washington, DC. nal of Psychiat~,141,367-371. Andrade, C., Gangadhar, B.N., Subbakrishna, D.K. Gangadhar, B.N., Janakiramaiah, N., Subbakrishna, et al. (1988) A double-blind comparison of D.K. et al. (1993) Twice vs. thrice weekly ECT in sinusoidal wave and brief-pulse electroconvul- melancholia: a double-blind prospective comsive therapy in endogenous depression. Convulparison.JournaZ of Affective Disorders, 27, 273-278. sive Therapy,4,297-305. Greenblatt, M., Grosser, G.H. & Wechsler, H. (1964) Barton, J.L., Mehta, S. & Snaith, R.P. (1973) The Differential responseof hospitalized depressed prophylactic value of extra ECT in depressive patients to somatic therapy. American J o u ~ aofl illness. Acta Psychiatrica Scandinavica, 49, 386Psychiatry, 120,935943. Gregory, S., Shawcross, C.R.& Gill, D. (1985) The 392. Beck, A.T., Ward, C.H.,Mendelson,M. et al. (1961) Nottingham electroconvulsive therapy study: a An inventory for measuring depression. Archives double-blind comparison of bilateral, unilateral of General Psychiatry,4,561-571. and simulated ECT in depressive illness. British Benbow, S.M. (1989) The role of electroconvulsive JournaZ of Psychiat~,146,520-524. therapy in the treatment of depressive illness in Hamilton, M. (1960) A rating scale for depression. old age. British Journal of Psychiatry, 155, 147J o ~ ~of aNle u r o l o ~Neurosurgery , and Psychiatry, 152. 23J 56-62. Brandon, S., Cowley,P., McDonald, C.et al. (1984) Hermann, R.C., Dorwart, R.A., Hoover, C.W. et al. Electroconvulsive therapy: results in depressive (1995) Variation in ECT use in the United States. illness from the Leicestershire trial. British American JournaZ of Psychiatry, 152,869-875. Medical Journal, 288,22-25. Johnstone, E.C., Deakin, J.F.W., Lawler, P. et al. Buchan, H., Johnstone, E.C., McPherson, K.et al. (1980) The Northwick Park electroconvulsive therapy trial. Lancet, ii, 317-320. (1992) Who benefits from electroconvulsive therapy? Combined results of the Leicester andKindler, S., Shapira, B., Hadjez, J. et al. (1991) FacNorthwick Park trials.British Journal of Psychitors influencing response to bilateral electroconvulsive therapy in major depression. Convulsive atry, 160,355-359. Carney, M.W.P., Roth, M. & Garside, R.F. (1965) Therapy, 7,245-254. The diagnosisof depressive syndromes and theKramer, B.A. (1987) Electroconvulsive therapy use prediction of ECT response. British JournaZ of in geriatric depression. The J o u ~ aofl Nervousand Psychiatry, 111,659-674. Mental Disease, 175,233-235. Crow, T.J. & Johnstone, E.C. (1986) Controlled Lambourn, J. & Gill, D. (1978) A controlled comtrials of electroconvulsive therapy. New York parison of simulated and realECT. British JourAcademy of Sciences, 462,12-29. nal of Psychiatry, q3,514-519. Devanand, D.P., Dwork, A.J., Hutchinson, E.R.et Lerer, B., Shapira, B., Calev, A. et al. (1995) Antidepressant and cognitive effects of twice- vs, al. (1994) Does ECT alter brain structure? American J o u r ~aofl Psychiat~,151,957-970. three-times-weekly ECT. American J o u r ~ a of l Devanand, D.P., Sackeim, H.A.& Prudic, J. (1991) Psychiatry, 152,564-570. Electroconvulsive therapy in the treatmentMedical Research Council (1965) Clinical trial resistant patient. Psyc~iatricClinics of North of the treatment of depressive illness. British America, 15,905-923. Medical Journal, i, 881-886. Feighner, JP., Robins, E., Guze, S.B. et aZ. (1972) Montgomery, S.A.& Asberg, M. (1979) Anew deDiagnosic criteria for use in psychiatric research. pression scale designed to be sensitive to change.
British ~ournalof Psychiatry,134,382-389. OLeary, D., Gill, D., Gregory, S. et al. (1995)Which depressed patients respond to ECT? The Nottingham results. ~ournalof A ~ e c t ~Disorders, ve 33,
Scott, A.I.F. (1989)Which depressed patients will respond to electroconvulsive therapy? The search for biological predictors of recovery. B~t~sh ~ o u ~of aPsychiatr~, l 154,8-17. 245-250. Scott, A.I.F. (1995)Does ECT alter brain structure? Pippard,J. (1992)Audit of electroconvulsive treatA~erican ~ournalof P s y c ~ i a t152,1403. ~, ment in two National Health Service Regions. Scott, A.I.F. & Whalley, L.J. (1993)The onset and British ~ournalof P s y c h i a t ~160,621-637. , rate of the antidepressant effectof electroconvulsive therapy: a neglected topic of research. Price, T.R.P. & McAllister, W. (1986)Response of depressed patients to sequential unilateral non- ~ritish ~ournal of Ps y c h ia t ~, 162,725-732. dominant brief-pulse and bilateral sinusoidal Scott,A.I.F., Rodger, C.R., Stocks,R.H. et al. (1992) ECT. ~ournalof Clinical Ps y c h i a t ~47, , 182-186. Is old-fashioned electroconvulsive therapy more efficacious? A randomised comparative study Prudic, J., Sackeim, H.A. & Devanand, D.P. (1990) of bilateral brief-pulseand bilateral sine-wave ~edicationresistance and clinical response to treatments.British ~ o u ~of aPsychiat~, electroconvulsive therapy. P s y c h i a t ~Research, l 160,36031,287-296.
364.
Robin, A. & DeTissera, S. (1982)A double-blind Seager, C.P. & Bird, R.L. (1962)Imipramine with controlled comparisonof the therapeutic effects electrical treatment in depression: a controlled of low and high energy electroconvulsive thera-trial. J o u r o~ fM ~ e ~ t Science, a~ 108,704-707. pies. British ~ournalof P s y c h i a t ~141, , 357-366. Spitzer, R.L., Endicott, J. &x Robins, E. (1978) Robin,A.A.& Harris, J.A. (1962)Acontrolled trial Research diagnostic criteria: rationale and reliability. Archives of GeneralPsychiat~,35,773of imipramine and electroplexy.~ o u r of ~ Menff~ tal Science, 108,217-219. 782. Rodger, C.R., Scott,A.I.F.& Whalley, L.J.(1994)Is Tomb, D.A. (1991)Other organic treatments. In: there a delay in the onset of the antidepressant Child and Adolescent Psychiat~:A Co~prehensive ~ e ~ t(ed. ~ oM.oLewis), ~ effect of electroconvulsive therapy? British Jourpp. 9x4917.Williams nal of Psychiut~, 164,106-109. & Wilkins, London. Royal Collegeof Psychiatrists Special Committee West, E.D. (1981)Electroconvulsive therapy in on ECT(1995)The ~ C T ~ a n d ~ Royal o o ~ College . depression: a double-blind controlled trial. Britof Psychiatrists, London. ish medica^ ~ournal,282,355-357. Sackeim, H.A., Decina, P., Canzler, M.et al. (1987) Wing, J.K., Cooper, J.E.& Sartorius, N.(1974)The Me a s u r e ~e n tand C l a s s ~ i c a t i oof~ Psychiatric Effects of electrode placement on the efficacy of titrated, low-doseECT. A~erican ~ournal of PsyUniversityPress, S y ~ p t o ~Cambridge s. Cambridge. chiatry, 144,1449-1455. ~ Sackeim, HA.,Prudic, J., Devanand, D.P. et al. World Health Organization (1978)M e n ~ aDisorders: Glossary and Guide to their C l a s s ~ c a ~ i in on (1993)Effects of stimulus intensity and electrode placement on the efficacy and cognitive effects Accor~ancewith the ~ of thehInteri Revision ~ ~ na~ional CZass~cation of Diseases. World Health of electroconvulsive therapy. New EngZand Organization, Geneva. ~ournulof Medicine, 328,839-846.
Mood stabilizers are drugs which are used in the prophylaxis of recurrent affective disorders. They include lithium, certain anticonvulsants-particularly carbamazepine and valproate-antipsychotic drugs and monoamine re-uptake inhibitors (MARIs) and, onless a scientificbasis, calcium antagonists, monoamine oxidase i ~ i b i t o r sand thyroid hormone. The main mood stabilizers arebest known for their acute a n t ~ a n iproperties, c but some have antidepressant actions. Their prophylactic use is primarily in bipolar disorder, but some are usedalso in unipolar recurrent depression. As described in Chapter 9, antidepressants are used in the prophylaxis of unipolar affective disorders.
The distinguish~gfeature of bipolar disorder is a history of an episode of mania or hypomania. Kraepelin’s concept of manic-depressive illness was too broad, and in 1957 Leonard proposed a distinction between bipolar patients (those with a history of mania) and unipolar depressives1979). (Leonard, Some patients with recurrent depression have hypomanic episodes (not requiring hospitalization), especially on recovery from depression, and these were describedas BP-11; those with a history of mania, as BP-I (Dunneret al., 197613). Unipolar mania accounts for5-10% of bipolar disorder but has not been establishedas being different in any way from BP-I disorder. Because of the different combinations of severity of manic and depressive episodes, Angst (1985) proposed four categories: MD (in which both manic and depressive episodes are severe enough to require hospitalization), Md, mD (BP-11) and md (cyclothy~a).Unipolar depressives with a family historyof mania are et al. (1985) has used sometimes calledBP-I11 or pseudo-unipolar. Kukopulos the sequence of mood changes to distinguish patients in whom mania is followed by depression, followedby a well interval (MDI), from those with depression followed by mania (DMI), those with a continuously circular pattern (CC) and those with completely separate affective swings.
~ h a ~ t13: e r~ i t h i and u ~ other drug t ~ ~ a t ~ e n t s Ta~le13.1 Diagnostic criteria for mania. Criteria A
B
Distinct period of elation orirritability Three of the following: over-activity, increased talkativenessor pressure of speech, flight of ideas or racing thoughts, inflated self-esteem orgrandiosity (whichmay bedelusional), decreased needfor sleep, distractibility, indiscreet behaviour with poor judgement (sexual, financial, etc.) Marked impairment in occupational social or function
The manic syndromeis one of the most clearly definedin psychiatry. Table 13.1 shows the inclusion criteria for the diagnosis according DSM-W to (American Psychiatric Association, 1994).
Mania isvery rare before puberty and it is doubtfulif it ever occurs before the age of 9 years, The peak age of first hospitalization is in the late teens, the median in the mid-zos, and the meanofage first hospitalization about the age of 26 years. There have often been earlier affective episodes sufficient to cause some impairment or to receive treatment outside hospital. There is a slight secondary peak of onset amongst women aged 45-50 and first episodes of mania continue to be seen in late life. An onset over the age of 60 is more likely tobe associated with organic brain disease (Stone, 1989).
The great majority of patients have more than one episode, confirming the view that bipolar disorder is a recurrent illness. The durationof manic episodes in the pretreatment era was usually 3-12 months with a mean of 6 months. These times are usually much shortened by treatment (~ebbington Ramana, 1995. In an individual, the episode duration tends to be stable through the course of the illness but the onset may become more rapid in later episodes(Post et al., 1981).
The interval from one episode to the next tends to decrease during the first five episodes. For instance, in Kraepelin’s series the average time between the
first and second episode was 5 years and between the fifth sixth, and 2 years. owever, in an individual there is great variability in the length between episodes, and a tendency for episodes to be clustered at particular timesin the patient's life (Cutler & Post, 1982); for instance, whenthey have difficulties coping with children, or when relationships are ending. It is possible that antidepressant treatments increase the tendency to switch from depression to mania, and have altered the natural course of the illness towards more frequent episodes. The work of Angst (1985) is citedin support of this; he found an increase in the incidence of manic switches, when the era before treatment was compared with thatafter the introduction of electroconvulsive therapy (ECT) and antidepressant drugs. ~a~~~cycling. If the patient has four or more affective episodes in a year, they are said tobe in a phaseof rapid cycling. This can occur at the onset of illness but is much morec o m o n later in its course. It is also commoner in females. Antidepressant medication (particularly tricyclic antidepressants) can increase the frequencyof cycling, and withdrawal of antidepressants can restore normal cycling in such patients (Wehr & Goodwin, 1979). Some cases of rapid cycling are associated with clinical or subclinicalh~othyroidism,although a causal relationshiphas not been proved; lithium treatment might contribute to this occurring (Baueret al., 1990). Rare cases exist-about 20 in the world literature-of patients who oscillate from mania to depression and back et al., 1981). Recently, ultradian again every 48 h (ultra-rapid cycling) (Dirlich cycling has been described in otherwise typical bipolar patients; mood change in such patients occur in a matterof minutes or hours and resemble the changesin borderline personalities (Kramlinger& Post, 1996).
Death by suicide occurs in about 15-20% of cases (Black et al., 1987) and a proportion of patients become socially and economically disadvantaged. In a 15-20 years Canadian study, patients lost11% of their productive time in the & Orn, 1982). There canbe a considerable after their index admission (Bland et al., 1987). social and economic burden on the family (Fadden Life events
A prospective study showed an increased rate of life events in the month et al., before mania,but the proportion of patients so affected was small (Hunt 1992). It has been suggested that the first episode is more likely to be trigis in gered by life events thanlater episodes (Sclare& Creed, 1987). This view keeping with the suggestion that a process of 'kindling' occurs, facilitating
"77
the developmentof subsequent episodes(see below), and Post(1992b)has reviewed the possible biochemical substrates for the progressiveof effects stress in recurrent affective disorders. Insomnia or sleep deprivationhas been suggested tobe an hportant factor thatmay trigger a manic episode (Wehr et al., 1987). This may be relevant to the observation that flying overnight from west to east is more likely to lead to mania than travel in the opposite direction (Jauhar &x Weller, 1982). A short course of sedative antipsychotic may be helpful in bipolar patients with transient sleep disturbance to reduce the risk of developing mania. It is sensible for patients to keep a regular structure to biological their daily activities and exercise, as this may help to entrain their rhythms.
Manic symptoms tend to be destructive to existing relationships. Although marriages often survive individual episodes of mania, the divorce rate in bipolar disorderis very high-y% in one series, whereas it was 8% in unipolar depressive patients (Brodie& Leff, 1971).Very few of the spouses of bipolar patients have seen the patient during a manic episode prior to marriage. Fortunately, effective treatment often improves the quality of the marriage. ~ l c o ~and o l drug abuse in recurrenta~ectivedisorder
Some patients increase and some decrease alcohol or drug abuse when manic or depressed rather than euthymic (Bernadt 8s: Murray, 1986). Alcohol and stimulants suchas amphetamine and cocaine are used by patients to restore hypomania during a dysphoric phase, or to heighten existing states of elatio (Cawin & Kleber, 1986). These drugs can alter the course of bipolar disorderby triggering mania; they diminish impulse control and impair judgement and are serious risk factors forsuicide,Therefore the recognition and treatment of alcohol or drug abuse in recurrent affective patients is a matterof urgency. Cannabis has been associated with an increase in psychotic symptoms in mania (Harding & b i g h t , 1973), and with the induction of mania (Rottanberget al., 1982).
Lithium is the best-investigated drug for the treatment of mania and the prophylaxis of bipolar and unipolar affective illness.
Chapter 13: ~ i t h i and u ~ other drug treat~ents 0-controlled studies of bipolar disorder
There have been four placebo-controlledstudies of lithium in mania which used a crossover design. These were conductedin Denmark by Schouet al. (1954) and later in double-blind designs in England by Maggs ( ~ 9 6 3and ) ~ in the USA (Goodwin et al., 1969; Stokes et al., 1971). In a total of 116patients on lithium, therewas an overall responserate of 78%, much greaterthan on placebo. Lithium usually has a few days’ delay its in onset, requires2-3 weeks to approach a full effecton mania, and may take longer. The study of valproate by Eowdenet al. (1994; see below) used lithium as a comparator in addition to placebo. ies of lithium prophyl ipolar affective disorder
There have been10 major double-blind comparative tria cebo in the prophylaxisof bipolar patients. Three types used. Two studies used double-blinddisco~tinuation,patie being assigned randomly either to continue on lithium or to switch to placebo. In thestudy by Eaastrupet al. (1970)~ 55% of patients who switched to placebo, and none of those continuedon lithium, relapsedwithin 5 months. About half of the relapses were manic and half depressive. However, this design is weakened by the occurrence of lithium withdrawal mania (see below). One study used a crossover design which also involved a lithium withdrawal phase (Cundall et al., 1972). A prospective design was used by Coppen et al. (1971) in the UK and by Prien et al. (1973)in the USA and in five other studies. Overall, in 204 patients on lithium prospectively, about 35% relapsed inthe study period (whichvaried from 4 months to 3 years), compared to about 80% of 221 patients on placebo (see Goodwin & Jarnison, 1990). In these studies the efficacy of lithium was more apparent for manic than for depressive relapses. However, retrospective mirror-image studies, designed to clarify this, suggested that the efficacy in preventing depression may even be greater than that in preventing mania (Pooleet al., 1978). Lithium improves both the severity and frequency of episodes. Usually it also stabilizes the mood between major episodes. Placebo-controlled studies of lithium in the prophylaxis of unipolar depression
The prospectivestudies of lithium in the prophylaxis of unipolar depression are limited by the broader conceptof unipolar depression and by the relatively short follow-up periods which in some cases blur the distinction
between the re-emergenceof the index episode and true prophylaxis. Two of the above prospective trials included unipolar depressives and found no difference in the efficacy of lithium compared to that in bipolar patients (Coppen et al., 1971; Prienet al., 1973). Prienet al. (1984) found imipramine superiorto lithium in the prophylaxis of the most severe depressive episodes, ofand equal efficacy for moderately severe depression. Other prospective double-blind comparative trials have found lithium superior to imipramine and to placebo (Kane et al., 1982), equivalent toamitrip~line(Glen et al., 1984), and superior to maprotil~eor mianserin(see Coppen & Abou-Saleh, 1983).In a n o ~ - b l ~ d randomized controlled trial, fluvoxamine was superior to lithium in preventing recurrences (Franchiniet al., 1994). ictin~ res~onse to lit Patients with typical bipolar disorder and complete recovery between epiet al., 1993). A family historyof bipolar sodes are more likely to benefit (Grof disorder is so strongly associated with prophylactic efficacy as to question whether 'secondary' bipolar disorder ever responds to lithium; indeed,neurological signs predict a poor response (Himmelhoch et al., 1980). Patients whose first episode was manic rather than depressivedo better on lithium (Prien et al., 1984), and the MD1 pattern predicts a better response than the DM1 pattern (Faeddaet al., 1991). It is thought that a good response to lithium in acute mania or depression predicts prophylactic efficacy. Manic patients who respond tend to be classical manics rather than mixed or schizoaffective ( H i ~ e ~ oetcal., h 1976; Swam et al,, 1,986).Elated-grandiose manics showed a better response than* destructive-paranoid manics in one study (Murphy& Beigel, 1974)~ but not in another (Swam et al., 1986). Dysphoric manics were less likely to improve (Postet al., 1989). Patients with a rapid-cycling phase of illness are oftenless responsive to l i t h i u ~(Dumer & Fieve, 1974). Other factors mitigating strongly against prophylactic efficacy are poor adherence to treatment, and drug abuse. In the prophylaxis of unipolar depressive disorder, a good response to lithium is predictedby a family history of mania or of response to lithium, stable premorbid personality, low neuroticism, and good interepisode functioning (Coppen& Abou-Saleh, 1983).
Lithium has numerous effects on biological systems, especially at high concentrations. It is unclear which of these are relevant its to therapeutic effects (see Wood & Goodwin, 1987). 280
~ h a ~13: t~ e i~t h i and u ~~ t h edrug r tre~~~ents Ionic meckanis~s
As the smallest alkaline cation, lithium can substitute for sodium, potassium, calcium and magnesium in several ways. It penetrates cells via sodium and other channels,but is extruded less efficiently than sodiumby the sodiumpotassium active transport system and other transporters. Thus the cell to plasma ratio for lithium (about 0.5 in red blood cells) is much higher than that for sodium. Within thecell, lithium can interact with systemsthat normally involve other cautions, including transmitter release and second messenger systems. Second messenger systems: adenylate cyclase
Many n e u r o t r ~ m i ~ eand r s hormones (e.g. antidiuretic hormone (vasopressin) (ADH) at V2 receptors, dopamine at D1 receptors and noradrenaline at beta receptors) interact with receptors that use cyclic-AMP as the second messenger. Lithium is known to inhibit cyclic AMP production in these systems. In humans, therapeutic levels of lithium have been shown to inhibit the glucagonet al., 1984). The inhibitionby lithium induced rise in plasma c-AMP (Waller of ADH-linked adenylate cyclase (Dousa, 1974) is thought to contribute to the polyuria and polydipsia (nephrogenic diabetes insipidus) which is a sideeffect. Goitre and hypothyroidism are due in part to interference with the action of thyroid-stimulating hormone (TSH, thyrotrophin)itsatreceptors in the thyroid (Wolffet al., 1970). ~ k o s ~ k o i ~ o s iturnover tide
This second messenger system involves the formation of inositol phosphates which control intracellular-free calcium levels. The phosphoinositide cycle provides the second messengers for several neurotransmitter systems including thyrotrophin releasing hormone (TRH) in the pituitary; acetylcholine at muscarinic M-1 receptors, noradrenaline alpha, at and 5-hydroxytryptamine (5-HT) at 5-HT2, receptors. Lithium inhibits inositol monophosphatase (Berridge, 1984; Hokin, 1993) and has an inhibitory or stabilizing effect on responses (for instance, to acetylcholine). The fact that lithium preferentially inhibits phosphatidyl inositol turnover in activated neural systems (Belmaker et al., 1996) may explainits mood stabilizing properties. ~ ~ e cont 5-HT s
Lithium can potentiate both the uptake of L-tryptophan (Das Chargas Rodrigues &C Zwicker, 1985) and its conversion to 5-hydroxytryptophan
~hapter13:~ i t h i and u ~ other drug treat~ents (Rastoge & Singhal, 1977)~ the precursorof 5-HT. It can also potentiate the release of 5-HT (Sharp et al., 1gg1) and can potentiate 5-HT~~-mediated responses (Goodwinet al., 1986); for instance, in man the rise in prolactin levels induced by L-tryptophan (McGance- kat^ et al., 1992)but not that induced by d-fenfluramine ispotentiated by lithium (Shapira et al., 1992).
Lithium can block the developmentof dopamine receptor super-sensitivity that normally occurs during prolonged treatment with dopa~e-blocking (antipsychotic) drugs (Klawans & Weiner, 1976). Ithas been suggested that lithium might reduce the developmentof tardive dyskinesia in bipolarpatients on antipsychoticdrugs. Although in two studies tardive dyskinesia was more common in patients who had a briefer exposureto lithium, in a third study the opposite was found (Dinan & Kohen, 1989). Side-effects
Lithium has actions on many bodily systems, at therapeutic even doses. These are important, as some require intervention and several contribute to noncompliance. The majority of patients on lithium experience at least one sideeffect, depending on the doses used (Vestergaard et al., 1980). All patients should be informedabout side-effects and s i p s of toxicity. Thyroid
Lithium tends to reduce thyroid function. The most sensitive laboratory index, increasedTSH, occurs in 23% of patients (Transbol et al., 1978). Thyroid enlargement (goitre) developsin about 5% (Myerset al., 1985), and clinical hypothyroidism in between 5% and 10% of patients, depending upon the dose and duration of treatment (Yassa et al., 1988). Patients with pre-existing thyroid antibodies or a family history of thyroid disease are at greater risk of developing hypothyroidism (Lazarus et al., 1981), and lithium treatment can increase antibody levels (Galabreseet al., 1985; Myers et al., 1985). Kidney
Polyuria and excessive thirst with polydipsia are noted by about one thirdof patients on lithium. The condition is usually reversible, but after long-term so (Bucht et al., 1980). Givinglithium once daily as treatment it is not always opposed to divided doses, was associated with lower dailyurine volumes in some studies, although others found no difference (Bowenet al., ~991).For
Cha~ter13:~ i t ~and i u~ t~h edrug r treat~ents patients in whom a reduction in dose is not appropriate in order to avoid polyuria, the loop diuretic frusemide or amiloride, with or without potassium supplements, may be helpful. In1977, Hestbech et al. reported histological changes in patients on lithium including glomerular damage, interstitial fibrosis and tubular atrophy (focal interstitial nephropathy). However, similar findings were later made in patients who had received no lithium treatment. Much further work hasshown that during long-term treatmentwith lithium, monitored at therapeutic doses, no deterioration occurs in glomerular filtration ratein the vast majority of patients. However, occasional cases of chronic renal failure have been reported and attributed by nephrologiststo lithium, even in patients whose lithium levels have been monitored carefully; this is thought to be a rare idiosyncratic reaction to lithium &(Waller Edwards, 1989). Episodes of lithium toxicity may produce renal damage with reduced glomerular filtration rates (Johnson, G., 1984). Central nervous system
A fine tremorof the hands occurs in about 25% of patients, and is similar to that in anxiety (Gelenberg& Jefferson, 1995). It can be worsened by tricyclic antidepressants. Beta-blockers such as propanol01 (starting at lomg twice daily) reduce thisand are probably best taken intermittently (Montastruc et al., 1994).Lithium can increase extra-pyramidal (parkinsonian)side-effects in patients on antipsychotic drugs (Tyrer et al., 1980)~and can itself produce cog-wheel rigidity in a small minority of patients (Asnis et al., 1979).In contrast to neuroleptic-induced parkinsonism, this does not improve with anticholinergic drugs. Cerebellar tremor and incoordination are signs of toxicity, as are more severe forms of fine tremorand parkinsonism. ent tal and cognitiveefects
There is some objective evidence of an effect of therapeutic levelsof lithium upon memory (Kocsiset al., 1993) but not all studies show this (Kolk et al., 1993). Memory problems are frequently affirmed by patients interviewed about possible side-effects. The useof ECT in patients on lithium has been associated with acute organicbrain syndrome or prolonged confusional states, but a retrospective case-control study did not find a higher frequency of adverse effects of ECT in patients on lithium @a et al., 1996). The possible effectof lithium upon creativity was explored by Schou (1979)~ who interviewed 24 successful artists and professionals taking lithium. Some did not want to continue lithium because of this effect, but the majority, although missing some hypomanic swings, consideredthat their long-term productivityand creativity were higherunder lithium treatment. Only sixthought they were
Cka~ter13: ~ i t k i and u ~ other drug treat~ents diminished. In therapeutic doses, lithium does not impair psychomotor coordination and is not a bar to driving private motor vehicles, although a diagnosis of manic-depressive illness excludes patients from driving certain public service vehicles (Cookson, 1989) and after an admission for mania, the patient may not drive for 6-12 months according to the UK Driver and Vehicle Licensing Agency. ~ardio~ascular efects
Lithium canproduce benign reversible T-wave flattening or inversion, pat-a tern similar to that with hypokalaemia. Cardiac dysrhythmias are rare with therapeutic doses, especiallyin younger patients but sinus node arrhythmius have been described (sick sinus syndrome) (Mitchell & MacKenzie, 1982). Gaution should be exercisedin using lithium in patients with cardiac failureand the elderly. The mortalityrate among patients on lithium is similar to that of manic-depressives beforelithium was introduced (Norton& Whalley, 1984). Skin
Lithium canproduce or exacerbate acne (Kanzaki, 1991) and psoriasis (Abel, 1992). Tetracyclines should be used with caution because of their possible interaction with lithium, but retinoids can be used. Hair loss and altered texture may also occur in about 12% of patients (Llau et al., 1995)~and there may be golden discoloration of the distal nail plates. ~arathyroid,bones and teeth
Lithium produces mild increasesin parathyroid hormone level and serum calcium (Fitzpatrick & Spiegel, 1988)and Stancer and Forbath (1989) reported clinical h~erparathyroidismin patients on lithium. No long-term effects on bone have been found in animals or man (Birch et al., 1982), although it is unknown whether thisapplies to the growing bonesof children. There is no evidence of a direct effect of lithium upon the teeth, but dry mouth or increased consumptionof sweet drinks will lead to caries. ~ e t a b o l iefects c and eight gain
About 25% of patients gain morethan IO pounds in weight. The mechanism is unknown and an attempt to replicate the weight gain in volunteers was unsuccessful (Chen& Silverstone, 1990). Although increased consumption of sweet drinks is cited,an increase in food intakeand altered metabolism are also possible. Lithium produces subtle alterations in glucose and insulin
C h a ~ t e 13: r ~ i t ~and i uother ~ ~ ~tu ~ e agt ~ e n t s metabolism, and may occasionally worsen control of diabetes. Fluid retention and oedema may occur especially with higher doses. Lithium may antagonize aldosteroneand increase angiotensin levels (Stewart et al., 1988). ~~stro-intestinal system
About one third of patients experience mild abdominal discomfort, sometimes with loose motionsduring the first few weeks of treatment, especially et al., 1970; Boneet al., 1980; Vestergaardet al., 1980). with higher doses (Schou By using divided doses, these side-effects can usuallyavoided. be Sometimes a slow-release preparation is tolerated better, but occasionally these themselves irritate the lower bowel. Severe or persistent diarrhoea suggests toxicity.
~mpairmentof sexual drive, arousaland ejaculation have beenattributed to lithium, but are thought to be rare (Blay et al., 1982). Theluteinizing hormone (LH) response the to releasing hormoneLHRH is potentiated by lithium treatment (Hunter et al., 1989),similarly tothe potentiation of the TSH response to TRH (Lombardiet al., 1993). ~euromuscularjunction
Lithium reduces acetylcholine release and impairs neuromuscular transmission. Normally the safety factor in neuromuscular transmission is sufficient to overcome these effects. Lithium potentiates neuromuscular blocking agents, including succinylcholine (Diamond & Brown, 1987), and exacerbates myasthenia gravis.
Lithium can produce respiratory depression patients in with chronic obstructive airways disease, especially at toxic blood levels (Lawler& Cove-Smith, 1986). ~ l o o and d bone ~ ~ r r o ~
Lithium producesa benign reversible leucocytosis (Prakash, 1987), probably by an effect on marrowgrowth factors.
C h ~ ~ t13: e r ~ i t h i and u ~ other drug t r e a ~ ~ e n t s
There are no absolute contraindications to lithium treatment but caution is required in people with renal failure, heart failure, recent myocardial arct ti on, electrolyte imbalance, the elderly and in patients who are unreliable in taking medication or prone to dehydration,
In the first trimester, lithium carries a risk of cardiac malformations, suchas Ebstein’s anomaly in the fetus. Recent cohort studies suggest the risk of major congenital abnormalities may be 4-12%, whereas it is 2-470 in women taking other drugs not known to be teratogenic (Kallen & Tandberg, 1983; Jacobson et al.,1992; Cohenet al.,1994). Screening tests, including high-resolution ultrasound and echocardiography examination of the fetusat 16-18 weeks’ gestation, are advisable in women exposed to lithium. The anticonvulsants carbamazepine and valproate in early pregnancy are associated with brain and neural tube developmental defects in about 170 of births (Rosa, 1991). Because of the possible developmental effects upon the child, pregnancy (in bipolar patients) should if possible be managed without psychotropic drugs. Antipsychotics are probably the safest, if antimanic medication is needed) when pregnancyis planned; there is arisk of transient extrapyrarnidalsideeffects in the neonateif these are continued. ptoms of ~ithiumtoxicity
Lithium toxicity is indicated by the development of three groups of symptoms-gastro-intestinal, motor (especially cerebellar) and cerebral (Table 14.2). Nausea and diarrhoea progress to vomiting and incontinence. Marked fine tremor progresses to a coarse (cerebellar or parkinsonian) tremor, giddiness, cerebellar ataxia and slurred speech, and to gross incoordination with choreiform movements and muscular twitching (myoclonus), upper motor neurone signs (spasticity and extensor plantar reflexes), elec~oencephalo~am (EEG) abnormalities and seizures (Reed et al., 1989). In mild toxicity thereis impairment of concentration but this deteriorates into drowsiness and disorientation, and in more severe toxicity there is marked apathy and impaired consciousness leading to coma (Gadallah et al., 1988). A Creutzfeldt-Jakoblike syndrome with characteristic EEG changes, myoclonus and cognitive &z deterioration has been described, and was in these cases reversible (Smith Kocen, 1988). Patients who have recovered from acute lithiu~ toxicity canbe left with permanent cerebellar signs (Schou, 1984; Adityanjee, 1989).
286
Cha~ter13: ~ i t h i and u ~ other drug ~ r e a t ~ e n t s Table 13.2 Symptoms of lithium toxicity. Symptom
Gastro-intestinal Severity
evere Nausea
Mild Moderate
oreiform/parkinsonian Apathy vomiting Severe ma movement
Diarrhoea Vomiting
Incontinence
concentration tremor Coarse Cerebellar ataxia Slurred speech
Drowsiness Disorientation
General muscle twitching (myoclonus) Spasticityn and cerebellar dysfunction EEG abnormalities and seizures
Practical aspectsof lithium treatment A physical examination and testsof the blood and urine should, if possible, precede drug treatment or take place soon after the patient is sedated in order to elucidate any intercurrent physical illness, especially infection, and any causes of secondary mania (e.g. drugs), and to determine baseline renal, hepatic and thyroid function. Selection o ~ ~ f f t i e n t s
Maintenance treatment should be considered after a second major episode of bipolar disorderor a third episode of unipolar disorder, especially if the interval between episodes was less than 5 years. Because the intervals between the first and second episodes tend to be longer than between subsequent episodes, maintenance treatment should only be used after afirst episode if the dangers of a subsequent episode are thoughtjustify to it-for instance, if the episode was severe and disruptive, had a relatively sudden onset and was not precipitated by external factors; if the person's job is very sensitive; or there is a suicide risk. ~ i t h i blood u ~ levels and ~onitoring
The narrow gap or overlap between therapeutic and toxic blood levels of lithium necessitates careful monitoring of lithium levels, usually based on samples taken l a h after the last dose. The phamacohetics of lithium involve rapid absorption, with a peakat 4h followed by distribution in body
fluids and slow penetrationof the intracellular space and brain. Elimination is largely by the kidney, and the plasma half-life varies from7 to 20 h in physically healthy individuals but is longer in the elderly or physically unwell. Thornhill and Field(1982) found it to range from15 to 55 h in euthymic psychiatric patients. Thus, on a regular dose, steady-state blood levels would be 2 and 9 days. Many of the featuresof toxicreached after a period of between ity may reflect high intracellular rather than extracelluar levels; hence, in assessing toxicity and efficacy, clinical judgement rather than blood levels sho be paramount. Recent studies have indicated that blood levels lower than those formerly used are sufficient in prophylaxis (Coppenet al., 1983). Thus, efficacy was preserved until levels were below 0.6 mE L-I. For somepatients, lower levels than this would suffice, although Gelenberget al. (1989) found that a group withlevels of 0.8-1.omE L-I had a better outcome than a group mE L-I. In the elderly, a level0.5ofmE L-" is r e c o ~ e n d e d with levels 0.4-0.6 of (Hardy et al., 1987). ~ecommendationsvary about monitoring lithium, renal and thyroid function tests, but even in the most stable patient these tests sh be performed at least once a year, and during less stable phases lithium levels should be done frequently. The development of h~othyroidismis often signalled by weight gain and lethargy and should be distinguished from depression. Treatment with L-thyroxineis usually straightforward. The occurrence of thyrotoxicosis during lithium treatment has also been described, and there may be a rebound exacerbation when lithium is discontinued. Antide~ressantsand l i t h i u ~
Depression occurring during lithium treatment can be treated with MANS. In patients withBP-I disorder, the courseof antidepressant treatment should be gradually discontinuedas the depression improves, in order to reduce the risk of triggering a manic episode (Quitkin et al,, 1985), and to avoid the induction of rapid cycling(Wehr &:Goodwin, 1979,1987). For patients with a predominantly depressive patternof bipolar disorder (BP-11), the combination of lithium and aMARI may be more effective in preventing depression than either drug alone (Shapiroet al., ~989).Kane et al. (1982) found lithium more effective than imipramine or placebo in preventing depressive recurfences in BP-I1 patients. There is some evidence that selective serotonin reuptake inhibitors (SSRIs) such as paroxetine areless likely to trigger mania & Roberts, 1994); however, some than tricyclic antidepressants (Montgomery individuals with bipolar disorder are readily switched into mania by SSNs.
~ ~ ~ g ~Lithium o s ~ stoxicity . should be assumed in patients on lithiurn with 288
vomiting or severe nausea, cerebellar signs or disorientation. Other evidence of likely toxicity includes poor concentration, muscle twitching, tremor, dysthymia, ataxia and hyperreflexia; convulsions and renal failure areserious signs of advanced toxicity. Lithium treatment shouldbe stopped immediately, and serum lithium, urea and electrolyte levels measured. However, the severityof toxicity bears little relationship to serum lithium levels (Hansen & Amdisen, 1978), and neurotoxicity can occur with serum inlevels the usual therapeutic range (West & Meltzer, 1979). Diagnosis should be base clinical judgement and not upon the blood level. Lithium should only be restarted (at an adjusted dose) when the patient’s condition has improved, or an alternative causeof the symptoms has been found. ~ ~ e a t m eOften, ~ t . cessation of lithium and provision of adequate salt and fluids, included saline infusions, will suffice. In patients with high serum reater than 3mEL-l) or coma, haemodialysis can speed the removal of lithium and reduce the riskof permanent neurological damage (Johnson, G., 1984). Outcome. Patients who survive episodes of lithium toxicity will often make a full recovery. However, a proportion have persistent renal or neurological damage with cerebellar symptoms, spasticity and cognitive impairment. This outcome is more likelyif patients are continued on lithium, while showing signs of toxicity or during intercurrent physical illnesses (Schou, 1984). Patients with persistent neurological damage had also shown more advanced signs of toxicity(HansenAmdisen,1978).Thesigns of toxicitydevelop gradually over several days during continued lithium treatment and, in some cases, continue to develop for days after treatment is stopped. Serum lithium levels can also continue to rise after treatment is stopped, probably through release of lithium from intracellular stores (Sellers et al., 1982). ~ acto ~s ~ to ~ lei t~k i us~toxicity. ~ o s ~Conditions ~g of salt depletion (diarrhoea, vomiting, excessive sweating during fever or in hot climates) can lead to lithium retention. Drugs which reduce the renal excretion of lithium include thiazide diuretics but not frusemide or amiloride, (Alexander & Perry, 1987), certain non-steroidal anti-inflammatory drugs (ibuprofen, indomethacin, et al., 1987) but not aspirin, piroxicam, naproxen and phenylbutazone (Imbs paracetamol or sulindac), and certain antibiotics (erythromycin, metronidazole and probably tetracyclines). These drugs should be avoided if possible. If they must be used, the doseof lithium shouldbe reduced and bloodlevels monitored. In patients with serious intercurrent illnesses, especially infections, lithium shouldbe stopped or reducedin dose, and carefully monitored until the patient’s condition is stable. ~astroenteritisis particularlyliable to lead to
toxicity. In the elderly, renal functionis decreased, lower doses are required, and toxicity can develop more readily (Stone, 1989).
Combinations of high levels of lit hi^ with high dosesof antipsychotics including haloperidol have been associated with severe neurological symptom hyperthermia, impaired consciousness and irreversible brain damage (Cohen & Cohen, 1974; Loudon & Waring, 1976). The conditions reported resemble both lithium toxicity and neuroleptic malignant syndrome. Antipsychotic drugs can increase intracellular lithium levels, suggesting a possible mechanism for this interaction (Von Knorring, 1990). Subsequent series have demonstrated thesafety of combining haloperidol (up to 30 mg per day) with lithium at levels of up to ImE L-l(see Johnson et al., 1990). In practice, when combining lithium with antipsychotics, the blood levels should generallybe be advised to observe and report the maintained below1 mE L+, staff should be temporarily development of neurolo~calsymptoms, and lithium should discontinued if they develop. The combination of antipsychotics and lithium in bipolar patients can also lead to troublesome somambulism requiring reduction of doses (Charney et d., 1979).
Symptoms of anxiety, irritability and emotional lability can occur following sudden discontinuation of lithium (King & Hullin, 1983). h a double-bl~d placebo-controlled crossover study, sudden cessation of lithium in bipolar patients led to the developmentof mania 2-3 weeks later in seven outof 14 patients (Mander& Loudon, 1988). A review of all published studies suggestedthat half the bipolar patients 5 months, usuallyof mania who discontinued lithium had a recurrence within (Suppes et d . , 1991). In occasional patients, discontinuationof lithium leads to recurrent affective swings which cannotbe controlled by the reintroduco nlithium should therefore tion of lithium (Post et d., 1990). ~ i s c o n ~ u a t i of be gradual. Patients whose mood has been stableless arelikely to relapse on stopping lithium than those who have continued to show mild mood swings ('metastable'). Lithium maybe reduced at the rate of one quarterto one eighth 2 months, to minimize the risk of precipitating withof the original dose every drawal mania; further studies are needed to clarify this. There are no clear ~uidelinesfor deciding when to advise patients to stop lithium, butif they have relapsesin spite of good adherenceto treatment and satisfactory blood levels, or if they have remained well on lithium3-4 foryears, thebenefits and risks of continuing lithium shouldbe reviewed. The possible occurrence of
lithium withdrawal should be part of the information given topatients on lithium. It has been estimatedthat to balancethe risk of mania arising from abrupt discontinuation, a first-episode manicpatient would need to continue on lithium for a minimum of 2 years (Goodwin, 1994).
~ a t u r a l o u t con o ~lei t h i u ~ Dickson and Kendell (1986) reporteda threefold increase in admissions for mania in Edinburgh during the period 1970-81 when the use of lithium increased. This highlights the difficulty of delivering an effective treatment to a community. A large proportion of patients at risk do not seek treatment, and many who do, adhere poorly to lithium. In addition, there is the risk of withdrawal mania in thosewho stop treatment too abruptly-for instance, when feeling no need forit during a mild upswing of mood. A naturalistic follow-upin the USA found no differencein outcome over 18 months between bipolar patients discharged on or off lithium (Harrow et al., 1990). However, thesepatients were not randomly assigned. Under circumstances wheresteps are taken to encourage and check adherence, relapse rates and affective morbidity on lithiurn,as low as thosein controlled trials, can be achieved (McCreadie & Morrison, 1985; Coppen & Abou-Saleh, 1988). This ispart of the rationale for specialist lithium or affectivedisorder clinics. A case recordstudy of 827manic-depressive or schizoaffective patients treated with lithium for morethan 6 months showedthat mortality was reduced to that of the generalpopulation (Muller-Oerlinghausenet al., 1992).A similar finding was reportedby Coppen et al. (19g1).
In the UK the use of lithium is only about 0.8 per 1000 population, even in centres with active lithium clinics (McCreadietk Morrison, 1985);about half the patients who comenced on lithium discontinued it within 1year, but a quarter remained on it for over 10years. The patients who are less likely to adhere tend to be younger, male and to have had fewer previous episodes of illness. The reasons they give for stopping are: drug side-effects, missing periods of elation, feeling well and in no need of treatment, feeling depressed or less productive, or not wanting to depend on medication. The side-effects most often given as reasons for non-adherence are excessive thirst and polyuria, tremor, memory impairment and weight gain. Inorder to increase adherence, the doctorshould take side-effects seriously, keeplithium levels as low as possible, educate the patients and their familiesabout their illness and the use of lithium, and discuss adherence with the patient. For nonadherent patients, the regular contact provided by counselling or psycho-
therapy canbe useful and has been shown to improve compliance and affective m o r b i d i ~(Glick et al., 1985). It may be h e l p f ~ to plot a'life char the patient (Squillace et al., 1984).
Even in favourable clinical trials, lithium pro ylaxis was unsuccessful in over 30% of patients. More recent studies putthe average failurera tk Gelenbe er and alternativet r e a ~ e n tare s clearly needed (Prien Post, 1990).
e mood-stabilizing e panese psychiatristsW acute mania, even in patients who were resistant to other drugs ( ~ k u met a al., 1973,1981). Ballenger andPost (1980) independently studied carbamazepine in bipolar disorder; these authors had developed the eory that affectiveillness might involve'kindling' a process in limbic brain areas, such as they had found with cocaine-induced behavioural changes in animals. There have only been small placebo-con~rolledstudies to confirm the antimanic efficacy of c amazepine, but studies comparing carbamazepine withantipsychotic dr or lithiumshow it approximately as effectiveas lithium with about 60% of patients doing well (Post, 1990). There is some delay in its action, but lessso than with lithium. In one controlled studyof treatment-resistant depression, about one third of patients showed some acute antidepressant responseover 3-4 weeks (Post et al., 1986). XiS
The first controlled studies of the useof ~ a r b a m a z e p in ~ ethe prophylaxisof bipolar disorder were thoseof Okuma et al. (1981) and Postet al. (1~83).The drug has now been shown be to superior to placebo in one cross-over study be of similar efficacy to lithiumatinleast five studies (Post et al., 1983), and to et al., 1992; Simhandlet al., 1993). Approxi~ately of bipolar disorder (Coxhead 65% of bipolar patients appeared to show a good response (Post, 199aa). The trials of carbamazepine have been reviewed critically (Prien & Gelenberg, 1989) and a meta-analysis concluded that further controlled studies are needed to define its efficacy and place in clinical practice (Dardennes et al., 1995). Silverstone and Romans (1996) suggest that carbamazepine may have been less effective than lithium in unipolar depressive disorder in the study by Watkins et al. (1987).
The cornrnonest side-effects of ca
rnia and ~ t e v e n s - J o ~ s o
orne; these contraindi~ate
can develop suddenly, and occ H ~ o n a t ~ e and m i ~water intox
l arb amaze pine in its own blood level bolisrn of other dr
from t ~ o sre ~ a t i ~ nwho t s res~ondto carbarnaze~inediffer sornew~at ing to lithium, and a history of non-response to lithium does
~ ~ a ~13t: ~e ir t h i and u ~ other drugt ~ e a ~ ~ e ~ t s
chances of responding to carbamazepine. Patients with a history of dysphoric mania, or mixed states, can benefit from carbamazepine. Patients with no fa ily historyof mania may have a greater chance of responding to carbamazepine (Post et al., 1987). Patients withbipolar disorder secondary tobrain damage can also benefit. In contrast to lithium, rapid-cycling patients asbenefit much from carbamazepineas do other bipolar patients. In longer-term use in some patients, there maybe partial loss of efficacy by the third year of treatment, although it is not clear to what extent poor adherence to medication is responsible. More recent findings, however, have questioned the efficacy of carbamazepine in some bipolar patients resistant to lithium (Calabreseet al., 1995). Dosage
The recommended starting dose is 1oomg twice daily, increasing at weekly intervals up to 400 mg twice daily. Gradual introduction of the drugis thought to reduce the incidence of side-effects, including nausea, rashes and blood dyscrasias. Patients should be warned of side-effects, and advised particularly to report any rashes, fevers or severe sore throats, which may herald agranulocytosis and require hxnediate discon~uationof the drug. ~
~ tests o
o
~
Serum levels are sometimes helpful in monitoring carbamazepine therapy but, as in epilepsy, clinical judgement is generally more useful in deciding on dose changes. No clear relationship between blood level and antimanic effect was foundby Post et al. (1987) or in prophylaxis by Simhandl et al. (1993). A range of 15-30 pmolL-I (5-1omg L-I) is generally citedin epilepsy. The differential blood count and electrolyte levels shouldbe monitored in the first few weeks. Hyponatremia requires a reduction of dose or discontinuation. Carbamazepine shouldbe stopped imediately if the total white cell count is less than 3000 per cubic millimetre or the neutrophil count is less than 1500.
so Many patients who failto improve when taking carbamazepine alone do when lithium is added (Krarnlinger & Post, 1989). Some patients appear to benefit more from the combination than from either drug alone, although combination may-as with neuroleptics-increase the risk of lithium neurotoxicity (Shuklaet al., 1984).
94
Valproate is also effective in a proportion of manic patients including nonresponders to antipsychotic drugs and lithium. Patients who respond to valproate do not necessarily respond to carbamazepine and vice versa. In the first large parallel-group placebo-controlled study (which included onlypa1 0 of 17 patients on tients who were unresponsive to, or intolerant of, lithium), valproate improved compared to only 3of 19 on placebo (Pope et al., 1991). Most of the improvement occurred within 1-4 days of achieving therapeutic levels. Bowdenet al. (1994) compared valproate(as divalproex) to lithium or placebo in a 3-week parallel-group double-blind study of 179 patients, half of whom had been unresponsive to lithium previously. The proportion of patients showing50% improvement was greater for valproate (48%) and lithium (49%) than for placebo (25%). Valproate was as effective in rapid-cycling manics as in other manic patients and equally effective in the patients previously judged responders or non-responders to lithium. Subsequent analyses indicate that patients with mixed affective states may benefit more from valproate than from lithium. Few patients in the study had a return to normal functioning within 3 weeks, and the of place valproate in routine practice seems likelyas in the caseof lithium-to be as an adjunct to antipsychotic drugs, perhaps particularly in those manic patients with a significant mixtureof depressive symptoms, andas prophylaxis.
in be Valproate has been studied less methodically in prophylaxis but useful it may et al., 1992). The drug those whoare resistant tolithiumor carbamazepine (McElroy is effective in a large proportion of patients during the rapid-cycling periods of their illness;in one study the efficacy against mania appeared greater than against depression (Calabrese& Delucchi, 1990). Gemer and Stanton (1992), reviewing the world literature, observed a 54% response rate in 375 patients. Some clinicians conlithium in preventing recurrences sider that valproate may be more effective than of depression or rnixed states in bipolar patients, and that elderly patients may tolerate valproate better than lithium.
The drug is thought to increase the function of the inhibitory transmitter g a ~ a - a ~ o - b uacid ~ i (GABA); c there is little direct evidence for increased GABA levels but GABA-B receptors may be up-regulated by valproate. The drug may also enhance central serotonin activity, and may reduce & Calabrese, 1994). adrenocortico~ophin(ACTH) and cortisol levels (Maes
Mild side-effects which are dose-dependent include nausea, vomiting and stomach cramps: these can be managed by reducing the dose of valproate. Lethargy, hair-thi~ing,elevated liver function tests and thrombocytopaenia are also dose dependent and can be mana idiosyncratic reactions to valproate which contraindicate further treatment with valproate include liver failure, pancreatitis and agranulocytosis. A fetal valproate syndrome has also been described with cardiac and other congenital abnormalities, and with jitteriness and seizuresin the neonate (Thisted& Ebbeson, 1993).
Patients who do not respond to lithium or carbamazepine, or who cannot tolerate them, should be tried on valproate. Those who have responded to valproate during mania may benefit, and include those with mixed affective disorders. The drug should be avoided during pregnancy. Dosage
The starting dose is 200mg two to three times daily, rising by 2oomg at 3-day a plasma has intervals, to2000 mg daily according to clinical response. The drug half-life of 8-zoh, which maybe prolonged by cimetidine, Blood test~ o ~ i t o ~ i n g
The recommended plasma concentration is in the ofrange 50-15omg L+. Liver function levels should be monitored before treatment and periodically during thefirst 6 months.
Some patients benefit from the combination of lithium and valproate, having not responded to either drug alone (Mitchell et al., 1994).
Lamotrigine isan anticonvulsant with beneficial effects on mood in epilepsy. An open study in 85 bipolar patientsgave encouraging results both in bipolar
Cha~ter13: L i t h i u ~ and o t ~drug e ~ t~~at~ents depression andin mania (Cooksonet al., 1996; Walden et al., 1996), and controlled studies are underway. It is generally well tolerated, without weight gain or cognitive impairment and probably without teratogenic risk. However, up to 10% of patients develop a rash which is potentially serious. The risk of serious allergic reactions may be reduced by increasing the dose slowly in steps of 25 mg per week toan average of 15omg per day. Higher doses are required in patients on carbamazepine, and lower in patients on valproate because of the effects of these drugs on hepatic oxidase enzymes which metabolise the drug.
action Lamotrigine blocks fast sodium channels, thereby reducing theofsize potentials in nerve fibres.It has also profound effects reducing calcium currents. These actions are thought to reduce the release of neurotransmitters, especially the excitatory transmitter glutamate whichis released at cortical projections in the limbic areas of the ventral striatum, an area which also receives an inhibitory input from the mesolimbic dopamine pathway.
Because of sedative effects and long-term neurological side-effects, antipsychotic drugs should be avoided if possible for long-term use in bipolar patients. However, for those who have frequently recurring episodes, and either do not benefit from, or do not adhere to, oral medication, depot formulations of antipsychotic drugs can provide a period of stability (Lowe & Batchelor, 1990).Two studies of bipolar patients using a mirror-image design et found that during depot treatment, there were less manic episodes (White al., 1993)~ or less manic and depressive episodes (Littlejohn et al., 1994). It has been suggested that patients with bipolar disorder are particularly susceptet al., 1983). Other authors have ible to developing tardive dyskinesia (Hamra found a prevalence (about20%) similar to that among patients with chronic schizophrenia (Hunt& Silverstone, 1991). However, tardive dyskinesia may be more preventable in bipolar patients since a wider range of alternative treatments are available than for schizophrenia. In practice, the combination. of an antipsychotic drug with lithiumis very common for the treatment of mania, and the majority of manic patients are discharged on antipsychotic medication (Licht et al., 1984) and remain on it after 6 months (Keck et al., 1996). The developmentof new antipsychotics withless propensity to cause tardive dyskinesia is important for bipolar patients.
Antipsychotics are primarily antagonists at dopamine D2 receptors, and to varying extents at alphal adrenoceptors. These actions, within the limbic system probably underlie their a n h a n i c properties (Cookson et al.,1981,1985) and their effects on hormone levels in mania (Cookson, 1985).
Maintenance treatmentwith clozapine may be of value in some casesof bipolar illness resistant to other treatments including other antipsychotics (Zarate et al., 1995). Weekly blood count monitoring is required because of the risk of agranulocytosis. Risperidone, which combines potent dopamine receptor blockade with blockade of 5-HT2receptors has not been investigated in controlled trials in bipolar disorder. However, the serotonin antagonist methysergide exacerbates mania (Coppenet al., 1969), and treatment with risperidone has been associated with the onset of mania in some bipolar schizoaffective patients (Dwight et al., 1994).
The calcium channel blocker verapamil has been found effective in acute mania, although few controlled studies have been conducted (Dinan et al., 1988; Hoschl Csr Kozeny, 1989). In a double-blind comparison, verapamilin doses of 40-12omg three times daily was less effective in mania than lithium (Waltonet al., 1996).Nifedipine and diltiazem have also been reported to have antimanic effects. Pazzaglia et al. (1993) have described mood-stabilizing effects of nimodipine (ad~ydropiperidine,which has greater liquid solubility) in a group of bipolar patients during phases of ultradian cycling, when there is also evidence of low cerebral metabolism in brain scans.of Ittheoris etical interestthat the cholinesterase inhibitor physostigminehas antimanic properties; however, its side-effects preclude clinical use (Janowsky et al., 1973). Likewise, theaddition of lecithin (a precursor of acetyl choline) may potentiate other antimanic medication, but its value in clinical practice is doubtful. L-tryptophan may have antimanic properties (Prange et al., 1974) but placebo-controlled trials have not consistently supported this. The use of clonidine in neuroleptic-resistant mania has been reported but placebo controlled studies have not confirmed its value (Janicak et al., 1989).
~ h a ~ t13: e r~ i t h i and u ~ other drug treat~ents
e There is evidence from placebo-controlled studies that L-thyroxine or triiodothyronine in replacement doses can potentiate the antidepressant effect in patients with resistant depression. Howof monoamine re-uptake inhibitors ever, in long-term prophylaxis this treatment eventually loses efficacy (Wehr et al., 1988). High-dose thyroid hormone treatment has also been advocated for resistant bipolar patients, especially rapid-cyclers, but prospective placebo-controlled studies are required to establish this as an effective treatment (Bauer& Whybrow, 1990).
si Recurrent affective disorder carries a poor prognosis in terms of frequency of recurrences, divorce, physicalillness and suicide, It is also compatible with periods of successful leadership, productivity and creativity (Andreason, 1987; Jamison, 1995; Post, 1996). Treatment with mood stabilizers can reduce the of personal frequency and severityof recurrences, minimizes the disruption function, and reduces the mortality and suicide risk. Sufferers from this condition should be educated concerning the natureof their condition and the treatments available, and should be encouraged to choose effective treatment and to use it to best advantage.
therapy.American~ournalof Psychiatry, 136,1225Abel, E.A. (1992) Diagnosis of drug-induced 1226. psoriasis. Seminars in D e r m a t o l o ~XI,, 269-274. M., Thomsen, Adityanjee (1989) The syndrome of irreversible Baastrup, P.C., Poulsen, J.C. Schou, lithium-effectuated neurotoxicity. P h a ~ m a c ~ - K. & Amdisen, A. (1970) Prophylactic lithium: double-blind discontinuation in manic-deprespsychia~ry,22/81-83. Lancet, sive and recurrent-depressive disorder. Alexander, B. & Perry, P.J. (1987) Diuretics. In: ii, 326-330. Lithium Combination~ r e a t m(ed. ~ t EN.Johnson), Carbamazepine Ballenger, J.C.& Post, R.M. (1980) pp. 177-200. Karger, Basel. in manic depressive illness. a new treatment. American Psychiatric Association (1994) Diagnostic and Statis~icalManual of Mental Disorders, 4th American ~ournalof Psychiatry,137, 782-790. edn. American Psychiatric Association; Wash- Bauer, M.S. & Whybrow, P.C. (1990) Rapid-cycling bipolar affective disorder. 11. Treatment of ington, DC. refactoryrapid-cyclingwithhighdose Andreasen, N.C. (1987)Creativity and mental illlevothyroxine: a preliminary study.Archives of ness: prevalencerates in writers and their firstGeneral Psychiatry,47, 435-440. degree relatives. American Journal of Psychiatry, Bauer, M.S., Whybrow, P.C.& Winokur, A. (1990) 144,1288-1292. Rapid-cycling bipolar affective disorder. Assomania: Angst, J. (1985) Switch from depression to ciation with Grade I hypothyroidism. Archives a recordstudy over decades between1920 and of General Psychia~ry,47, 427-432. 1982. psycho pa tho lo^, 18,140-155. Asnis, G.M., Asnis, D., Dunner, D.L.& Fieve, R.R. Bebbington, P. & Ramana, R. (1995). Theepidemiology of bipolar affective disorder. Social Psy(1979)Cogwheel rigidityduring chronic lithium
chiatry and Psychiatr~c~ p i d e m i o l o30, ~ , 279292. Y., Agam, G., Levine, J. Belmaker, R.H., Berdusky, Kofman,0. (1996)How does lithium work in manic depression? Clinical and psychological correlates of the inositol theory. Annual re vie^ of ~ e d i c i n e47, , 47-56. Bernadt, M.W. & Murray, R.M. (1986)Psychiatric disorder, drinking and alcoholism. British Journal ofPsyckiatry, 148,393-400. Berridge, M. (1984) Inositol triphosphate and diaglycerol as second messengers. Biockemical
patients treated with lithium. A ~ e r ~ c a n J o ~ r n a i of Ps y c h ia t ~,142,1318-1321. Calabrese, J.R., Bowden, C., &r Woyshville, M.J. (1995) Lithium and anticonvulsants in the treatment of bipolar discords. In: Psychop h a r ~ a c ~The ~ oFourth ~: Genera~ionof Progress (eds F.E. Bloom & D.J. Kupfer), pp. 1099-1111. Raven Press,New York. Charney, D.S., Kales, A., Soldatos, C. & Nelson, J.C. (1979)Sornnambulistic-like episode, secondary to contained lithium-neuroleptic treatment. ~ r i t i s h J o uof~Pa sl ~ c h i a t135, ~ / 418-424. Chen, Y. & Silverstone, T.(1990) Lithium and Birch, N.J., Horsman, A. & Hullin, R.P. (1982) weight gain. ~nternationaiClinical Psychopharm a c o l o ~5,217-225. , Lithium, bone and body weight studies in longCohen, W.J. & Cohen, N.H.(1974)Lithium carbonterm lithium-treated patients and in the rat. ate, haloperidol, and irreversiblebrain damage. ~ e u r o p s y c ~ ~ o b i8/86-92. olo~, & Nasraliah, M.A, (1987) ~ournalof the A m e ~ c a nMedical Association, 230, Black, D.W., Winokur, G. 1283-1287. Suicidein sub-types of major affectivedisorder: a comparison with general population suicide Cohen, L.S., Friedman, J.M., Jefferson, J.W.et al. (1994) A reevaluation of risk of in utero expo~ / 878mortality. Archives of General P s y c h i a ~44, l Ame~icunMedical sure to lithium. J o ~ r n a of 880. Association, 271,146-150. Bland, R C . 8r Om,H. (1982)Course and outcome Cookson, J.C.(1989)Manic-depressiveillness and of Psyof affective disorders. Canadia~ Jo~rnai e 7, 105-108. driving. Travel~ e d i c i nlnternatio~ai/ chiatry, 27,573578. Blay, S.L., Toledo Ferraz, M.P. & Calil, H.M. (1982) Cookson, J.C., Silvestone,T. & Wells, B. (1981)A double-blind comparative clinical trial of Lithium-induced male sexual impairment: two pimozide and chlo~romazine in mania: a test 43, 497case reports.J o u ~ aofl Clinical Psyckiat~, of the dopamine hypothesis. Acta psychiatric^ 498. Scandinavica,64, 381-397. Bone, S., Rose, S.P., Dunner, D.L. & Fieve, R.R. T., Williams,S. 8r Besser (1980) Incidence of side effects in patients on Cookson, J.C./ Silverstone, G.M. (1985) Plasma corticol levels in mania: long-term lithium therapy. Amerian ~ o ~ r nofa l associated clinical ratings and change during Psychia~ry,137, 103-104. treatment with haloperidol. Brikish Journal of Bowden, C.L., Brugger, A.M., Swann, A.C. et al. Ps y c h ia t ~, 146,498-502. (1994) Efficacy of Divalproex vs. lithium and J.C., Kundu, S., Anderson, J. et al. (1996) l placebo in the treatment of mania. ~ o ~ ~ofn a Cookson, Lamotriginein treatment-refractorybipolar disfhe American ~ e d i c aAssociation, l 271,918-924. 10 (Suppl. order. J o ~r n aofl Psychopharmacolo~, owen, R.C., Grof, P. 8r Grof, E. (1991) Less fre3)A39. quent lithium administration and lower urine Coppen, A. & Abou-Saleh, M.T. (1983) Lithium in ofPsychiatry, 148,189volume. American~ournal prophylaxis of unipolar depression: a review. 192. ~ o ~of ~the nRoyal ~ Society l of ~ e d i c i n e76, , 297Brodie, H.K.H.& Leff, M.J. (1971)Bipolar depres301. sion: a comparative study of patient characterof Psychiat~,127, 1086- Coppen, A. & Abou-Saleh, M.T. (1988) Lithium istics. A~erican ~ournal therapy: from clinicaltrials to practical manage1090. 78,754-762. ment. Acta PsychiatricaScandi~avica, Bucht, G., Wahlin, A., Wentzel,T. & Wimblad, B. Noguera, (1980) Renalfunction and morphology in long- Coppen,A., Prange, A.J., Whybrow, P.C., Methysergide in mania: a R. & Paez, J.M. (1969) term lithium and combined lithium-neuroleptic controlled trial. Lancet, 2, y@-340. i c a381, treatment. Acta ~ e d i c a S c ~ n ~ i n a v208, Coppen, A., Noguera, R., Bailey, J.et al. (1971)Pro385. phylactic lithium in affectivedisorders: controlCalabrese, J.R. & Delucchi, G.A. (1990)Spectrum led trial. Lancet, ii, 275-279. of efficacy of valproate in 55 patients with rapidcyclingbipolar disorder.American ~ournalof Psy- Coppen, A., Abou-Saleh, M.T.,Milln, P., Bailey,J. & Wood, M.K. (1983) Decreasing lithium dosage chi at^, 147,431-434. reduces morbidity and side-effects during proCalabrese, J.R., Gulledge,A.D., H h , K. et al.(1985) isor~~s, phylaxis.~ o u ~ofaAl ~ e c ~ i v e ~5,353-362. Autoimmune thyroiditis in manic-depressive
~ h a ~ t1e3:y ~ i t h i and u ~ other drug t y e ~ t ~ e n t s affective illness. Biological Psychiatry,11,31-42. Coppen, A., Standish-Barry, H., Bailey, J. et al. (1991) Does lithium reduce the mortality of Dwight,M.M.,Kech, P.E., Stanton, S.P., recurrent mood disorders? Journal of Afective Strakowski, S.M. & McElroy, S.L. (1994) Antidepressant activity and mania associated with ~isordeys, 23,1-7. risperidone treatment of schizo-affective disorCoxhead, N., Silverstone, T. & Cookson, J.C.(1992) Carbamazepine versus lithium in the prophyder. Lancet, 3441554-555. laxis of bipolar affective disorder. Acta Psychia- Elphick, M., Yang, J.D. & Cowen, P.H. (1990) Effects of carbamazepine on dopamine and trica Scandinavica,$5,114-118. serotonin-mediated neuroendocrine responses. Cundall, R.L., Brooks,P.W. & Murray, L.G.(1972) A controlled evaluation of lithium prophylaxis Archives of General Psychiat~,47,135-143. in affective disorders.Psychological Medicine,2, Fadden, G., Bebbington, P. & Kuipers, L. (1987) Caring and its burdens: a study of spouses of 308-311. depressed patients.British Journalof Psychiatry, Cutler, N.R.& Post, R.M.(1982) Life courseof illness in untreated manic-depressive patients. 151,660-667. Faedda, G.L., Baldessarini, R.J., Tohen, M., Comprehensive Psychiatry,23,101-115. Strakowski, S.M. & Waternaux, C. (1991) EpiDardennes, R., Even, C., Bange, F. et al. (1995) Comparison of carbamazepine and lithium in the sode sequence in bipolar disorder and response to lithium treatment. American Journa~of Psychiprophylaxis of bipolar disorder: a meta-analysis. British Journalof Psychiatry, 166,378-381. atry/ 148,1237-1239. Das Chargas Rodrigues,R & Zwicker, A.P. (1985) Fitzpatrick, L.A. & Spiegel, A.M. (1988) Normal Chronic lithium prevents REM sleep-deprivaregulationof parathyroid hormone secretion. In: tion-induced increased responsiveness to Lithium and the Endocrine System (ed. F. N. apomorphine.~ournal of Pharmaceutics and PharJohnson), pp. 194-201. Karger, Basel. Franchini, L., Gasperini,M. & Smeraldi, E.(1994) m a c o l o ~37,210-211. , A 24-month follow-up study of unipolar subDiamond, B.I. & Brown, R.L. (1987) Anaesthetic jects: a comparison between lithium and fluvoxaagents and associated medications.In: Lithium mine. Journal of Afective Disorders, 32,225-231. Combination ~ y e a t ~ e (ed. n t F.N. Johnson), pp. Gadallah,M.F., Feinstein, E.I.& Massry, S.G.(1988) 217-232. Karger, Basel. Lithium intoxication: clinical course and theraDickson, W.E. & Kendell, R.E.(1986) Does maintenance lithium therapy prevent recurrences of peutic considerations. Mineral Electrolyte Metabomania under ordinary clinical conditions? Psylism, 14,146-149. Gawin, F.H. & Kleber, H.D. (1986) Abstinence chological Medicine,16,521-530. Dinan, T.G.& Kohen, D.(1989) Tardive dyskinesia symptomatology and psychiatric diagnosis in in bipolar affective disorder: relationship to cocaine abusers. Archives of General Psychiatry 43, lithium therapy.British Journalof Psychiatry, 15, 107-113. Gelenberg, A.J. & Jefferson, J.W. (1995) Lithium 55-57. Dinan, T.G., Silverstone, T. & Cookson, J.C.(1988) tremor. Journal of Clinical Psychiatry,56,283-287. Cortisol, prolactin and growth hormone levels Gelenberg, A. J., Kane, J.M., Keller, et al. M.B. (1989) with clinical ratings in manic patients treated Comparison of standard and low serum levels with verapamil. ~nternatio~al Clinical Psychoof lithium for maintenance treatment of bipolar pharmacolo~,3,151-156. disorder. New England Journal of Medicine, 321, Dirlich, G., Kammerloher, A., Schulz, H. et al. 1489-1493. (1981) Temporal coordination of rest-activity Gerner, R.H. & Stanton A. (1992) Algorithm for cycle: body temperature, urinary free cortisol patient management of acute manic states: and mood in a patient with @-hour unipolar lithium, valproate or carbarnazepine? Jouynal of depressive cycles in clinical and time-cue-free Clinical Psychopharmacolo~/12 (Suppl.), 57-63. environments. Biological Psychiat~/ 16,163-179. Glen, A.I.M., Johnson, A.L. & Shephered,M.(1984) Dousa, T.P. (1974) Interaction of lithium with Continuationtherapywithlithiumand vasopressin-sensitive cyclic AMP system of huamitriptyline in unipolar depressive illness: a man renal medulla. Endocrinology/ 95, 1359randomised double-blind controlled trial. Psychological Medicine,14,y-50. 1366. et al. (1985) A Dunner, D.L.& Fieve, R.R. (1974) Clinical factors Glick, I.D., Clarkin, J.F., Spencer, J.H. in lithium prophylaxis failure.Archives of Gencontrolled evaluation of inpatient family intervention: Preliminary results of the six-month eral ~ s y c h i a t30~22g-233. ~, follow-up. Archives of General Psychiatry, 42, Dunner, D.L., Gershon, E.S. & Goodwin, F.K. (1976b) Heritable factors in the severity of 882-886.
~ h a ~ t1e3:r ~ i t ~and i uother ~ drug ~ r e a t ~ e n ~ s tive disorders:a controlled,double-blind study. Goodwin, F.K. & Jamison, K.R. (1990) Manic-Depressive Illness.Oxford University Press, Oxford. Biological Psychiatry,25,128-140. Goodwin, F.K., Murphy, D.C. & Bunney, W.F. Hunt, N. & Silverstone, T. (1991) Tardive (1969) Lithium carbonate treatment in depresdyskinesia in bipolar affective disorder: a catchment area study. International Clinical Psychosion and mania: a longitudinal double-blind pharmacolo~,6,45-50. study.Archives of GeneralPsychiatry, 21,486-496. Goodwin, G.M. (1994)Recurrence of mania after Hunt, N., Bruce-Jones, W. & Silverstone, T. (1992) lithium withdrawal: implications for the useof Life events in bipolar affective disorder.Journal of Afective Disorders, 25,13-20. lithium in the treatment of bipolar affective disHunter, R., Christie, J.E., Whalley, L.J.et al. (1989) order. British Journal of Psychiatry,164,149-152. Luteinizing hormone responses to luteinizing Goodwin, G.M., De Souza, R.J., Wood, A.J. & Green, A.R. (1986) The enhancement by lithium hormone release hormone (LHRH) in acute of the 5-HTI,-mediated serotonin syndrome in mania and theeffects of lithium on LHRHand the rat: evidence for a post-synaptic mechanism. thyrotropin releasing hormone tests in volunPsychopharmacolo~, go, 488-493. teers. PsychologicalMedicine, ~9~69-77. M. Grof, P., Alda, M., Grof, E., Fox, D.& Cameron, P. Imbs, J.L., Danion, J.M., Schmidt, M., Welsch, (1993) The challenge of predicting responseto & Singer,L. (1987)Non-steroidal anti-inflamash of u~ ~reatment(ed. stabilising lithium treatment.~ r ~ t iJournal tory drugs. In:~ i t h i Combination F. N. Johnson), pp.201-216. Karger, Basel. Psychiatry,163 (Suppl.21), 16-19. Hamra, B., Nasrallah, H., Clancy, J. & Finn, R. Jacobson, S.J., Jones, K.,Johnson, K. et al. (1992) (1983)Psychiatric diagnosis andrisk fortardive Prospective multi-centre study of pregnancy outcome after lithium exposure during first dyskinesia.Archives of General Psychiatry,40, 346trimester. Lancet, 339,530-533. 347. Hansen, H.E. & Amdisen, A. (1978) Lithium in- Jamison, K.R. (1989)Mood disorders andseasonal patterns in British writers and artists. Psychitoxication (Report of 23 cases and a review of atry, 52,125-134. loo cases fromthe literature). Quarterly Journal e t Picador, Jamison, K.R. (1995) An ~ n ~ u iMind. of Medicine/ 47,123-144. London. Harding, T. & Knight, F. (1973) Marijuana-modiJ.E. Janicak, P.G., Sharma, RP., Easton, M., Comaty, Archives of General Psychiatry, 29,635fied mania. & Davis, J.M. (1989) A double-blind, placebo 637. controlled trial of clonidine in the treatment of Hardy, B.G., Shulman, K.I., Mackenzie, S.E., , acute mania. Psychopharmacology ~u lle t in25, Kutcher, S.P.& Silverberg, J.D.(1987) Pharmacokinetics of lithium in the elderly. Journal of 243-245. Janowsky, D.S., El-Yousef, M.K., Davis, J.M. & Clinical Psychopharmacolo~, y, 153-158. Sekerke, H.J. (1973) Parasympathetic suppresHarrow, M., Goldberg, J.F., Grossman, L.S. & sion of manic symptoms by physostigmine. Meltzer, H.Y. (1990)Outcome in manic disorder: Archives of General Ps y c h ia t ~, 28,542-547. a naturalistic follow-up study. Archives of GenJauhar, P. & Weller, M.P.I. (1982)Psychiatric moreral P s y c h i a t ~47,665471. , bidity and h e zone changes:a study of patients Hestbech, J., Hansen, H.E., Amidsen, A. & Olsen, from Heathrow Airport.Bri~ishJournal of PsyS. (1977) Chronic renal lesions following longchia~ry,140,231-235. term treatment with lithium. Kidney InternaJha, A.K., Stein, G.S. & Fenwick, P. (1996) Negational, 12,205-213. Himelhoch, J.M., Mulla, D., Neil, J.F., Detre, T.P. tive interaction between lithium and electroconvulsion therapy: a case-controlled study. & Kupfer, D.J. (1976) Incidence and significance British Journal of Psychiatry,168,241-243. of mixed affectivestates ina bipolarpopulation. Johnson, D.A.W., Lowe, M.R. & Batchelor, D.H. Archives of General Psychiatry/33,1062-1066, Himelhoch, J.M., Neil, J.F., May, S.J., Fuchs, C.Z.(1990) Combined lithium-neuroleptic therapy for manic-depressive illness. Human Psycho& Licata, S.M. (1980) Age, dementia, dyshesias, 262-297. p h a r ~ c o l o5~ (Suppl.), , and lithium response.American Journal of PsyJohnson, EN. (1984)The History of ~ ~ t h ~ u m T ~ r a p y . chi at^, 137,941945H o h , L.E. (1993)Lithium increases accumulation Macmillan, London. Johnson, G. (1984)Lithium.MedicalJournal of Ausof second messenger inositol 1,q,5-trisphosphate tralia, 141,595-601. in brain cortex slices in species ranging from G.F.S., Hunt, G.E., Duggin., G.G., Hoovalt, , mouse tomonkey. Advanced Enzyme ~ e ~ l a t i o n Johnson, J.S. & Tiller, D.J. (1990) Renal function and 33, 299-312. lithium treatmentinitial and follow-uptests in Hoschl, C.& Kozeny, J. (1989) Verapamilin affec-
manic depressive patients. Journal of Afective cine, 11,8592. Leonard, K. (1979) The Class~cationo ~ E n ~ o g ~ o ~ s Disorders, 6,249-263. Lithium and pregKallen, B. & Tandberg, A. (1983) Psychoses (ed. E. Robins; trans. R. Berman). nancy: a cohort study on manic-depressive Irvington Publishers, New York. ~i~ 68,134-139. women.Acta P s y c h ~S~ndinavica, Licht, R.W.,Gouliea, G., Vesterfaarl P., Dybbro, J., Kane, J.M., Quitkin, F.M., Rifkin, A., RamosLumd, H. & Membdet, L. (1994) Treatment of manic episodes in Scandinavia: the use of Lorenzi, J.R., Nayak, D.P. & Howard, A. (1982) Lithium carbonate and imipramine in the neuroleptic drugs in a clinical routine setting. prophylaxis of unipolar and bipolar I1 illness. Journal of A f e c t i ~ Disorders, e 32,179-185. Archives of General Psychiatry,39,1065-1069. Littlejohn, R., Leslie, F. & Cookson, J.C. (1994) Kanzaki, T. (1991)Acneiform emption induced by Depot anti-psychotics in theprophylaxis of bi, lithium carbonate. Journffl of D e r m a t o l o ~18, polar disorder. British Journal of Psychiatr~,65, 827-829. 481-483. Keck, PE.,McElroy, S.L., Strakowski, S.M., Llau, M.E., Viraben, R. & Montastruc, J.L. (1995) Balistrem, T.M., Kizer, D.L. & West, S.L. (1996) Drug-induced alopecia: review of literature. Therapie, 50,145-150. Factors associated with maintenance anti-psychotic treatment of patients with bipolar disor- Lombardi, G., Panza, N., Biondi, B. et al. (1993) 57, , 147-151. Effects of lithium treatment on hypothalamicder. Journal of Clinical P s y c h i f f t ~ Ketter, T.A., Post, R.M. & Worthington, K. (1991) pituitary-thyroid axis: a longitudinal study. Journal of Endocrinological In~estigfftions, 16, Principles of clinically important drug interactions with carbamazepine:part I. Journal of Clini259-263. cal P s y c h o p h a r ~ a c o l o1 ~1,,q8-203. Loudon, J.B. & Waring, H. (1976) Toxicreactions to lithium and haloperidol. Lancet, ii, 1088. King, J.R. & Hullin, R.P. (1983)Withdrawal symptoms from lithium: four case reports and a ques- Lowe, M.R.& Batchelor, D.H.(1990) Lithium and neuroleptics in the management of manic-detionnaire study. British Journal of Psychiatry,143, pressive psychosis. Human Psychopharmacolo~, 30-35. Klawans, H.L. & Weiner, W.J. (1976)The pharma5,267-274of cology of choreatic movements. Progress in Maes, M. & Calabrese, J.R. (1994) Mechanisms action of valproate in affective disorders. In: N e u r o ~ i o l o6,49-80. ~, Kocsis, J.H., Shaw, E.D., Stokes, P.E. et al. (1993) Anti-con~ulsan~s in Mood Disorder(eds R. T.Joffe & J. R. Calabrese) pp. 93-110. Marcel Dekker, ~europsychologicaleffects of lithium discontinuation. Journfflof Clinical Psychopharmacolo~, New York. Maggs, R. (1963) Treatment of manic illness with 13, 268-275. Kolk, A., Kathmann, N. & Greil, W. (1993) No lithium carbonate. Br~tishJournal of Psychiatry, short-term changes of cognitive performance log, 56-65. and mood after single doses of two different Mander, A.J. & Loudon, J.B. (1988) Rapid recurlithium retard preparations. Pha~acopsychi~try, rence of mania followingabrupt discontinuation of lithium. Lancet, ii, 15-17. 26,235-239. McCance-Katz, E., Price, L.H., Charney, D.S. & Kramlinger, K.G. & Post, R.M. (1989) Adding Heninger,G.R. (1992)Serotonergicfunction durlithium carbonateto carbamazepine:anti-manic ing lithium augmentation of refractory depresefficacy in treatment-resistant mania. Acta Psy108,9397. sion. Psychopharmacolo~-~erlin, chiatrica Scffndina~ica, 79,378-385. Kramlinger K.G. & Post, R.M. (1996) Ultra-rapid McCreadie, R.G. & Morrison D.P. (1985) The impact of lithium in South-West Scotland:British and ultradian cycling in bipolar affective illness. J o u ~ foff lPsychiatry,146,7074. British J o u ~ aofl Psychiatry,168,314-323. & MullerKukopulos, A., Minnai, G. McElroy, S.L., Keck, P.E. Pope, H.G. et al. (1992) Oerlinghausen, B. (1985) The influence of maValproate in the treatment of bipolar disorder: nia and depression on the pharmocokinetics of literature review and clinical guidelines. Journal of Clinical P ~ r m o c o l o12,42S-52S. ~, lithium: a longitudinal single-case study. Journal of A8ective Disorders, 8,159-166. Mitchell, J.E. & MacKenzie, T.B. (1982) Cardiac JourLawler, P.G.,& Cove-Smith,J.R. (1986).Acute reseffectsof lithium therapy in man: a review. piratory failure following lithium intoxication. nal o f C l ~ n ic Psychiatry, ~1 43,47-51. Anaesthesia, 41, 623-627. Mitchel, P,Withers, K., Jacobs, G.et al. (1994)Combining lithium and sodium valproate for bipoLazarus, J.H., John R., Bennie EH, Chalmers, RJ, Crockett, G.(1981)Lithium therapy and thyroid lar disorder. Australian and New Zealand Journal of Psychiatry,28: 141-143. function: a long-term study. PsychologicalMedi-
303
Cha~ter13: ~ i t h i and u ~ other drug t r e a t ~ ~ n t s W.E. Jr (1981) Slow and rapid onset of manic Montastruc, J.L., Llau,M.E., Rascol, 0. & Senard, J.M. (1994) Drug-induced parkinsonism: a reepisodes: implications for underlying biology. view. Fundamentals of Clinical Pharmacolo~,8, Psychiatry Reseaych,4,229237. Post, R.M., Uhde,T.W., Ballenger J.C.& Squillace, 293-306. Montgomery, S.A.& Roberts,A. (1994)SSRIs: well K.M. (1983) Prophylactic efficacy of carbamatolerated treatment for depression. ~ u m a n zepine in manic-depressive illness. American Psychopharmacolo~,g, S7-S10. ~ournalof Psychiatry, 140,1602-1604. Muller-Oerlinghausen,B., Ahren, B., Grof, E. et al. Post, R.M., Uhde, T.W., Roy-Byme,P.P. & Joffe, R.T. (1992) The effects of long-term lithium treatment(1986) Antidepressant effects of carbamazepine. on the mortality of patients with manic-depres-American Journalof Psychiatry, q3,29-34. sive and schizo-affective illness. Acta Psychiatrica Post, R.M., Uhde, T.W., Roy-Byrne,P.P. & Joffe, R.T. Scandinavica, 86,218-222. (1987) Correlates of antimanic responses to Murphy, D.L.& Beigel, A. (1974) Depression, elacarbamazepine. Psychiatry Research,21,71-84. tion, and lithium carbonate responses in manic Post, R.M., Rubinow D.R., Uhde, T.W. (1989) patient sub-groups. Archives of General PsychiDysphoric mania: clinical and biological correlates. Archives of General Psychiatry,46,353-358. atry, 31,643448. Post, R.M., Leverich, G.S., Rosoff, &A.S. Myers, D.H. Carter, R.A., Burns, B.H., Armond, Altshuler, A., Hussein, S.B. Chengapa, V.K. (1985) A proL.L. (1990) Carbamazepine prophylaxis in respective studyof the effects of lithium on thyfractory affective disorders: a focus on long-term roid function and on the prevalance of thyroid follow-up.Journal of ClinicalP s y c h o p ~ a r ~ c o l o ~ , antibodies. Psychological Medicine,15,5541. 20,318-327. Norton, B. Whalley, L.J. (1984) Mortality of a Post, R.M., Weiss, S.R.B., Chuang, D.M. & Ketter, T.A. (1994) Mechanism action of of lithium-treated population. British ~ournalof Psychiatry, 145,277-282. carbamazepine in seizure and affective disorOkuma, T., Kishimoto, A., Inoue, K. et al. (1973) ders. In: Anticonvulsants in Mood-Disorders(ed. Anti-manicandprophylacticeffects of R.T. Joff & J.R. Calabrese), pp. 43-92. Marcel carbamazepine (Tegretol) on manic depressive Dekker, New York. psychosis: a preliminary report. Folia Psych~trica Prakash, R. (1987) Blood. In: Depression and Ma~ e u r o lJapan, o ~ 27,283-297. nia: Modern~ithiumTherapy (ed. F. N. Johnson), Okuma, T., Inanaga, K., Otsuki,S. et al. (1981) A pp. 218-220. IRL Press, New York. preliminary double-blind study of carbama- Prange, A.J., Wilson, IC., Lynn, C.W., Alltop, L.B. & Stikeleather, R.A. (1974) L-tryptophan in mazepine in prophylaxis of manic-depressive nia: contribution to a permissive hypothesis of illness. Psychopharmacolo~,73,95-96. affective disorders. Archives of General PsychiPazzaglia, P.J., Post, R.M., Ketter,T.A. et al. (1993) atry, 30,5642. Preliminary controlled trial of nimodipine in ultra-rapid cycling affective dysregulation. Psy- Prien, R.F. & Gelenberg, A.J. (1989) Alternatives to lithium for preventive treatment of bipolar chiatry Research,49,257-272. disorder. American Journal of Psychiatry,146,840Pope, H.G., McElroy, S.L., Keck, P.E. & Hudson, 848. 3.1. (1991). Valporate in the treatment of acute Prien, R.F., Caffey, E.M.Jr & Klett, C.J. (1973) Promania: a placebo-controlled study. Archives of phylactic efficacy of lithium carbonate in manicGeneral Psychiatry,48,6268. depressive illness.Archives of General Psychiatry, Post, F. (1996) Verbal creativity, depression and alcoholism: an investigation of one hundred 28, 337-341. Prien, R.F., Kupfer,D.J., Mansky, P.A. et al. (1984) American and British writers.British ~ournalof Drug therapy in the prevention of recurrences Psychiatry, 168,545-555. in unipolar and bipolar affective disorder: rePost, R.M. (1990) Prophylaxisof bipolar disorders. port of the NIMH Collaborative Study Group ~ n t ~ n a t i o nReview al of Psychiatry, 2,227-320. comparing lithium carbonate, imipramine and Post, R.M. (1992a). Anticonvulsants and novel a lithium c a r b o n a t ~ ~ p r a m i combination. ne drugs. In:~ a n ~ ~of oAfective o k Disorders (ed. E.S. Archives of General Psychiatry,41,1og6”1104. Paykel), pp. 387-417. Churchill Livingstone, Quitkin, EM., Kane, J., Rifkin, A., Ramos-Lorenzi, Edinburgh. J.R. & Mayak, D.V. (1985) Prophylactic lithium Post, R.M. (199213) Transduction of psychosocial carbonate with and without imipramine for bistress into the neurobiologyof recurrent affecpolar 1 patients: a double-blindstudy. Archives tive disorder.A ~ e r i c a n J o ~ of r nPsychiat~, al 149, of General Psychiatry,38,902-907. 999-1010. Post, R.M., Ballenger, J.G., Rey, A.C. & Bunney, Rastoge, R.B.& Singhal, R.L. (1977) Lithium modi-
C ~ a ~ t13: e r ~ i t ~and i u ~ t~ ~drug e r treat~e~ts fication of behavioural activity and brain ment of bipolar disorder.Drugs, 51, 367-382. biogenic amines in developing hyperthyroid Simhandl, C., Denk, & E.Thau, K.(1993) The comrats. Journal of Phar~acologyand Experi~ental parative efficacy of carbamazepine low and high serum level and lithium carbonate in the ~herapeutics,201~92-102. prophylaxis of affective disorders. Journal of Reed, S.M., Wise, M.G. & Timmerman, I. (1989) Choreoathetosis: a sign of lithium toxicity. JourA~ectiveDisorders, 28,221-231. nal of Neuropsychiatryand Clinical Neurosciences, Smith, S.J.M. & Kocen, R.S.(1998) ACreutzfeldtJakob like syndrome due to lithium toxicity. Jour%57-60, and P s y c ~ i a t51, ~, Rosa, F.W. (1991) Spina bifida in infants of women nal of N e u r o l o ~Neurosurgery treated with carbamazepine during pregnancy. 120-123. Stancer,H.C. & Forbath, N.(1989) HyperNew EnglandJournal of Medicine, 324, 674477. parathyroidism, hypothyroidism, and impaired Rottanberg, D., Robins, A.H., Ben-Arie, O., Teggan, renal function after IO to 20 years of lithium A. &Elk, R. (1982) Cannabis-associated psychotreatment. Archives of ~ n t e ~ nMedicine, al 149, sis with manic features.Lancet, ii, 1364-1366. Schou, M.(1979) Artistic productivity and lithium 1042-1045. prophylaxis inmanic-depressive illness.British Stewart, P.M., Atherden, S.M., Stewart, S.E., Whalley, L., Edwards, C.R.E & Padfield, P.L. Journal of Psychiatry, 135, 97-103. (1988) Lithiumcarbonate: a competitive Schou, M. (1984) Long-lasting neurological aldosterone antagonist? British Journ~lof PsychiActa Psychisequelae after lithium intoxication. atry, 153,205-207. atrica Scandinavica,yo, 594-4502. Stokes, P.E., Shamoian, C.A., Stoll,P.M. & Patton, Schou, M., Juel-Neilson, N., Stromgren, E. & M.J. (1971) Efficacy of lithium as acute treatment Voldby, H. (1954) The treatment of manic psyof manic-depressive illness. Lancet, i, 13q-1325. choses by administration of lithiurn salts. Journal of Neurolo~, ~eurosurgery and Psychiatry,xy, Stone, K.(1989) Mania in the elderly.British Journal of Psychiatry, 155,220-224. 250-260. G.L. & Tohen, Schou, M., Baastrup, P.C., Grof P., Weis,P. Angst, Suppes, T., Baldessarini, R.J., Faedda, J. (1970) Pharmacological and clinical problems M. (1991) Risk of recurrence following discontinuation of lithium treatment in bipolar disorof lithium prophylaxis. British Journal of Psychider. Archives of GeneralPsychiatry,48,1082-1088. atry, 116,615-619. Swann, A.C., Sctcunda, S.K., Katz, M.M. et al. (1986) Sclare, P. & Creed, F. (1987) Life events and the Lithium treatment of mania; clinical characteronset of mania. British Journa~of Psychiatry,156, istics, specificityof symptom change, and out508514. come. Psychiatry ~esearch,18,127-141. Sellers, J., Tyrer, P., Whiteley, A., Banks, D.C. & Barer, D.H.(1982) Neurotoxic effects of lithium Thisted, E. & Ebbesen, F. (1993) Malformations, withdrawal manifestations and hypoglycaemia with delayed rise in serum lithium levels. Britafter exposure to valproate in utero. Archives of ish Journal of Psychiatry, 140, 623-4525. Shapira, B.,Yagmur, M. J., Gropp, C., Newman, M.Disease in Childhood, 69,288-291. Thornhill, D.P. & Field, S.P. (1982) Distribution of & Lerer, B. (1992) Effect of clomipramine and lithium elimination in a selected population of lithium on fenfluramine-induced hormone repsychiatric patients.EuropeanJournal of Clinical lease in major depression.Bioiogical Psychiatry, P ~ a ~ a c o l o g21,351-354. y, 3% 975-@3. J., Baldessarini, R.J.,et al.(1995) Shapiro, D.R., Quitkin, EM. & Fleiss, J.L. (1989). Tohen, M., Castillo, Blood dyscrasias with carbamazepine and Response to maintenance therapy in bipolar illvalproate: a pharmacoepidemiological study of ness: effectof index episode.Archives of General 2,228 patients at risk. A ~ e r i c a Journal n of PsychiPsyc~iatry,46,401-405. atry, 152,413-418. Sharp, T., Bramwell, S.R., Lambert, P. & GrahameSmith, D.G.(1991) Effect of short- and long-termTransbol, I., Christiansen, C. & Baastrup, P.C. (1978) Endocrineeffectsoflithium: I. administration of lithium on the releaseof enHypothyroidism, its prevalence in long-term dogenous 5-HT in the hippocampus of the rat treated patients.Acta Endocrinolo~,879,759-767. in vivo and in vitro. N e u r o p h a r ~ a c o l o3, ~ ,977Tyrer, P.,Alexander, M.S., Regan, A. & Lee, I. (1980) 984. An extrapyramidal syndrome after lithium Shukla, S., Godwin, C.D., Long, L.E.B. & Miller, therapy.British Journalof Psychiatry,136,191--194. M.G. (1984)Lithiutn-carbamazepine neurotoxicity and risk factors.A ~ e r i c a nJournal of Psychiat~, Vestergaard, P., Amdisen, A. & Schou, M. (1980) Clinically significant side effects of lithium treat141,1604-1606. ment: a survey 237 of patients in long-term treatSilverstone,T. & Romans,S. (1996) Long term treat-
ment. Acta Psychiatrica Scandinavica, 62, 193-
Wehr, T.A., Sack, D.A. & Rosenthal, N.E. (1987) Sleep reduction as a final common pathway in the genesis of mania. AmericanJourna~of PsychiVon Knorring, L.(1990) Possiblemechanisms for atry, 144,201-204. the presumed interaction between lithium and Wehr, R.A., Sack, D.A., Rosenthal, N.E. & Cowdry 5,287neuroleptics.~u~~ Psychopharmucolo~, R.W. (1988) Rapid cycling affective disorder: 292. contributing factors and treatment responses in Walden, J., Von Wegerer, J., Beger, M. & Grunze, H. (1996) Efficacyof antiepileptic drugs in the 51 patients. A ~ e r i c a nJournal of Psychiatry, 145, treatment of psychiatric diseases. EEG Labor, 18, 179-184. West, A.P. & Meltzer, H.Y. (1979) Paradoxical 32-47. Waller, D.G. & Edwards, J.G. (1989) L i ~ i u mand lithium neurotoxicity:a report of five cases and a hypothesis about risk for neurotoxicity.Amerie, the kidney: an update. Psycho~ogical~ e d i c i ~19, can J o u ~ aofl Psyckiatryf 136,963-966. 825-831. Waller, D.G.,Albano, J.D.M., Millar, J.G.B. & Polak, White, E., Cheung, P. & Silverstone, T. (1993) Depot antipsychotics in bipolar affective disorder. AMP responsesto parathyroidhormone A. (1987) rnternationa~Clinical Psychop~armacologist,8, and glucagon during lithium treatment. Clinical Science, 66,557-559. 119-122. & Brook, S. (1996)Superior- WOE,J., B e r n . S.C. & Jones, A.B. (1970) Inhibition Walton, S.A. Berk, M. ity of lithium over verapamil in mania: a conof thyrotropin-stimulated adenyl cyclase activtrolled trial. European Psychiatry, 11 (suppl. 4), ity of beef thyroid membranes by low concenc a 2~ i o p h y s ~ tration of lithium ion. ~ ~ o c h e m ~and 294s. cal Research ~ommunica~ionsf 39,7742. Watkins, S.E., Callendar, K.,Thomas, DR et al. (1987) The effectof carbamazepine and lithium Wood, A.J. & Goodwin, G.M. (1987) A reviewof sh the biochemical and neuropharmacological on remission from affective illness. ~ r ~ t iJournal of P s ~ c h i u t150,180-182. ~, actions of lithium. Psyckfflogica2Medicine, 17, Wehr, T.A. & Goodwin, F.K. (1979) Rapid cycling 579-600. in manic-depressives induced by tricyclic anti- Yassa, R., Saunders, A., Nastase, C. & Camille, Y. (1988) Lithium induced thyroid disorders: a f depressants. Archives of General P s y c k i a t ~36, l ~in~ca~ Ps~ch~atry, prevalence study. J o ~ r ~ofa C 555-559. 49, 14-16. Wehr, T.A.&Goodwin, EK, (1987)Can antidepressants cause mania and worsen the course of Zarate, C.A., Tohen, M.,Banov, M.D. et al. (1995) affective illness? American Journal of P s y c h i a ~ ~ , Is clozapine a mood stabilizer? Journal of Clinical Psychiatry,56,108-112. 144,1403-1411. 200.
SCOTTWEICHANDANTONY
MANN
Thirty years ago the first substantial study of psyc.batric morbidity in primary carewas published by Shepherdet al. (1966).A one-in-eight sampleof patients attending 46 general practices in London were assessed. Of the 15ooo patients at risk, one in seven were consulting for a psychiatric condition (largely, depression and anxiety). The authors’ recommendation was not to call for more psychiatrists,but to strengthen the primary care team to deal with these conditions, which most of the collaborating general practitioners (GPs)regarded as their own problem. It is surprising, therefore, that UK in the and the USA, only in the last few years, has there a general been acceptance by health service planners, psychiatrists and primary care personnel that neurotic disorders do constitute a significantburden to the individual sufferer, the general practiceand the public purse. With this acceptance comes exciting possibilities for improvement for all three should effective treatments be introduced in primary care. The intervening 30 years have seen the publication of the resultsof a programme of research that has defined more preciselythe type and extent of psychiatric morbidity seen by GPs and, more recently;begun the evaluation of strategies for better detectionand treatment. This work has been greatly facilitated by the creation of standard measures: the General Health Questionnaire (Goldberg & Williams, 1988) and the Clinical Interview Schedule (Goldberg et al., 1970; Lewis et al., 1992) which were specifically created for case-finding among primary care patients. Eighty to ninety per cent of patients with mental disorders that present to GPs suffer from depression, anxiety or a mixture of the two states. Only a minority of patients present their distressin psychological terms,the majority through a processof somatization (Goldberg& Bridges, 1988). Patients are likelyto present at times of social stress and also when physically ill. The GP’s picture, therefore, is a complicated one and difficult to unravel in the brief consultations in the surgery. Goldberg and Huxley (1992)have constructed a hierarchical model to show the relationship of this primary care morbidity to that seen in secondary care. Using existing prevalence data, they indicated that only about 50% of the
apter er 14: ~ a n a g e ~ine general ~t practice
morbidity is, in fact, detected in primary care of and, the detected, only about 5-1oY0 is referred for specialist help.A large volume of psychiatric activity, therefore, happens outside the scope of the psychiatric services, approximately three patients consulting seen in primary careat a point in time, with a diagnosable psychiatric disorder, for any one being seen all in components of the psychiatric services combinedat that time. Over the30 years, a well-rehearsed response to this demonstrationof the scale of primary care psychiatry has been that the disorders seen there are relatively mild and self-limiting. GPs were stated to be responsible for patients with socioeconomic problems and the 'worried well', in contrast to the work of the communi^ psychiatrist (Bennett, 1973). These prejudices have been overcome with the demonstrationof the severity of many cases of depression among the spectrum of disorders seenin primary care. Some 5-10% &:Clare, 1987) andthe recent suffer from a major depressive disorder (Blacker World Health Organization (WHO) multicentre study indicated that 10.4% can be classed as having a current depression using ICD-lo World Health et al., Qrgan~ation(WHO) criteria, with a further2% with dysthymia (Ustun 1995). The outcome for the patients is not as good as might be imagined. Only a minority recover quickly; the majority of sufferers will tend have to a fluctuating course and approximately one-fifth become chronic, u~emittantcases and high users of primary care (Mann et al., 1981). Finally, the cost to the economy of this primary care morbidity through disability days and the cost to the health service through frequent consultation and prescription have bee shown tobe considerable (Croft-Jefferys&:Wilkinson, 1989). Recognition nowof the potential for benefit for thepublic health and the individual patient has led to recent activity. h the TJK and other European countries, a campaignto defeat depressionhas been launched by the Royal &:Priest, 1992).A Colleges of Psychiatrists and General Practitioners (Paykel series of research initiatives has been funded with the aim to improve detection and treatmentin primary care under the general rubric S t r e ~ g t h e ~the i~g Primary Care Team. Also, the changes in the structure of the National Health local a level to Service in the UK has placed more pressure on psychiatrists at collaborate more closely with their GP colleagues, thereby restoringpsychiatric attention to the common disorders in general practice, often through setting up liaison consultations in the surgery. This chapter will first discuss the evidence that increasing detection of depression in primary careis of value and second, the status of management in primary care and the research studies that are underway to improve the situation.
tes of detection of depression in primary care vary greatly in diffe the world (Ustunet al., 1995). ~ l t h o u g hdetection rates of 75% have been reported in one or two centres (vonKorff e f al., 1987;Ustun e f a 1995)~ most studiesof routine primary care attenders in ~ e s t Europe e ~ an North ~ m e r i c have a reportedthat primary care around one-half of prevalent casesof depression et al., 1985; Blacker& Clare, 1987; Ormel et al., 1991; Goldberg Priest, 1992; Ustune f al., 1995). Since evexi. the 'best miss a large proportion of the cases thatpass through. their surge
Evidence c o n c e ~ i n gthe conse~uencesof GP reco comes from two sources: naturalistic outcome studies rmel and his colleagues (1990) fo duction in PS chiatric sympto 179 'new' cases of psychiatric disorychiatric diagnosis in the12 months tion), irrespective of diagnosis. These authors also ition and abriefer mean durationof o or me^ et al., 1991). of anxiety but not depression on outcome did not appear to be confounded by iations between recognition and of the index episode of psych the beneficial effects of reco the least severe disorders A weaker positive association between recog~tionand (better) outcome was et al., 199 also observed3 years after the index consultation (Ormel ition was associated with a greater li~elihoodof receiving a mental intervention (most c o ~ o n l ysychotropic medication), this did not nfound) the association between. recognition an an do mi zed con~olled trials in whic themental state of patientswithpreviouslyunreedpsych.iatricdisorhave der produced contradictory results. studies found th informing GPs about presence the of 'U chiatric m o r b i ~ i ~ (as detected by a screening~ u e s t i o ~ a i rwas e ) associated with higher rates of
recovery at 1 (Zung et at,, 1983) and 3 months (Johnstone6r: Goldberg, 1976), and briefer overall episode duration (Johnstone & Goldberg, 1976), subseet al., 1984; Shapiro quent studies have failed to replicate these results (Hoeper et al., 1987). In an excellent review of this literature, Gold (1992) concluded that feeding back the results of a psy alone wasof little use in the absenceof advice aboutpatient management, a & Priest, 1992;Wright, 1994; Lewis, 1992). conclusion echoed elsewhere (Paykel Two recent r ~ d o m i z e d trials in this area support this interpretation. Dowrick and Buchan (1995) found no association between disclosure of Beck Depression Inventory (BDI) score and outcome among casesof de ression not recognized by GPs. In a naturalistic arm of this study, unreco depression had lower BD1 scores at the outset than patients sion was ’spontaneously’ recognized by their GP, and this differencerease^ se^ during the la-month study period. h a randomized study of consecu~ve attenders, Lewiset al. (1996) found that while providingGP@with details of subjects’ General Health Questionnaire (GHQ) score had little effect on outcome, feedbackof a more detailed computerized psychiatric assessment based on the Revised Clinical Interview Schedule was associated with a modest improvement in outcomeat 6 weeks and 3 months, though not at 6 months. GPs tendto detect the more severe cases of depression while overlooking those cases most likely to remit spontaneous ich limits the capacity of randomized trials of disclosure of G scores to detect an improvement in outcome for previously undetected cases compared with no treatment. Taken together, these findings confirm the view of Goldberg and Huxley that detectionof depression per se has at best a modest effecton outcome, and none at all in the absenceof advice or training on the mana of depression (Goldberg & Huxley, 1992). Indeed, it may be that the ’often haphazard and inadequate’ treatmentof patients with conspicuous psychiatric morbidity in primary care first described by Shepherd et al. in 1966, and confirmed elsewhere by others (Ormel et al., 1991; Katon et al., xyp), is of greater significancein explaining the persistently high prevalence of depression inboth the community and primary care, despiteavailabili~ the of treatments of proven effectiveness (Brugha, 1995). The detection of depression by GPs is a necessarybut not sufficient precursor to adequate treatment, and the imperative must surely be to provide GPs and patients ~with o ~ a t i about o n the nature and treatment of depressionthough a combinationof professional and public education (Katon et al., 1992).
Any account of management in primary care has to take into account the specifics of the situation-brief consultations inbusy surgeries, patients with
depression of varying levels of severity which is usually part of a complex clinical picture with a concurrent physical illnessor marked social stresses. Patients in primary care do not necessarily see themselves as ’psychiatric’, and indeed may be afraid of receiving treatment for a mental illness, whether it be medication or a psychological treatment. Negotiation of treatment and monitoring of compliance and prognosis is therefore necessary but, in dayto-day practice, often hard to achieve. ~ ~ be e ~prei c a t i o ~ The GP needs to decide whethera ~ t i ~ ~ ~ e s s a ~ should psycho-a scribed and, if so, how to ensure compliance. In addition, or instead, logicul ~ ~ e or u atsocial ~ ie ~~ ~~ e ~to ~ameliorate e ~ ~ itheo patient’s ~ circumstances may be appropriate. Here, the GP has to consider how this is best achieved in the surgery or the locality, given the enormous variation of availability of skilled personnel to deliver this approach. In the background will be the need to consider whether any current physical illness its or treatment is promoting or maintai~ngthe depressionor whether the treatment for depression could be medically risky. Finally, he/she will have to decide whether a speciulist ~ e ~ eis~now ~ a appropriate l and, in particular, to be alert to the potential for suicide. Specific management approaches under these areas be discussed will more hlly below. The GP, therefore, is the linchpinto the approach to managementof depression in primary care. Several initiativesinare progress to make this management as sophisticated as possible and to improve the confidence of the GP. Increased input concerning the diagnosis and treatmentof mental illnesses into the postgraduate programmes is being promoted by a network of regional GP fellows, trained and supported by a seniorGP funded for this purpose by the Department of Health and the Mental Health Foundation. Guidelines for treatment have been prepared as part of the ‘Defeat Depression’ campaign,both for older and younger adults (Paykel& Priest, 1992; Effective Health Care, 1993; Katona et al., 1995).Computerized treatment algorithms have been devised that will allow theGP on his/her own surgery computer to access specialist help quickly (Lewis et al., 1988). Facilitation (a process whereby a professional in certain areas is attached to a practice to introduce changesin that practice over a short period) is another approach. Facilitation hasbeen successful in introducing better practice for 1992).It is now being evaluated for assessing cardiovascular risk (Armstrong, its impact upon the detection and management of anxiety and depression by means of a randomized controlled trial conducted in 18 London practices, Early results suggestthat detection at least is improved as the result of facilitation (Bashir, personal communication). There is, however, a risk. GPs, already under pressure after contractual changes, may resent bombardment by well-intentioned guidelinesor visitors from outside the practice.It is, perhaps, not adequately recognized that GPs
~ ~ a14:~~ a~n ae g re ~ine general n~
actic ice
themselves are not a homogeneous group in their attitude to treatment of depression. The Depression Attitude u e s t i o ~ a i rwas e recently applied to a random sample of GPs who were part of the Medical Research Council’s Research Framework (Botega et al., 1992). Their responses, when subject to cluster analysis, suggested three groupings: those who had interest and sympathy for their depressed patients and wanted to personally do more, along the psychotherapeutic lines; those with a more strictly medical/organic view of depression who wanted clear-cut lines of physical treatment; those who found depressed patients very heavy-going and wished the treatment to be conducted by others. Any strategies from outside to improve management in general practice should, therefore, take account of this variation. An approach suitable for one GP is likely to irritate another GP. It is also clear GP, that guidelines and protocols must be developed in partnership with the allowing individual variations in attitude andlocal resources. The burdenof detection and care can alsobe shared by practice nurses and health visitors who work regularlyin the practiceand/or by specialists, suchas counsellors or psychologists who may be employedby the practice for this purpose. This ~ ~ Z t i ~ i s c i ~ l i n a in ~ primary tea~~ care u r will ~ be discussed later.
This section will not rehearse the evidence that antidepressant medication helps patients with depression, nor discuss which antidepressant should be preferred. These are discussed elsewhere in this book (Chapters 9 and lo). This section will focus on the evidence that depressed patients seenin primary care can benefit from medication, but also discuss obstacles to effective prescribing. The vast majorityof studies of efficacy of antidepressants have been conducted on out-patient samples with mild to moderately severe depression. A n tHealth and Human Servmeta-analysis publishedby the U S ~ e ~ a r t m eof ices (1993) in Cli~ical Practice G ~ i ~ e l i n e s : ~ e ~ inrPe s sr i o i~Care ~ pointed ~ ~ to the efficacyof antidepressant medication over placebo aand ~ i o l y t ifor c cases of major depression. They also reported on a subanalysis of seven studies that took place in primary care, where they showed a drug efficacy of 57.8% compared with placeboof 35.6%-the latter being higher than the studies in UK,some specific studies have shown the superiorin-patient settings. In the ity of amitriptyline (mediandose, 125 mg) over placebo, when taken for a6week period by patients who meet research criteria for depression with a (HWD) score of 6 (Hollyman minimal Hamilton Rating Scale for Depression et al., 1988). Subsequent research has focused on comparative studies between the various forms of antidepressants for use in primary care. The metaanalyses of research in these settings would su~gestequal benefit between
apter er 14: M ~ n a g e ~ ein n t~ e ~ e rpractice al tricyclic antidepressants (TCAs) and selective serotonin re-uptake inhibitors (SSRIs), but with a tendency for lower dropout from theSSRI group (Song et af., 1993).Prescription needs to be continuedat least for 4-6 months (Effective ealth Care, 1993; Old Age Depression Interest Group, 1993). While it is, therefore, established that antidepressants are useful for depression that meets a particular severity criterion theon HRSD, it is not at all clear how effective they are forthe group of patients with mild to moderate depression who are commonly seen in primary care (Paykelet al., 1988). There seems to be a tendency for a small dose to be given to a 'little' To depression. answer this question, the Medical Research Council is currentlyfunding a study taking place in 20 practices, in which antidepressant medication is being compared, for benefit among patients with mild depression, with problem-solving-a psychological approach conducted by practice nurses ( ~ ~ o r s - W a l leti sal., 1995). The evidence remains that antidepressant medication is beneficial in primary care for thosewith a depression of moderate severity. However,in actual practice, dosages are often prescribed at subtherapeutic levels and compliance is poor: less than 60% of patients actually ma~tainingtreatment for morethan 3 weeks (Johnson, 1981). A recent survey of 80 ooo prescriptions conducted by Donaghueand Tylee (1996)shows the mean dose for tricyclic 56mg per day, but there is huge a range between GPs prescription is currently on the length of time for which these are prescribed. In contrast, the dosage for the newer antidepressants and the SSRIs are usually at the therapeutic level. It seemsthat the concernof GPs and complaints bythe patients about side-effects toTCAs are a major factor in explaining this deficiency. Both may be overcome if the predominance of tricyclic medication yields to SSRIs in prescription frequencyin general practice. The latter have fewer side-effects and a narrower dose range. Thus, this change will probably result in more patients receiving antidepressant medication in an effective range of treatment, despite lack of evidence forsuperiority of the SSRIs over TCAsin conditions of full dosageand maximal compliance. However, the medication itself not is the complete story for difficulties in effective useof pharmacotherapy in general practice. A study reported from South Wales compared a sample of GPs and psychiatrists for their attitudes to depression and their treatment strategies (Kerret al., 1995). Therewere predictable differencesin attitude to depression between psychiatrists and GPs, but it was clear that there was a worrying difference between GPsand psychiatrists in their use of antidepressant medication. Approximately50% of the GPs were reported as prescribing less than the recornended dosage of antidepressant medication and 40% for lessthan 4 months-the recommended continuation time."his group of GPs were comparedwith their colleagues, who claimed to treat in full doses for longer time, for their beliefs. The low
apter er 1.4:~ a n a g e ~ e in n tgene~alpractice prescribers tended tohave a strongbelief in the valueof psychotherapy and much less conviction about the biochemical basis of depression. Patients, too, are reluctantto take antidepressant medication, particularly the older patients, associating these drugs with a potential for addiction. Explanation of the prescription and whyit is necessary, with follow-up and support, would seem mandatory after prescription. Thisis hard to deliver in abusy general practice where many patients can get 'lost'. Here is a possible role for the practice nurse who can work with GPs as of the partmultidisciplinary effort ( ~ i ~ ~ o et al., 1993). The reluctance to take antidepressant medication by older patients was clearly demonstrated in a recentinstudy, which ac o ~ u n i tnurse y acted as a case manager for depressionin primary care. Only a thirdof the prescriptions for antidepressant medication were accepted by the study population drawn from a community sampleof people over 65 (Blanchard et al., 1995)Therefore, although antidepressant medication is clearly effective for moderate to severe depression, the attitude of some GPs and patients makes them unwilling or uncertain participants in this form of treatment. For this reason, psychologital treatments are often more popular and will be now considered.
Like social interventions (see below), evaluation of psychological interventions for depressionin primary care has been hamperedby a lack of clarity et al., 1994). and specificity in the treatments which have been studied (King Though we have distinguished between 'social' and 'psychological' interventions for the purposes of this chapter on the grounds that the former usually includes practical assistance in modifying the social environment (e.g. help in securing benefits or re-housing), this distinction is somewhat artificial, since social work also comprises 'psychological' activities such as providing 'support by encouragement and listening, [and] help to understand feelings' (Scott & Freeman, 1992). Despite the recent vogue for 'counselling' (Corney, 1992; Sibbald et al., ~ 9 9 3there ) ~ is little evidence about the effectivenessof psychological interventions in primary care, a stateof affairs which reflects the combined difficulties associated with evaluating psychological treatments in general and conducting large trials in general practice (Brewin & Bradley, 1989; Greenberg, 1991; Tognoni et al., 1991; King et al., 1994). Among published trials, several psychological interventions have been found to be superior to placebo (no treatment) in the management of major depression in primary care, though there is no evidence thatany are more orless effective than treatment with a TCA at a dose equivalent to amitriptyline 15omg daily at least within the first 3 or 4 months of commencing treatment.
Chapter 14: ~ a n a g e ~ ein n general t p~actice Two studies have consideredthe outcome of psychological inte~entions for general practice patients suffering from a range of non-psychotic psychiatric disorders. Brodatyand Andrews (1983)found no difference in outcome between patients with non-psychotic psychiatric morbidity of at least 6 months’ duration (defined as a persistently elevated GHQ score) who received eight weekly half-hour sessions of ’problem-oriented’ psychodynamic psychotherapy compared with those who saw theirGP for a similar amount of time, and those who received ’standardGP care’, even after adjusting for baseline differences betweengroups. In addition to its non-randomized design, interpretation of the findings from this study were hindered further by a high dropout rate and a failure to analyse the results an intention-to-treat on basis. Gournay and Brooking (1994)compared the outcome forpatients with ’nonpsychotic psychiatric problems’ randomized to immediate intervention from a community psychiatricnurse (CPN), a waiting list controlgroup or continuing standard GP care. All three groups were found to have improved significantly on awide variety of psychiatric and social measuresat 6 months, though no differences in outcome werefound between groups. Though the work of the x 1 CPNs involved in the study was not prescribed (or described), the majority ’saw ”counselling’’ as central to their role’. In a randomized controlled trial, Teasdale et al. (1984)found that general practice patientswith RRC major depression (Spitzer et al.,1978)who received an average of 15h of cognitive therapy from a clinical psychologist in addition to standard GP care recovered morerapidly than patients who received standard GP care alone, though the difference in outcome between these groups disappeared within 3 months of the cessationof cognitive therapy. This study was limited by small numbers, a relatively high dropout rate in the cognitive therapy group and a decision to analyse results only for subjects who accepted the inte~ention.As described above, Scottet al. (1992)found no differencein outcome at 16 weeks forpatients who received an average of xo sessions of cognitive-behaviour therapy from a clinical psychologist compared with those who received eitherstandard GP care or amitriptyline from a psychiatrist. The dropout rate was higher, and overall satisfaction lower, in the cognitive therapy than in the group allocated to social work interventio~.More recently, Mynors-Wallis et al.(1995)reported the results of a randomized controlled trial of six fortnightly sessions of brief ’problemsolving treatment’, delivered by a psychiatristGP, orcompared a with amitriptyline (150mg daily) and a drug placebo for primary care patients with major depression. Though superior to placebo, no difference in outcome was found between problem-solving and amitriptyline at 6 or 12 weeks. Patients in the problem-solving group reported high levels of satisfaction, and were less likely todrop out of treatment than subjects in either of the other groups.
~lthough both socioeconomic adversity, suchas poverty, poor housing and ployment (Platt et al., 1990; Goldberg zer et al., 1995) and difficulty in inte rown et al., 1986) are c o ~ o an S and conse~uences of this dis own et al., 1986; Warr, 1987; al., ~ 9 9 7there ) ~ have been depression in primary careor elsewhere. Whileit seems likely that social interventions maybe part of standard GP care (the control conditionin many trials in primary care), there havebeen few attempts to evaluate individual components of such care. Indeed,in all primary care trialsof social interventions (of which we are aware) the experimental condition has been allocation to treatment by an individual from a specific discipline (e.g. social worker, health visitor or practice nurse), rather to any one t e of social intervention (Corney,1984; J o ~ s t o n & e Shepley,1986;Corneyrray,1989;Goldberg Huxley, 1992; Scott& Freeman, 1992;Wil~nson,1992). The most detailed study of this nature was a randomized controlled trial by Corney (1984) of allocation to a social worker for women withacute depression. Though no overall difference was found in clinical or social outcome for subjects (women aged 18-45) allocatedto a social worker attached to a GPcomp~redwith GP management alone, subgroup analyses revealed a benefit from social work t r e a ~ e n for t women with ’acute-~n-chronic’depression with major marital difficulties. Though analyses were based on sm numbers, this group were described as having more social problems at outset, and were both more likely to accept the social work t r e a ~ e n and t to receive practical help than other jects. In a review of this and other social inte~entionstudies, Corney and rray argued that intervention by a social worker islikely to be of greatest benefit to those with the most long-standing and severe depressive illnesses,on the grounds that ’earlyinte~entionmay e harmful, the social worker ~terfering with the client’s own abilities or affecting the support received from informal sources’ (Corney & 1989). This view was challenged in a more recent trial where social work intervention was compared with standardGP care, cog~tive-behaviourtherapy from a clinical psychologist and antidepressant medication prescribedby a psychiatrist among patients in primary care recruited when they were ’about to start treatment for a depressive illness’ (Scott ipr Freeman, 1992). criticized for failingto adequately describethe social work ~tervention,the artificiality of the psychiatrist arm of the study (restricted to prescription vice on matters related to pharmacotherapy)~and a randomization
differences in depression of overall recover^ but outcome was found in this group included the highest proportion of ean severity of depression at baseline. ese potential sources of confounding. The ects who received the social work interfied with their t r e a ~ e nthan t otherstu
in collaboration in the pracsurvey of general practices % had direct links social worker; 17% with a counsellor;15% with a a psychiatrist (Sibbaldet al., 1993). These attachvenly spread. Some practices had three or four
r patients with depres-
for GPNs from psychotic patients to those at one is normally linked to five practices; can only come into contact with a nxinority of patients study indicated that CPNs working with ents produced no extra benefit to an ordinary outpatient referral a llors are usually referred patients with ip di~icultiesof which the core sy~~tom unsellors’ approaches to treatment are ining received. It has been reported that tients (~nderson& Hasler, 1979) and ation ayde den field conducted randomized counsellors havebeen an important flecting a public demand for psychofor statesof depression and anxiety.
Most general practices employ nurses, either to work in the surgery or in the c o ~ u n i t yThese . nurses will come into contact with many patients at risk for depression: the physi~allyill, the elderly, new mothers, etc. Potentially therefore, these nurses could contribute to the management of depressed patients being treated by theGP. A role is already established fornurses to run clinics in general practice for sufferers from chronic diseases such as asthma, hypertension and diabetes. Nurses could run a support clinic for patients with depression, allowing time for education of patient and family, monitoring compliance toantidepressants and the detectionof social factors that might be maintaining depression.A group of practice nurses has been recently trained to assess depression and to usea structured manual to follow up such patients, working as an adjunct to the GP. After a successful pilot ilkin ins on ef al., 1993)~this role was evaluated by means of a randomized controlled trial (GP treatment alone vs. GP and nurse working together) in 20 practices (Mann e f al. submitted). Five hundred and seventy patients with conspicuous depression, the majority sufferinga major from depression, were recruited. "he outcome of the study confirmed thefeasibili~and popularity of the nurse clinics, but no added benefit in outcome could be detected. It seemed that the presenceof these practice nurses, who were trained to assess and then discuss the depressed patient with the GP aspart of the intervention arm, had an important effect upon the GP who changed practice for the patients in the controlarm that he/she was seeing alone.As a result, allparticipantsin the trial benefited froma high prescription rate of a n t i d e p r e s ~ ~ t s (74%)with excellent compliance(79%).Perhaps havinga practice nurse interested in andavailable to helpin the careof depressed patients increases the assurance of GPs in their owntreatment. ultidisciplinary teamwork in primary care is still in its infancy, but it ly that patients with depression will benefit from the extra time for assessment and monitoring that nurses could carryout. In theory, the specific skills of attached professionals should improve the addition of specific inte~entionsand the outcome for depressedpatients. However, evaluative studies of this work is awaited. A demonstration of a lack of specificity of approach by attached workers is worrying. During a survey of 4.0 practices in South London, R. Corney (personal communication) asked the professionals attached to these practiceswhat types of therapy they offered. "he responses from the GPNs, counsellors and psychologists indicated that all three groups interchangeably offered psychodynamic therapies, stress management, cognitive behaviour therapy and anxiety management. Only one tivity seemed to be job-specific: that of giving depot injections by the nurses, uch more research is required if the continuation of this range of attachments is justified.At the moment, although the GP may believe that something different is being delivered by a counsellor or a psychologist to a
~ ~ a ~14:t ~e a rn a g e ~ ein n general t practic~ depressed patient, it may be that they are receiving similar, unspecified intervention from both groups.
As was highlightedearlier in this chapter, the nature and scale of the burden of psychiatric morbidityin primary carehas been documented with increasing precision in the 30 years following the seminal study by Shepherd et al. (1966). It has been estimated recently that each of the 30000 GPs in Britain sees 300-600 anxious and depressed patients each year, a number far beyond the capacityof the 2000 consultant psychiatrists and5000CPNs, even if they & Jenkins, 1995). In view of the spent all of their time in primary care (Lloyd large numbersof specialist mental health care professionals working in priet al., 1993)~substantial concerns have also been mary care settings (Kendrick expressed about the care of patients with chronic psychotic illnesses, it and is likely that secondary care personnelwill come under increasing pressure to devote more time to this group (Gournay & Brooking, 1994; Lloyd tk Jenkins, 1995). While there is some evidence that shifting psychiatric out-patient clinics into general practice reduces demand on secondary care services (Tyrer et al., 1984), evaluationof a service in which the secondary care team was based in primary care suggested that GPs referred patients they might et al., 1993). There is therefore otherwise have managed themselves (Jackson a clear need to strengthen the primary care team through the introduction of specialist knowledge on the detection, diagnosis and management of the c o m o n mental disorders, anxiety and depression, in ways which do not involve diverting scarce secondary care personnel away from patients with severe mental illness. two experimental interventions designed to We have already highlighted strengthen the primary care team, namely the introductionof ent tal ~ e a l t ~ ~ a c ~ l ~and ~ atraining ~ ~ r sfor ~ r a c t ~nurses ce in the assessment and support of depressed patients. Other ways in which it has been suggested that this might be achieved are through specific training packages for GPs (Gask, 1992), the development and dissemination of clinical practice guidelines (including guidelines on the indications for referral to secondary care) (Effective Health Care, ~ 9 9 3and ) ~ through the use of information technology (Lewis, 1992). There has been a proliferation of clinical guidelinesin many western countries across a rangeof specialities (Haines & Feder, 1992), including psychiatry (Effective Health Care, 1993; Paykel & Priest, 1992; US Department of is consistent evidencethat consensus Health & Human Services, 1993). There statements based on expert opinion and disseminated through publication in journals have little impact on clinician behaviour (Grol, 1990). Instead, guidelines are mostlikely to be acted uponif those who are to use them participate
in their development(North of England Study of Standards in General Practice, 1992), and if doctors are given patient-specific reminders out treatment at the time of consultation (Haines& Feder, 1992;Grimshaw Russell, 1993). Gomputer-based clinical decision support systems able of combining patient information with treatment gui~elinesto patient-specific prompts (Johnston et al., 1994). The application of this technology in psychiatry has been limited by the difficulty of obtaini ardized clinical data on which to base these guidelines, giventhe extreme variability in doctors' clinical assessments. It has now been demonstrated that self-ad~stered co~puterized assessments of c o ~ o mental n disorder are reliable, unbiased and valid, and are acceptable and easy to use for patients within primary care (Lewiset al., 1988; Lewis, 1992; Lewis, 1994). Though GPs vary greatly in their rates of referral to secondary care services (Greedet al., 1990; Fertig et al., 1993)~ comparisons of such rates are notoriously difficult to interpret ( ~ a r i n k eetr al., 1988; knowledge about the characteristicsof both patients associated with psychiatric referral (see Goldberg is limited empirical evidence on which to base gui psychiatric referral, or that such guidelines would reduce referral rates substantially (Russell & Grimshaw, 1992; Fertig et al., 1993). While risk to self or others is likely to remain an absolute indication for referral, other indications are a matter for local negotiation between stakeholders at the primary-secondary interface, including GPs, psychiatrists and all those responsible for plamin and purchasing mental health services.
' ~ n d e r t r e a ~ e n tor ' , perhaps more correctly 'unmet need for treatment', is probably the single mostimportant issue inthe management of depression in general practice (Brugha, 1995). We would argue that it is necessaryto move beyond the study of case recognition to consider both physician (e.g. unwillingness to prescribeantidepressants at high enough doses for longenough) and patient (non-compliancewith treatment) variables associatedwith treatment adherence and clinical outcome. One area whichhas not yet been exlored fully is the effect of patient attitudes on the outcome of depression in eneral practice (Katon et al., 1992); depressed patients frequently electnot to consult theirGP at all, while those who do rarely present psychological complaints. This may be the result of misconceptions about (among otherthings) the nature of depression, the ability (or willingness) of their GP to help,and the likelihood of being offered physicalrather than psy~ologicaltreatment (or vice versa). We simply do not know enough about patient preferences, or how these influence consultation behaviour or compliance with ~ e a ~ e n t .
A second unresolved issue concerns the effectiveness of alternatives to pharmacological interventions, either alone or in combination with interventions. Clear conclusions have been hampered by, first, th specificity in the treatments which have been evaluated, second, general difficulties associated with evaluating psychological treatments (including subject motivation; radley, 1989) and the difficulty of d i s ~ g u i s h ing between effects to the intervention and those which are due ist), and finally the problems associated with ~ d e r t a ~ largeng n general practice (especially non-participation by GPs who to recruit patients; Tognoni et at., 1991). Though there have been specific therapies (e.g. cognitive therapy and problem-solving therapy), most studies have evaluated the outcome for patients refer professionals from a articular discipline (e.g. social worker or CPN). I we are not aware of any evaluationsof specific social i n t ~ ~ e n t i oin n s primary care (e.g. debt counselling or befriending). espite the very significant and exciting developments in the primary care chiatric disorder which have taken place in the 30 years since the study by Shepherd and collea ues (Shepherd et al., 1966), there islittle evidence of any su~stantialreducti in the prevalence of depression, either in primary care settings or in the community (B gha, 1995). The heaviest burden rests xley; 1992; Shah, 1992)~ perhaps even on the shouldersof GIs' (Goldberg & As a d ct result of changes in the deliveryof specialmore now than in 1966. ist psychia~istservices, sec0 ry care staff are now occupied almost exclusively with the care of patients with the most severe psychiatric disorders, which commonly means the psychotic disorders. It is ironic thatcommunit~ psychiatrists should have so little time to devote to those illnesses which are most prevalentin, and costlyto, the community. It isof the utmost importance that psychiatrists should not abandon their vital role in the primary care of depression. Psychiatrists are the one best placed to ensure that only treatments of proven effectiveness (i.e. those which are 'evidence-based') are implemented in primary care settings. As medical colleagues, GPs are naturally inclined to look to psychiatrists for advice in this area. Should this leadershipbe not forthcoming, the vacuum is likely tobe filled by more vocal advocatesof interventions, suchas CO ling, which have yet to be shown to work.
Anderson, S. & Hasler, J. (1979) Counselling in Bennett,D.H.(1973) Community mental health services in Britain. American ~ o ~ r nofa lPsychigeneral practice. ~ournalof theRoyalCollege of General Prac~itioners,29,352-356. atry, 130,1065-1070. Armstrong, E.(1992) Facilitators in primary care. Blacker, C.V.R. & Clare, A.W. (1987) Depressive ~n~erna~ional Review of Ps~chiatry,4,33y342. disorder in primary care. British ~ournalof Psy-
Donaghue, J.M. & Tylee, A. (1996)The treatment c k i a t ~150, , 737-751. of depression: prescribing patterns of antideBlanchard, M.R., Waterreus, A. & Mann, A.H. (1995)The effectof primary care nurse intervenpressants in primary care in the UK. Britisk Journal of Psyckiat~,168,164-168. tion upon older people screened as depressed. International Journal of Geriatric P s y c k i a t ~10, , Dowrick, C. & Buchan, I. (1995) Twelve month 289-298. outcome of depression in general practice: does Blazer, D.G., Kessler, R.C., McGonagle, K.A. & detection or disclosure make a difference?BritSwartz, MS. (1994) The prevalence and distriish Medical ~ o u ~ 311,1274-1276. al, bution of major depression in a national com- Effective Health Care (1993) The treatment of k munity sample: the National Comorbidity depression in primary care. Efective ~ e a l t Care Bulle~in,5, 1-12. l 151, 979Survey. American ~ o u ~ofaPsychiatry, 986. Fertig, A., Roland, M,,King, H. & Moore, T. (1993) Mann, A., Blizard, R. & Wilkinson, G. Understanding variation in rates of referral Botega, N., (1992)Generalpractitionersand depression:first among general practitioners: are inappropriate use of the Depression Attitude Questionnaire. referrals important and would guidelines help International ~ournalof Methods in Psyckiatric to reduce rates? ~ r i t i s kMedical ~ournal,307, Researck, 2,169-180. 1467-1470. Brewin, C.R. & Bradley, C. (1989) Patient prefer- Freeling, P., Rao, B.M.,Paykel, E.S., Sireling, L.I.& ences and randomised controlled trials. Britisk Burton, R.H. (1985)Unrecogniseddepression in Medical J o u ~ a l299, , 313-315. general practice. Bri~iskMedical Journal, 290, Brodaty, H. & Andrews, G. (1983) Brief psycho1880-1883. therapy in family practice. Britisk Journal of Gask, L. (1992) Training general practitioners to Psyckiat~,143.11-19. detect and manage emotional disorders. Interna~ionalReview J o u ~ aolf Psyckiat~,4,2~y”-oo. Brown, G.W. & Harris, T. (1978) Social Origins of Depression. Tavistock, London. n Goldberg, D. & Huxley, P. (1992) C o ~ m o Mentffl Disor~ers:A Biopsyckusocial A p ~ r u a cRoutledge, ~. Brown, G.W. & Moran, P. (1994) Clinical and London. psychosocial origins of chronic depressive epiGoldberg, D.P. & Bridges, K. (1988) Somatic pressodes. I: Acommunity survey. Britisk Journal of entation of psychiatric illness in primary care Psyckiat~,165, 447-456. setting.Journal of Psyckosomatic Research,32,137Brown, G.W.,Andrews, B., Harris, T., Adler, 2.& Bridge, L. (1986)Socialsupport, self-esteemand 144. Goldberg, D.P. & Williams, P. (1988) Tke User’s 16,813-831. depression. Psyckolo~cal Me~icine, Guide to the General ~ e a l t Questionnaire. k NFER(1991) PovBruce, M.L.,Takeuchi,D.T. & Leaf, P.J. Nelson, Windsor. erty and psychiatric status. Arckives of General Goldberg, D.P., Cooper, B., Eastwood, M.R., p s y c k i a t ~48,470-474. , Kedward, H.B. & Shepherd, M. (1970) A Brugha, T.S. (1995) Depression undertreatment: standardised psychiatric interview for use in lost cohorts, lost opportunities? ~syckological community surveys, Britisk Journal of Preventive Medicine, 25, 3-6. and Social Medicine, 2418-23. Corney, R. (1992)The effectivenessof counselling in general practice. Inte~ationalReview of Psy- Gournay, K. & Brooking, J. (1994) Community psychiatric nurses in primary health care. ~ r i t i s k c k i a t ~4,331-337. , Jou~nalof P s y c k i a ~165,231-238. ~, Corney, R.H. (1984) The effectiveness of attached social workers in the management of depressed Greenberg, G. (1991)Clinical trials in general practice. Britisk Me~icalJournal, 303,940. female patients in general practice. PsyckologiGrimshaw, J.M. & Russell, I.T. (1993) Effect of cal Me~icine,Supplement 6,i-47. guidelines on medical practice: asystematic reCorney, R. & Murray, J. (1989) The evaluation of social interventions.In:n e Scope of E p i d e ~ ~ o 1 o ~ - view of rigorous evaluations. Lancet, 342, 13171322. cal Ps~ckiatry(eds P. Williams, G. Wilkinson & K. Ramsley), pp. 343-354. Routledge, London. Grol, R. ,(1990) National standard setting for quality of care in general practice: attitudes of Creed, F., Gowrisunkur, J., Russell, E. & Kincey, J. general practitioners and response to a set of (1990) General practitioner referral rates to disstandards. Britisk Journal of General Practice, 40, trict psychiatry and psychology services. Brit361-364. ish Journal of General Practice, 40,450-454. Guidance on guideCroft-Jefferys, C.& W i o n , G. (1989)Estimated Haines, A. & Feder, G. (1992) lines. Bri~iskMedical Journal, 305, 785-786. costs of neurotic disorder in UK general pracHoeper, E., Nycz, G., Kessler,L., Burke, J. &Pierce, al 19,549-558. tice. P s ~ c k o l o ~ cMedicine,
W. (1984) Theusefulness of screening for mental illness. Lancet, i, 33-35. Hollyman, J.A., Freeling,P., Paykel, E.S., Bhat, A. & Sedgwick, P. (1988) Double-blind placebocontrolled trial of a m i ~ i p t yamong ~ e depressed patients in general practice. Journal of the Royal College of General Practitioners,38,393-397. Jackson, G., Gater, R., Goldberg, D., Loftus, L. & Taylor, H. (1993) A new community mental health team based in primary care: a description of the service and its effects on service use in the first year. British Journal of Psychi~try,162, 375-384. Johnson, D.A.W. (1981) Treatment compliance in general practice. Acta Psychiatrica Scandinavica, 63 (SUPPl. 2091,447-453. Johnston, ME., Langton, K.B., Haynes, R.B. et al. (1994) Effects of computer-based clinical decision support systems on clinici~,performance al and patient outcome.Annals of ~ n t e ~Medicine, 120,135-142. Johnstone, A. & Goldberg, D. (1976) Psychiatric screening in general practice: acontrolled trial, Lancet, i, 605-608. Johnstone, A. & Shepley, M, (1986) Theoutcome of hidden neurotic illness treated in general practice. ~ournalof the Royal College of General Practitioners,36,413-415. Lin, E., Bush, & T.Ormel, Katon, W., von Korff, M,, J. (1992) Adequacy and duration of antidepressant treatment in primary care. Medical Care, 30,67-76. Katona, C., Freeling, I?.,Hinchcliffe,K., Blanchard, M. & Wright, A. (On behalf of the Consensus Group of the Royal Colleges of General Practitoners & Psychiatrists)(1995) Recognition and management of depression in late life in general practice: consensus statement. Primary Care Psychiatry, 1,107-113. Kendrick, T., Sibbald, B., Addington-Hall, J., Brennerman, D. & Freeling, P. (1993) Distribution of mental health professionals working on site in English and Welsh general practices. British Medical Journal, 307, 544-546. Kerr, M., Blizard, R. & Mann, A. (1995) General practitioners and psychiatrists: comparison of attitudes to depression using the Depression Attitude Questionnaire.Bri~ishJournalof General Practice, 45,89-92. King, M., Broster, G., Lloyd, M. & Horder, J. (1994) Controlled trials in the evaluation of counselling in general practice. British J o u ~ aofl General Practice, 44,229-232. Lewis, G. (1992) Computers in primary care. Interna~ionalReview of P s y c h i a t ~4,307-310. , Lewis, G. (1994) Assessing psychiatric disorder
with a human interviewer or a computer. Jour~ C o m ~ u n i t y ~ e a l48, th, nal of E ~ i d e m i o l oand 207-210. E. et al. (1988) The Lewis, G., Pelosi, A.J., Glover, development of a computerised assessmentfor minor psychiatric disorder. Psycho~ogicalMedicine, 18,737-745. Lewis, G., Pelosi, A.J., Araya, R. & Dum, G. (1992) Measuring psychiatric disorder in the community: a standardised assessment for use by lay al 22,465-486. interviewers.Ps ~c h o lo ~cMedicine, Lewis, G., Sharp, D., Bartholomew, J. & Pelosi, A.J. (1996) Computerised assessment of common mental disorders in primary care: effect on clinical outcome.Family Practice,13,120-126. Lloyd, K. & Jenkins, R. (1995) The economicsof depression in primary care: Department of Health initiatives. British Journal of Psychiat~, 27 (Suppl.), 60-62. Mann, A., Jenkins, R., Cutting, J. & Cowen, P. (1981) The12 month outcome of patients with neurotic illness in general practice. Psychological Medicine, 11,839-847. Marinker, M,, Will&, D. & Metcalfe, D.H. (1988) Referral to hospital: can we do better? British Medical Journal, 297,461-464. Marks, J.N.,Goldberg, D.P. & Hillier, V.F. (1979) Determinants of the ability of general practitioners to detect psychiatric illness. Psychological Medicine, 9,337-353. Meltzer, H., Gill, B. & Petticrew, M. (1995)OPCS t ~ Britain. Surveys of Psychiatric~ o r ~ i dini Great Report no. 2. The re valence of ~sychiatricm o r ~ i ~ ity among adults aged26-64 living in private households in Great Britain. HMSO, London. Mynors-Wallis, L.M.,Gath, D.H., Lloyd-Thomas, A.R. & Tomlinson, D. (1995)Randomised controlled trial comparing problem solving treatment with amitriptyline and placebo for major depression in primary care. British Medical ~ournal,310,441-445. North of England Study of Standards and Performance in General Practice (1992) Medical Audit in General Practice. 1: Effect on doctors’ clinicalbehaviour for common childhood conditions. British Me~icalJournal, 304, 1480-1484. Old Age Depression Interest Group (1993) How long should the elderly take anti-depressants? A double-blind placebo-controlled study of continuation/prophylaxis therapy with dothiepin. British Journa~of Psychiatry, 162,175182. Ormel, J., van den Brink, W., Koeter, M.W.J.et al. (1990)Recognition, management and outcome of psychological disorders in primary care: a
naturalistic follow-up study. PsychologicalMedi- Sibbald, B., Addington-Hall, J., Brenneman, D. & cine, 20,909-923. Freeling, P. (1993) Counsellors in English and Ormel, J., Koeter, M.W.J., van den Brink, W. & van Welsh general practices: their nature and distride Willige, G. (1991) Recognition, management bution. Bri~ish~ e d i c aJournal, l 306,2933. Song, F., Freemantle,N., Sheldon, T.A. et al. (1993) and course of anxiety and depression in general practice. Archives of General P s y c h i a t ~ , Selective serotonin reuptake inhibitors: rnetaanalysis of efficacy and acceptability. British 700-706. Ormel, J., Oldehinkel, T., Brilman, E.& van den Medical ~ o ~ r n 306,683-68~. al, Teasdale, J.D., Fennell, M.J.V., Hibbert, G.A. & Brink, W. (1993) Outcome of depression and anxiety in primary care:a three-wave 3+ -year Amies, P.L. (1984) Cognitive therapy for major study of psychopathology and disability. depressive disorder in primary care. British Archives of General P s y c h i a f ~50,759-766. , ~ournalof Psychiafry, 144,400-406. Paykel, E.S. & Priest, R.G. (1992) Recognition and Tognoni, G., Alli, C., Bettelli, G. et’ al. (1991) management of depression in general practice: Randomised clinical trials in general practice: consensus statement.BrifishMedical ~ o u ~ 305, lessons from a failure. British Medical ~ournal, al, 1198-2202. 3031 969-971. Tyrer, P.,Sievewright, N. & Wollerton, P. (1984) Paykel, E.S., Hollyman, J.A., Freeling, P. & Sedgewick, P. (1988) Predictors of therapeutic General practice psychiatric clinics: impact on benefit from ami~iptylinein mild depression: psychiatric services.Bri~ish ~ o ~ r ~ a ~ o ~ ~ s ~ a general practice placebo controlled trial. Jour145,15-19. nal of A ~ e c f i vDisorders, e 14,8395. US Department of Health and Human Services Platt, S., Martin, C. & Hunt, S. (1990) The mental (1993) D ~r e s s io nin P r i ~ Care. a ~ Vol. 2: Treafhealth of women with children living in demenf of Major Depression. Agency for Health Care prived areas of Great Britain: the role of living Policy and Research, Rockville, Maryland. conditions, poverty and unemployment. In:The Ustun, T.B. & Sartorius, N.(eds) (1995)menta^ IllP u ~ l i Health c ~mpactof Mental Disorder (eds D. ness in General Healfh Care. John Wiley & Sons, Chichester. Goldberg & D. Tantam), pp. 124-135. Hogrefe von Korff, M., Shapiro, S., Burke, J.D.et al. (1987) and Huber, Toronto. Roland, M.(1992) Measuring appropriateness of Anxiety and depression ainprimary care clinic. Archives of General P s ~ c h i a t44,152-156. ~, hospital referrals. In: Hospifal Referrals (eds M. nf Me~tal Roland &A. Coulter), pp. 137-149. Oxford Uni- Warr, P. (1987) Work, ~ n e ~ p l o y m eand Health. Oxford Science Publications, Oxford. versity Press, Oxford. Russell, 1. & Grimshaw,J. (1992) The effectiveness Waydenfield, D.& Waydenfield, S.W. (1980) Counselling in general practice. ~ o u r n aof~the Royal of referral ~ i d e l i n e sa: review of the methods College of General Pracfifioners,30,671477. and findingsof published evaluations. In: Hosp i f ~ReferraZs l (eds M. Roland and A. Coulter), Weich, S., Churchill, R., Lewis, G. & Mann, A. (1997) Do socio-economic risk factors predict the pp. 179-211. Oxford University Press, Oxford. incidence and maintenance of psychiatric disScott, A.I.F. & Freeman, C.P.L. (1992) Edinburgh order in primary care? ~ s y c h o l o ~ c a l ~ e d27, icine, primary care depression study: treatment out73-80. come, patient satisfaction and cost after 16 Wilkinson, G. (1992) The role of the practice nurse weeks. British Medical ~ournal,304,883-887. in the management of depression.~ ~ f e r n a f i o n a l Scott, J., Moon, C.A.L., Blacker, C.V.R. & Thomas, Review of P s y c h i a f ~4,311-316. , J.M. (1994) A.I.F. Scott & C.P.L. Freeman’s Wilkinson, G., Allen, P., Marshall, E., Walker, J., ’Edinburgh Primary Care Depression Study’. Brown, W. & Mann, A.H. (1993) The role of the British ~ournalof P s y c h i a t ~164,410-415. , practice nursein depression in general practice: Shah, A. (1992) The burden of psychiatric distreatment adherenceto antidepressant medicaorder in primary care. ~nfernational Review of tion. PsychologicalMedicine, 23,229-237. P s ~ c h i a t4,243-250. ~, Shapiro, S., German, P,,Skinner, E. et al. (1987)An Wright, A.F. (1994) Should general practitioners be testing for depression? Brifish ~ o ~ r ofn a ~ experiment to change detection and managegenera^ Practice, 44,132-135. ment of mental morbidi~ in primary care.MediZung, W.W.K., Magill, M.,Moore, J.T. & George, cal Care, 25,327-339. D.T. (1983) Recognition and treatment of depresShepherd, M., Cooper, B., Brown, A.C.& K a h n , sion ina family medical practice. ~ournalof CliniG. (1966)PsychiatricDisordersin GeneralPractice. First edn. Oxford University Press, Oxford.
3
le 15.1 The structural componentsof a comprehensive mental health service. Identification and needs assessment systems Range of hospital and community beds Case management and assertive outreach Day care, rehabilitation, education and work Crisis response Assessment and consultation Carer and community education and support Primary care liaison Community agency liaison User involvement and advocacy
ea
S
GMHTs have their origin inthe American Community ental Health Centre programe of the 1960s. Theteams weremandated to could respond to the needs of the severely mentally ill (§MI). However, when located in community mental health centres, the focus of work shifted to ‘the worried well’ resulting in the neglect of those with severe mental illness (Sayce et al., 1991). Perhaps exacerbated by the strength of its primary care infrastructure where the vast majority of mental health morbidi in Britain have followed a similar pathway. A Sainsbur identified over300 GMHTs practising across England, comprised of at least four members fromtwo or more disciplines,with a caseload of individuals with mental healthdisorders largely residing the in community 1994). In 1994 focus on the work, organization, service dev proaches and management styles in ritish teams formed the subject of intensive studies by a numberof nationa gencies (Audit Gommission, 1994; 3rd nquennial Community Psychiatic Survey; White, 1994; Health of The ons, 1994; Mental Health Foundation,3-994; Ritchie, 1994). the reviews comment on how variable the targeting of the appears to be. In some instances, less than 10%of the comunity psychiatric nurses working with the S I in community services le with either manic-depressive disord or schizophrenia,
A strategic approach to the planning and implementation of services and working practices is essential if a CMHT is to effectively focus its work on the most vulnerable patients with the most complex needs. Primarily these include patients with major depressionand schizophrenia. Thosewith moderate degrees of depression not responding to management by primary care
physicians, and patients presenting with somatic problems repeatedly to their general practitioners (GPs) also need to be regarded as a priority. A large number of those with less severe depression remain unseen by specialist services mainly because of resource limitations, atypical presentations or inability to benefit from the existing style of service delivery. Women with young children, ande t h i c minorities, for example, fail to optimally benefit and dis-are cussed in more detail below. Yet GPs (primary care physicians) report that they are unable to influence the illness episodes of many patients who are ineligible for care by secondary specialist services. This has resulted in the increased use of counselling and alternative therapies which are purchased by GPs. This occurs in the absenceof an effective and financially viable role for the secondary specialist mental health services. These specialist services have focused largely on those suffering from severe depressive disorders with overt suicidality and behavioural disturbance. Yet, such service providers identify the independent deployment of counselling and psychotherapy services in primary careas detrimental to the provision of existing secondary services and as being of questionable long-term benefit. A comprehensive strategy must therefore aim to reach those groups not served by current services. Any strategy shouldbe informed by an understandingof the theoretical evaluation/ research and practice findings in community care programmes. It is equally importantto have information detailing where, and by what adaptations, such international findings can inform good practice suited to unique local service requirements and patients with special needs. Such an evaluation must take account of the limitations inherent in well-funded 'model' programmes and harness transportable pragmatic working practices (Test &C Scott, 1990; Audit~ o ~ s s i o1994; n , Wood, 1996). Table15.2 outlines the key elements such a strategy should embrace. First, a system for the assessment and recognition of the needsof patients with severe, moderate and mild depression is necessary. Second, a network of service structural components developed to take accountof the research evidence and with a clear ~derstanding of the outcomes they aim to achieve. Third, specific attention should be given to the availabilty of a rangeof effective therapeutic interventions. Such interventions must address the wider
Table 15.2 A strategy for the effective deployment of CMHTs. Knowledge of the epidemology of the local area Strategy for the identification of those with depression Development of the structural components Use of effective interventions Effective organization and management of the CHMT Planning for staff skill mix, training and support
needs (including basic social re~uirements)of those with depre§sion and should have demonstrated improvements in health andsocial outcomes. This assumes that those delivering these interventions are trained todo so in an optimal m a ~ e rNext, . explicit and effective o r g a ~ ~ a t i o nand a l management t e c h ~ i ~ umust e s be used with^ the C HT to th~roughlyaddress the service demands within the alloca budgets. Fundamentally, this will involve clear interdisciplinaryproceduuidelinesaswellas mu~ally understood,rereed roles. Finally, an analysisof the skill mix of the CMHT ge of effective interventions held by the team must be carried and i n t e ~ e n ~ oshould ns meet the needs of the target popu~g a~a~ment uld be s~ucturedo portunity forc o n ~ n u skill ref~ement ~ o u personal ~ h ~ a development ~ gp r o g r ~ efor s all staff. urnout among§t community teams drastically reduces the momentum of gic plans. Preventive management strategies should be em~loyedto lessen thepossibility of an eventual manpowercrisis.
The range of needs of those withmental health disorders.
Needs and effective ~te~entions
The long-term menta~lyill
Case identification and case register recall +++ Crisis interventionin the community and at home +++ Medication +++ Respite and crisis community beds +++ +++ Family education and support + Marital/sexual counselling + ~ehavioural/co~itive therapies ++ Supportive counselling +++ Case ~anagementand assertiveoutreach Physical care +++ Social skills and stressm ~ a g e m e n t +++ Day care +++ ++ Welfare benefits advice ++ Mousing + Legal advice ++ Support group/self-help group Information and education about illness and services+++
Common primary care conditions
+ + + + ++ +++ +++ +++ + ++S +C+
+ ++ ++ + +++ +++
Table 15.3 illustrates the similarities between the specific potential umet needs of those with severe affective disorders (manic-depressive disorder or major unipolar depression), and those of patients with more common forms of depression and neurotic disorders, althou h differences in degree and profile r e ~ a i nSuch . a body of needs cannot alliances between health, social services, ho
3
roviding a comprehensive, complementary and cost-efficient service framework.
In order to plan strategically,co~munity a team must first assess the full extent of the potential demands upon them. As a starting point the C develop, over 2-5 a year timetable, a detailed profile of the number and needs of all individuals with the most severe disorders in the defined locality. In the context of affective disorders these include those with bipolar and unipol depression, a historyof suicide attempts, repeated deliberateself harm an previous violent acts. Special emphasis should be placed on those with clearly defined comorbid conditions which are undermining to their capacity for optimal work with providers and place them at higherofrisk suicide on the basis of empirical epidemiological and clinical findings. In Britain, the Office of Population and Census National vey produced s ~ ~ findings ~ ssuggesting ~ g a wider variati ence of psychosis in the UK than had previously been identified (OPCS, 1995). This varies from0.2% in affluent areas to0.9% in more deprived areas. Overall about14% of adults had some sort of neurotic health problem as measured by the CIS-R, Depressive ideas were present amongst 5-14% of cases with 71 respondents per thousand reporting mixed anxiety and depressive states. It clearly demonstrates that one in seven of the population suffer from a neurotic disorder, most often depression. By using locally applied epideical techniques, teams can estimate the expected number of individuals with depression for their area and at least plan with some reliability. GPs are responsible for treating the bulk of psychiatric disorders. Up to a third of GP attendees may have psychosocial problems but approximately~ half of these may not be detected (Goldberg & Huxley, 1992; Wright, 1993). Depression and anxiety states form 90% of the mental illness caseloadGPs, of with 5% reaching D~M-111-~ caseness and10%having impaired functionin because of depressive symptoms (Paykel & Priest, 1992). Of those with major psychiatric morbidity,only a proportion willbe in contact with the specialist services. In their long-term follow-upof a cohort of Maudsley patients with (1988) found that over 50% had lost contact with depression, Lee and Murray mental health services. Between 25 and 40% maybe in contact only with their GI>,social services, the housing agencies, police, local churches or user groups. A. fundamental building block in establishing an effective team is the de . opment of an identification or case finding component of the service. establishment of lar liaisons with all other agencies who may be serving corn^^^ is essential. Table 15.4indicatesthe theneeds of theinthe likely contact points of people with mental health disorders.
Table 15.4 Identifying peoplewith mental health disorders: contact sources. Mental health service contacts Out-patient and domiciliary visit records CPN case-load and depot clinic patients Mental Health Act and in-patient audit data Crisis attenders, e.g. A/E attenders CPA, Supervision register,S.117 (in GB) Residents of long-stay institutions Primary care team contacts Practice register diagnoses Repeat psychotropic drug prescriptions of antipsychotics and antidepressants Frequent emergency and other consultations Hostel/group home/sheltered residence populations CMHT attenders and health visitor contacts Social service contacts Area social worker case-loads Care management clients Housing department clients causing concern ~ o ~ ~ ~ sector t a and r y other agency contacts Residents of sheltered accommodation Individuals in distress presenting to churches Individuals causing local police officers concern Imprisoned and homeless people Probation officer case-loads Drop-in and other casual facility users User groups
Many services in Britain have used the implementation of the Care Programme Approach (CPA) as a vehicle to audit how the number of SMI in contact with the service compares with that expected from the epidemiological data.
Having identified individuals with depression, the nextisstep to introduce a system of needs assessment. Table 15.5indicates the full range of needs of people withmental health disorders.If an assessment failsto include attention to the social as well as clinical needs, then long-term outcomes are ~emonstrablypoorer (Craig et al., 1994). Scales commonly used include the Hospital Anxiety and Depression Scale (HADS), the Beck Depression Inventory, the Montgomery Asperg Depression rating scale and the Zung selfrating scale.We refer those interestedin a more detailed and comprehensive list of scales and their specific application to examine Freeman Tyrer,and 1989. In Britain, the basis for an agreed interagency definition of the service priority group and categories for needs assessernent has been formed by services in whichlocal health and social services colleagues have integrated
330
le 15.5 Effective interventionsfor people with depression. Housing with adequate support Welfare benefits and financial advice Rehabilitation, practical skills and support Work and education Physical and dental care Medication and psychoeducational programmes ~ e h a v i o u r a ~ - c o ~ itherapy tive for psychotic symptoms Cognitive analytic therapy Identification and development of coping strategies Crisis prevention and relapse prevention Family problem-solving therapy Counselling Formal psychoanalytic psychotherapy Complementary practitioners Spiritual/religious support
the CPA and Care Management. The routine collection of a clinical data set and care planning assessment and review is vital. When this has been incorporated into the C HT culture it can ensure that service developments are based on the aggregated needs of patients and can inform the case for resources It also facilitates multiagency co-ordination of service planning and delivery.In sum, it is vital at a nationallevel for the mental health agenda (Kingdon, 1994).
The central taskof the CMHT is to establish systems which ensure a continuof care. ing focus on the severely mentally ill and provide effective packages In many areas, traditional ’stand-alone’ and tertiary rehabilitation services are being decreased in size or replacedby the CMHT. A range of methods are employed by the CMHT in order to ensure that they continue to prioritize the severely mentally ill. They may have an agreed case-mix for each team member’s caseload, or use a small number of dedicated case managers. In areas with high levels of morbidi~,small separate teams of case managers providing an assertive outreach function are regardedas the most effective way of maintaining a continued focus on the most vulnerable client group (Ford et al., 1996) Case mana~ement as a service(or at least a technique)is essential to combine all the necessary service elementsinto one coherent packageof care. In practice, case management for the long-term mentally ill has developed into 1 2 different axes (~o~croft, a rangeof approaches that can be described along 1990).These techniques aim to ensure that patients receive consistent and ~ontinuing services foras long asis necessary. Several studies have compared
Lelliott and Wing’s (1994) study of the new long-stay found that 61% of patients were considered to be inappropriately placed in hospital beds, 47% required a community-based residential setting, and of these overhalf were in hospital because no suitable c o ~ u n i t placement y was available. Several analyses of acute bed unit use indicate that 30-70% of those in the beds are ’revolving door regulars’, well known to services. A significant proportion had been readmitted within 3 months of discharge due to a lack of high support community accommodation and effective discharge planning mechanisms. The successof communi~-based services is crucially related to the nature and availability of accommodation with appropriateof support, levels whether provided in individual’s homes,by local authorities orby independent sector HTs must havea wide rangeof hospital andCommunity beds available to them. These should range from responsive and communitysensitive secure facilities for mentally disordered offenders, through to acute hospital beds. Twenty-four-hour staffed units in hospital hostels and community places for the new long-stay especially those with mild to moderate challenging behaviours needbetoprovided. Finally,a range of community beds from crisis diversion, quarter- and half-way hostels and respite facilities through to residential and supported permanent accommodation are also vital (~trathdee,1996). CMHTs are ideally placed to play a lead in ensuring the planning of such facilities if they are able to present concrete local data on the numbers and needs of their patients and the number and range of bed provision needed.
In addition to planning and ensuring the development of an adequate number and range of beds, community teams have aused range of techniques to prevent unnecessary and inappropriate acute bed use. The organizationof their service ona defined sectorbasis facilitates comprehensive case identification, needs assessment and effective interagency planning.In a study where the CMHTs’ senior and experienced clinicians carried outinitial assessments in the patient’s home instead of hospital out-patient clinic sites, admission was et al., 1993). Gate-keeping (i.e. being involved in reduced by 50-7070 the decision to admit a patient) by senior doctors and nurses can assist junior staff who may be less experienced in risk assessment and the deployment of safe alternatives to admission. Consistent research findings demonstrate that integrated hospital and c o ~ ~ u ~services i t y are essential to maximize the use of resources. Unless community teams have control over their own hospitalbeds, both bed use and length of stay are significantly increased.
urns
333
uals with depression arevulner can exacerbatesymptoms and foster r support and help for the client and their family members/carers e crisis, while maintainingthe client’s status as a functioningcomer. Studies of users, carers and GPs produce almost interdings. All want crisis services to be accessible from a single or face-to-face contact. They would value services providin facilities as a viable alternative to hospitalization on a 24our, 7-day basis (Phelan et al., 1995). uch services need to be m a ~ e by d experienced and known mental health professionals. However, on a national scale this type of service is provided by a very limited number of teams (Johnson & Thornicroft, 1995). S
In a comprehensi~ereview, Holloway (1988) defines several possible functions of the day hospital. It canprovide an alternati~eto hospital admission when acutely ill (a model well elaborated by Creed (1995)), support, supervision and monitori during the transition between hospital and home. It is usually a sourceof g-term structure and support for thosewith long-term ifficulties and a site forbrief intensive therapy for thosewho require shortterm focused rehabilitation.Finally, it can offeran i~ormation,training and communication resource.Wood (1994)~as the occupational therapist responsible for the development of day-care services for aC ~ H Tstresses , the importance of user consultation and the e s t a b l i s ~ e nof t formal liaison with other localproviders of work, educationa d day care. In one service, model ecialists arepart of the c o ~ u nmental i ~ health centre t services in the c o ~ u nteam i ~ site. Emphasis is plac client preferences, rapid job findin ,continuous assessment, CO employment, integrated work settin S and follow-along supports rake, 1994). Furthermore, the range of social needs differ little between those with affective disorders and those with schizophre~aand length of duration of contact with a servicehelps to re-establish socialand family supports.
Up until the past few decades,the majority of consultation services were conita1 out-patient settings. The evaluation of these services indiction with communication patterns and clinical and referrer outcomes. l on-attenda~ce rates wereshown to beup to 45%. The response
33
of clinicians, uni~ormedby policy or research considerations, was to work on a sessionalbasis in primary care centres. The sessional clinics were criticised early on for running a service that could result in psychiatrists serving only ‘the worried well’, particularly those with depression and anxiety. ever, the evidence now refutes this. They generally provide a service to many of those who either did not have a service or who had dropped of contact out with the specialist services. Such isolated groups include women with longterm disorders, young men with schizophre~ia,the homeless with psychosis and those with paranoid disorders who feel stigmatisedby attendance at a hospital site (Tyrer, 19
In Britain where there is a large primary care i~rastructure,CM developed a number of models to establish a system of liaison. In ar GP fundholders are common some have helped to fund small, integrated and practice-based primary care liaison teams. More commonly, co~unity teams have dedicated specialist workers who act as the single access point and gatekeeper for referrals. Others have evolved a system whereby each member of the team is the ’link worker’ for a particular general practice. nearly all cases the appropriate model is best determined by local circu stance. Liaison is easier in areas where the primary care teams are based in large health centres. WhereGP surgeries are widely dispersed (particularly in h e r city areas), the establishmentof feasible systems becomes far more problematic. Table 15.6 proposes a range of initiatives to improve joint working between primary health care teams and the CMHT (Strathdee 1995). The measures are based on the successful shared care model developed for other patient groups with longer-term disorders such as diabetes and asthma. A study examining the effectsof a community mental health serviceon the practice and attitudes of general practitioners indicated that GPs with access to a psychiatric team were significantly more satisfied with the specialist support services. They were also more likely to give high priority to community psychiatric nurses and psychiatric social workers working as part of et a primary health care team if they facilitated access to the service (Warner al., 1993). Ensuring that specialist mental health workers liaise with primary care services has not always proved easy. Community psychiatric nurses in many c o ~ u ~services t y were the first to develop attachments to particular general practices, Where the nurses have been hospital based and work as mem80% of their referrals are from psychiatrists, bers of the secondary care team, Thus, individuals with §MIform a considerable proportion of their work-
335
T~~l 15.e6 Interagency and primary care liaison in the care of individuals with depression.
C o ~ ~ u n i c a t istrategy on A directory of street boundaries, sector team names, roles and contact numbers Booklets describing therapies available Named contacts to advise on appropriate referrals Crisis response Single and well-publicized telephone contact access point Agreed crisis response time-scales, roles and responsibilities Assess~ent and out-pa ti en^ services Agreed referral-appointment interval Stated referral criteria for members of the CMHT Clear comm~ication with stated objectivesof management, predicted response, complications and side-effects 6-monthly review plans for long-term patients Clearly stated roleof GP and specialist in treatment Clarification of prescribing responsibilities S ~ a r e care d for individuals with d ~ r e s s i o n Set up a joint case register similar to diabetes Prioritize the most severely ill Agree annual physical care reviews Ensure GP verbal/physical/fax input into CPA care planning reviews Identify GP/practice nurse role in development of relapse prevention plans Agree protocols for group homes/hostels Discuss consequencesof expensive new psychotropic drugs and ECRs Obtain support for appropriate range of beds andc o m ~ t alternatives y to hospital admission
load. Where nurses, although employed by secondary care services, are attached to particular general practices, the referral pattern shifts, with 80% of GPs. Caseloads tend tobe large and their referrals coming directly from the composed of patients with neurotic and a d j u s ~ e n tdisorders. s Although the caseloads of both hospital and primary care-based CPNs remain similar in terms of the number of individuals with schi~ophrenia,the mean time in contact with psychotic patients amounts to one-third of the time spent with nonpsychotic patients and almost entirely limited to the administration of injections.
Whatever the structural links between primary and secondary care teams ar the important question is how best to provide effective, evidence-based interventions to the maximum number of patients. A two-pronged approach seems appropriate. First, the skills of the primary care team who already provide the majority of care need to be enhanced and appropriate training given. Second, those with more severe and moderate difficulties will need and
benefitfromdirectcare by the C Thecommunitymentalhealthteam is particularlyhelpful in reducinrdenposedbypatientswithneurotic and psychosocial problems, but can result in de-skilling the general practitioners who end up undertaking less mental health work themselves (~~rner et al., 1993).An educational approach therefore seems most likely to lead to overall improved care in the longer term. Table 15.7illustrates some of the training found tobe useful and effective for primary health care staff. Table 15.7 Developing effective interventionsin primary care. CMHT initiatives Provision ofa resources directory of all statutory and other mental health resources locally, e.g. day care, dropins, counselling services Development of jointgood practice protocols for common conditions such as depression Employment and supervisionof primary care-based counsellors, psychologists, etc. Provision training to PHCT Provision of updates on new psychotropic medication, therapies and information leaflets, videos, etc. PHCT initiatives: increased training for: Practice nurses Recognition and management of depression, stress management Mental state assessment Administration and reviewof medication Assessment of tardive dyskinesia Relapse prevention Psychoeducation ~ e a visitors ~ ~ h Recognition and prevention of depression in young mothers/children Recognition of depression in the elderly and chronically physically ill Problem-solving for mentally ill parents Problem-solving for parents of adolescents District nurses Recognition of depression in the elderly and physically ill Receptionists: Dealing with difficult patients
One i ~ p o r t a nreason t for persistent depressive symptoms is the failure of patients to adhere to the agreed treatment plan. of The psychotropic use medication to treat depressive symptoms has proven effective, particularly if combined with a psychotherapeutic and social problems management approach. Yet up to 50% of patients may not be taking their medication in sufficient amounts. Strategies to establish a dialogue with patients about their objections to certain medication regimens are an essential part of the community management of depression where intensive 'medication' supervision is not available. Poor compliance canbe influenced by specific factors related to the medication, the patient, the illness experience or the treat-
337
rnent setting (Fawcett, 1995). Strategies desi ed to improve the uptake of include keeping the regimen simple, tailoring to daily rituals, explicit written instructions and implementing drug regimes in providing process and patient olving the therapeutic thefeedback comp of goals for 1985).A fuller accountof nonalso helps to encourage compliance (~hurchill/ compliance with psychotropic medication is given elsewhere (Frank et al., 1995; Sair et al., 1995).
anaging the depressed patient inc the o ~ ~without t y admission requires a careful assessment and reviewof suicide risks as wellas the risks of each t~eatmentdecision. One should consider each of the risk factors associated with suicideas illustrated in Table 15.8. .8
Risk factors associated with suicide.
History of depression or schizophrenia History of alcohol and/or substance misuse Family historyof depression, schizophrenia, DSH, completed suicide Family historyof alcoholism Previous DSH/suicide attempt Stressors and capacity to deal with them: intelligence and supports Comorbidity: physical illness, bereavement, chronic pain Presence of behaviour d~~rbance/violence/impulsivity Level of supervision availableif stays at home and providers of supervision Past rangeof treatments and response to treatments Previous compliance with treatment Expressing suicidal ideas Objective measuresof depression (scales) and suicide intent scales Are family and supporters in crisis themselves such that admission is necessary?
If adequate supervision can be provided at home by a patient’s family and rofessionals during home visits, one shouldexplicitly document the risk factors, the potential outcomes, the likelihood of each occurring and the potential severityof each outcome.hexplanation as to why the treatment packcommuni^ treat~entrisk also needs to be age should effectively contain the included. The motivation for the decisions and the ofstages reasoning should e clear instead of continuing to only document the final decisions. Each decision shouldbe allocated a risk time period after which. the decision its ons sequences are reviewed.A plan maybe reviewed every12 to 24 h initially, depending on the level of risk carried and what can clinically be ed as a safe risk period. nce some improvement has occurred and the level of despair expressed isless, then the risk periods could be increased. In addition, ones h o ~ l d also involve the patient in a crisis plan if it is anticipated
33
that they may experience profound suicidal ideas. ffective communication with the primary care teams c also help prevent a crisis and re risk as theycancarry out so the reviewvisits. Daycaremayhelpto relieve relative§ and carers an patients a sense of support. ~urthermore, day care services can help assess levelof dysfunction and provide occupational and social i n t e ~ e n ~ o which ns may beof specific importance forthe individual.Some suicide scales enable the enumeration of risk ple scales can be constructed for use by specific teams. Am example of ris assessment in the CO management of mental1 di§ordered offenders has been well desc rson, 1993) and could state ies for the community care of de must take accountof known epid
In all the major epidemiological surveys women have been found to have 1.4-3 times higher rates of depression (Juel-~ielsonet al., 196%; 1961; Paykel et al., 1970).This is largely accounted forunipo~ar by as the sex ratio for bipolar dis der is 1 :1. Several hypotheses have been advanced to explain these find (Paykel, 1991). omen may be more li to seek help when depressed, and re ch is being carried out into the impact levels of female sex hormones a netic differences atso son 1990). Social ~ u ~ e r a b i absence l i ~ , of support, low self-esteemand women’s general social position are thought to beimport~ntfactors influencing mild to moderatede ression ( rown & Harris, 1978; Brown & Prudo, 1981; Brown rris et al., 1991). A final theory suggests that each sex perceives, interacts with and responds to distress in a manner which is depressogenic in women and gives rise to other disorders in men. The latter describes a h ~ o ~ e sisisupported s by ~ ~ oet al.~ (1971) u who r sive spectr~mdisorder’ which is manifest as alcoholism and personality disorder in men and depressio~in women. ~urthermore,the diagnostic of women presenting GPs to differs from their male counterparts. The of the intervention can also differ: specifically, women ma be offere benzodiazepines more often, their distress being conceptualiz ety state (Ashton, 1991). omen developing severe and chronic mental illness are less likely to receive an effective inte~entionwhich limits and appropriately ends their illness episode (Perkin R o w l ~ d1991). , ~hronicityof § ~ p t o mmay s therefore be a functionof i ible services structures, as well as a failure todi nose depres§ion accurately an administer the a propriate treatment.
339
social needs of women should be addressed in conjunction with social services, housing, child care services, educational establis~ents,parenting services and relationship counselling to ensure specific provisions are made. A sound infrastructureof flexible and responsive health service delivery must exist if the services are tobe fully effective. ome t r e a ~ e n t are s increasin~ly demonstrated to be a patient preference. For mothers or women who receive their main source of support from their immediate communities, assessments and treatment are most appropriately delivered at home. This will help to maintain and reinforce already established social support rather than undermine and replace long-standing support structures with less flexible, less familiar and rationed statutory equivalents.omen-only services including admission wards are especially valuable when considering the range of services suitable for ethnic minorities or specific religious groups.
Evidence showsthat the African and Caribbean populations of Britain have a 3-6 times higher incidence of diagnosed schizophrenia than white people arrison et al., 1988; Littlewood&:Lipsedge, 1981; Wessley recent study extendedthis finding to all ethnic minorities( By comparison, the rates of depression and anxiety disorders as identified amongst ethnic minorityGP attendees andby hospital admissions, appears to be lower than those identified for white attendees ( ebbington et al., 1981; Cochrane, 1977; Birchwood et al., 1992). Gillamet al. (1989) demonstrate^ that Black and Asian people attended their GP as often as the control group,but often attended for physical complaints rather than for specific mental illness. The outcome seemsto be somewhat paradoxical, as factors such as social adversity, including racism, poor housing and employment, which are thought to contribute to higher ratesof psychosis are also expected to lead to higher rates of depression (Lloyd,3-993). Little attention has been given to the systematic identification of depression amongst black communities. ~ o r s e n i n gsuicide rates in the USA, especially amongst young black men, has alerted British psychiatrists to the possibili~that black people appear to present when in crisis. Thus, effective interventions may not occur at the early of stage an illness episode or the intervention sought by a depressed patient is not that offeredby existing health services. Distress experiences may also present differently amon~stminorities because of an inherently different senseof self and object relatedness (Bhui et al., 1995); first rank symptoms, for example, areof less diagnostic significance amongst some communities (Fernando, 1988). Unfamiliar idioms of distress employedby ethnic minorities may remain unrecognized, causing a significant delayin the implementationof therapeutic interventions.
an women are increas
s~ategiesshould be developed in alliancewith patients and their relatives. Most depression scales developed the on basis of symptom profiles are likely to beas insensitive tode ression across cultures as clinical assessments. suggested solution is to employ culture experts or advocates to help with the of children and other family reco~ition and diagnosis of symptoms. The use members (e.g.parents or siblin~s)should be discouraged asit is invariably likely to give riseto a distorted version of events and symptom appraisal.
e management of the majority of individuals with depression has taken ce in the community settin ,The challenge forthe future is to determine the most effective ways in which agencies working in the liaise and work together.This applies particularly C to o m ~ Teams and Primary ealth Care Teams. Agenciesand teams need to examine how best they can acquire and use the effective ~terventionsknown to achieve optimal outcomes for the patients they serve.
Ashton, H. (1991) Psychotropic prescribing in lni~iative:An Evalua~ion.HMSO, London. 158 (Suppl.), Creed, F. (1995)Acute day hospital care in E ~ e r g ~ c y women. British J o u ~ aofl Psychiat~, Mental Health Services in the Community, (eds 30-35. Audit Commission (1994) Finding a Place: A RePhelan, M.,Strathdee, G. & Thornicroft, G.) view of Mental Health Servicesfor Adults. HMSO, Cambridge University Press,Ca~bridge. London. Dean, C. & Gadd, E. (1990) Home treatment for Bal, S. (1984) The Symptomatolo~of Mental 11lness acute psychiatric illness.g~itishmedica^ JournaZ, amongst. Asians in the West ~ i ~ l a nBA ~ sDis. 301 (6759), 1021-1023. sertation. Departmentof Economics and Social Fawcett, J.(1995)Compliance: definitions andkey Sciences, Wolverhampton Polytechnic. issues. Journal of Clinical Psychiatry, 56 (Suppl. Becker, D.& Drake, R. (1994)IndividualPlacement l), 5-10. center Fernando, S. (1988) Race, Culture and Ps y c h ja t ~, and Support:a community mental health Commuapproach to vocational rehabilitation. Routledge, London. nity Mental Health Journal, 30 (2), 193-206. Ford, R., Ryan,P.,Norton, P., Beadsmore, A., Craig, Bebbington, P., Hurry J. & Tennant, C. (1996) Psy- T & Muijen, M. (1996) Does intensive caremanchiatric disorders inselected immigrant groups agement work? Clinical, social and quality of , in Camberwell. Social Ps y c h i a t ~16,43-51. a controlled study. Journff~ of life outcomes from Belliapa, J. (199%)Illness or distress? Alter~fftive Mental Health, 54,361-368. models of Mental Health. Confederationof Indian Frank, E.Kupfer, D. & Siegal, R, (1995) Alliancenot Organisations(UK). compliance: a philosphy of outpatient care. Bhopal., R. (1986) The Inter-relationshipof Folk, ~ o u of~C ~ a i ~ i c a l P s y c56 h ~(Suppl. a ~ ~ , I), 11Traditional and Western Medicine within an 17. Asian Community in Britain. Social Science in Freeman, C. & Tyrer, P. (1989) Resear~h Met~ods Psychiatry. Royal College of Psychiatrists, Me~icine,22 (no. x), 99-105. Bhui, K.,Christie, Y. & Bhugra, D.et al. (1995) London. Essential elements in culturally sensitive psy- Gillam, S., Jarman, B. & White, P. et al. (1989) chiatric services.lnternationalJournal Social PsyEthnic differencesin consultation rates in urban chi at^, 41(4), 242-256. British ~ e d i c aJournal, l 299,953general practice. p tAsian o ~ s patien~s. Bhui, K.(1994) S o ~ a t i c s y ~ in 957. Goldberg, D. & Huxley P. (1992) C o m ~ o Mental MSc thesis. UMDS, University of London. n Disorders. Routledge, London. Birchwood, M., Cochrane, R. MacMillam, E et al. (1992). The influnce of ethnicity and family Goldman, H. (1981), Defining and outlining the chronically mentally ill. Hospitaland C o ~ ~ u n ~ t y structure on relaspefirst in episode schizophrePsychiatry,32,21"27. nia. gritish J o u ~ aofl P s y c h i a t ~161,783-790. , Brown, G.,& Harris, T,, (1978)Social origins of de- Goldman, H. & Taube, C. (1988) High users of pression. A study of psychia~vicdisor~erin omen. outpatients mental healthservices, I1 implications for practiceand policy. American Journal of Tavistock, London. Ps y c h ia t ~, 145, 24-28. Brown, G., Andrews, B., Harris, T. et al. (1986) Harrison, G. et al. (1988) A prospective study of Social support self esteem and depression. severe mental disorder in Afro-Caribbean PsychologicalMedicine, 16,813-831. Brown, G.& Prudo, R. (1981)Psychiatric disorder Patients. PsychologicalMedicine, 18,643-657. in urban and rural popula~ons:I. Aetiology of Health of the Nation (1994) Key Area H a n d ~o o k ~ Mental Illness. D o H ~ M S OLondon. depression. Psyc~ologicalMedicine, 11,581-589. Burke, A. (1986) Racism, Prejudice and Mental 111- Health of the Nation(1996) S~ectrumof Care: Local al Problems. Servicesfor People with M e ~ ~ tHealth Psychiatry (ed. J. L. Cox). ness. In: ~ranscu~tural DOH London. Groom Helm,London. Carson, D. (1993)Risk Taking with Mentally Disor- Holloway, F. (1988) Day Care and Community dered O~enders. SLE Publications, Southampton. t y in Practice (eds Support. In: C o ~ ~ u n iCave A. Lavenders & F.Holloway). John Wiley, Cochrane, R. (1977) Mental illnessin immigrants to England and Wales: an analysis of mental hos- Chichester. Hoult, J. & Reynolds, I. (1984) Schizophrenia: a pital admissions.Social Psychiatry,12,25-35. comprehensive trial of community oriented Churchill, D.N. (1985) Compliance-how to measand hospital oriented psychiatric care. Acta ure it. Modern Medicine in Canada,40, 1203-1211. Ill 69,359-372. Psychia~rica Scandin~vica, Craig, T. (1994) TheHomelessMentally
(1991)British ~ournaZ o fp s y c k ja t ~, 158 (suppl, IO), Hunter, D. & Wistow, G. (1987) Mapping the organisational context. 1 Central departments, 69-74. M., Marks, I., Connolly, J. & Audini B. boundaries and responsibilities.In:C o ~ n ~ u n j ~ Muijen, y (1992) Homebased care and standard hospital Care in Britain: ~ariationson a Theme, pp.18-46. care for patients with severe mental illness: The King Edward Hospital Fund, London. a randomised, controlled trial. ~ r ~ tMed~caZ i ~h Johnson, S. & Thornicroft, G. (1995) Service mod~ournal, 304,749-754. els in emergency psychiatry: An international Mumford, D, (1992) Does somatisation explain view. In:EmergencyMental Health Services in the anything? P s y c k i a t ~ in Practice Spring,xx(x), 11C o ~ ~ u n i tCUP, y . Cambridge. Juel-Nielson, N., Bille, M. & Flygenring, J.(1961) 14. of depresOdegaard, 0.(1961) The epidemiology Frequency of depressive states within geosive psychosis. Acta Psychiatrica Scan~inavica, graphically delimited populations groups. Acta 162,33-38. Psychiatrica Scandjnavica,162,69-80. OPCS (1995) The ~ a t i o n aPsychiatric l Morbidity Kingdon, D. (1994) The Care Programme ApSurvey: Household Survey.HMSO, London. proach. Psychiatric ~ulletin, x8 (2), 68-70. Onyett, S., Heppleston, T. & Bushnell, D. (1994) King, M.,Coker, E.,Leavey, G., Hoare, A. & Johnson-Sabine, E. (1994) Incidence of psychotic The Organisation and Operation of C o ~ m u n i t y ~ ~ in k Englan~.The Sainsbury illness in London: comparison of ethnic groups.Metal ~ e a Teams Centre for Mental Health, London. British Me~ical Journal, 309,1115-1119. Paykel, E.,Herman, G. & Prusoff, B. (1970) TreatKrause, I.B. (1989) Sinking heart: a Punjabi ment setting and clincial depression. Archives of communication of distress. Social Science and General Ps yc h ia t ~, 22,11-21. Medicine, 2,563-575. Paykel, E. (1991) Depression in women. British Lee, A.S. & Murray, R.M. (1988) The long term ~ o u ~of aPls y c k ia t ~, 158 (suppl.), 22-29. outcome of the Maudsley depressives. British Paykel, E. & Priest, R. (1992) Recognition and Journal of Psychiatry, 153,741-751. Leff, J. (1988) Psychiatry A r o u n ~the Globe: A management of depression in General Practice. Consensus statement. British Medical ~ournal, tra~sculturalview. Gaskell Series. Royal College 305,1198-1202. of Psychiatrists, London. Lehman, A., Herron,J., Schwartz, R. & Myers, C. Perkins, R. & Rowland, L, (1991) Sex differences (1993) Rehabilitation for adults with severe men-in service usage in long term psychiatric care. tal illness and substance use disorders. A clini- Are women adequatly served.B~itish~ o u ~of a l cal trial.Jou~nalof Nervous andMental Disorders, Psychiatry,158 (suppl.IO), 75-79. Phelan, M., Strathdee,G. & Thornicroft, G. (eds) 181 (2),86-90. Lelliott, P. & Wing, J. (1994) National audit of new (1995) E ~ e r g e n c yMental Health Services in the long-stay psychiatric patients. 2. Impact on serC o ~ m ~ Cambridge n i ~ , University Press, Camvices. ~ r i t i s Journal h of P s y c h i a t ~165,160-169. , bridge. Littlewood, R. & Lipsedge, M.(1988) Psychiatric Raleigh, V., Balusa, L.& Balarajan, R.et al. (1990) illness among British Afro-Caribbeans, British Suicides amongst immigrants from the Indian 296,95o-g51. ~ e d i c a~ournal, l sub-continent. British ~ o u r n of a ~Psychiatry,r56, Lloyd, K. (1993) Depression and anxiety amongst 46-50. Afro-Carribeans general practice attenders in Ritchie, J. et al. (1994) The Report of the Inquiry into Britain, InternationalJournal Societyof Psychiatry, the Care and Treat~entof Christopher Clunis. NE & SE Thames Regional Health Authorities/ 39r1-9. Mental HealthTask Force (1994) Mental Health in DOH, London. London: Priorities for Action. London. Sayce, L., Craig, T. & Boardman, A. (1991) The deMeltzer, D., Hale, A.S. & Malik, S.J. et al.(lggz) velopment of community mental health centres Community care for patients with schizophrein the UK. Social Psychiatry and Psychiatric nia one year after hospital discharge. British E p i ~ e ~ i o26,14-20. ~o~, Medical ~ o u ~303,1023-1026. a ~ , Sair, A., Bhui, K. & Strathdee, G. (1997) CompliMental Health Foundation (1994) Creating Commuance or collaboration: the use of psychoeducanity Care: Report of the Mental Health Foundation tion in psychiatry. A Brief Review, Psychiatric Inquiry into the Carefor People with SevereM ~ t a l ~ u Z ~(in e ~press). ~n Illnesss. Mental Health Foundation, London. Soreff, S. (1983) New directions and added dimenMenil, J., Owens,J., & Wynne, S. et al.(1990) Asian sions in home psychiatric treatment American suicides.British~ournalof P s y c k ~ t156,748-749. ~, ~ o u r nofpsychiatry, a~ x40 (9), 1213-1216. Morris, R., Woods, R., Davies, K. & Morris, L. Strathdee, G. (1990) The delivery of psychiatric
343
care. Journal of the Royal Society of Medicine, 83, practitioners. British Journa~of General Practi222-225. tioners, 43 (377), 507-511. Strathdee, G. &Jenkins, R. (1995) Contracting for Watson, N. 8r Studd, J. (1990) Gonadal hormones primary care mental health services.PurchasIn: and depression in women. In:Current Approaches (eds Thornicroft, G. & ing Mental ~ e u l t Services h Series: Prediction and T r e a ~ e noft Rec~rrentDeStrathdee, G.) HMS0,London. press~on(eds J. Cobb & N. Goeting). Duphar Test, M.& Scott, R. (1990) Theoretical and research Medical Relations, Southampton, bases of community care programmes. In:MenWessley et al. (1991) Schizophrenia and Afrotal ~ e a l t Care h Delivery:Innovations,Impedi~ents Caribbeans: a case control study. B ~ ~ t i~ournaZ sh and Implementation (eds Marks, I. & Scott, R. of Psychiatry,159,795-801. Cambridge University Press, Cambridge. Westermeyer. (1991) Working with an interpreter Test, L. & Stein, M. (1980) Alternative to Hospital in psychiatric assessments. J o u r ~ofl Nervous and Treatment. 1 Conceptual model, treatment proMental Diseases, 178 (12), 745. gram and clinical evaluation. Archives of General Wood, H. (1994) C o ~ ~ u n iMental ty ~ e a l t hTeam Manager, pp. 1-6. Mental Health Foundation, psychiatry, 37, 392-397. Thornicroft, G, (1990) The concept of case manLondon. agement for the long-term mental illness. InterWhite, E. (1994)3rd QuinquennialC o ~ ~ u nPsyi~y national Review of Psychiatry/3, 125-132. chiatric Nursing Survey Manchester University Thornicroft,G. & Breakey,W. (1991) TheCOSTAR Department of Nursing, Manchester. programme: Improving social networks of the Wilson, M. (1994) ent tal ~ e a land t ~ Britain’sBlack long-team mentally ill. British JournaZ of PsychiaCommunities. NHS Management Executive: Mental Health Task Force and Kings Fund fry, 159, q5-249. Tyrer, P.(1984) Psychiatric clinics in general prac- Centre. tice: an extension of communityB care. r i t ~Jours~ Winoker, G., Cardoret, R., Dorzab, J. & Baker, M. nal of P s ~ c h i a t149, ~ , 449-457. (1971) Depressive Disease: a genetic study. Warner, R., Gater,R., Jackson, M.& Goldberg, D. Archives of General Ps y c h ia t ~, 24,135-144. (1993) Effects of a community mental health Wright,A. (1993) Depression: Recognitionand Manservice on the practice and attitudes of general agement in General Practice. RCGP, London.
344
Up to the early 19805, depression in children and adolescents had not been studied since it was considered a very rare occurrence. Most investigators considered that, due to an immature ego and superego development, depression could not occur before the final phases of adolescent development. Over the last 15 years, the picture has changed dramatically and research in epidemiological, clinical, genetic, neurobiological, pharmacological and psychological aspects of childhood depression has blossomed. This chapter starts with a brief overview of background knowledge on the assessment, distribution, risk factors and outcome of affective disorders in young people; it then continues withan up-to-date reviewof treatment techniques used with children and adolescents suffering from an affective disorder.
Depressive manifestations are often found among children c o n j ~ in ction with other forms of psychopathology or they maybe experienced as part of normal development. It is therefore crucial to differentiate, within the affective phenomena, varying levelsof definition and severity. "he first level is that of depression as a syrnptom. High levels of depressed mood are experienced by over a thirdof adolescents surveyedin community studies (Petersenet al., 1993); and feelingsof sadness and misery are also very child re^ and adol~scentssuffering from other forms psyof chopatholog~be they emotional or disruptive disorders (Harrington et al., 1996). In several studies comparingclinical samples to c o ~ u n i t samples, y depressed mood has been found to be the individual psychiatric feature or the most predictiveof referral status(~chenbach,1991). The second levelis that of a depressive syndrome, which is generally identified in multivariate analysis of symptom scales and ~ u e s t i o ~ a i r eThis s , broad construct of negative affect often failsto differentiate specific anxious and depressive dimensions and it is therefore of limited clinical usefulness. The third level of definition isthat
345
Chapter 1-6:~ a n a g e ~ e nint~hildrenand adoles~ents of depressive disorder, a psychiatrically defined cluster of symptoms persisting over a minimal period of time, associated with substantial impairment in functioning, as defined in current classification schemes. The use of strict criteria to diagnose depression among children is, howA few decades ago, unreliable formulations such ever, a recent phenomenon. as 'masked depression' or 'depression equivalent' were used to account for childhood affective disturbances. Attempts to devise diagnostic criteria for by the late I ~ ~ Oseveral S, children ( ~ e i n b e r et g al., 1973) were shortlived and, studies had shown that major depressionbecould recognized in children with the same diagnostic criteriaas for adults. Along with their adult colleagues, child psychiatrists use the DSM-IV and ICD-IO definitions of affective disorders. In order to account for slight differences in the clinical picture amongst younger subjects, it is noted in the DSM-IV definitionof Major Depression that depressed mood in children and adolescents is often substituted with irritable mood.In addition, the duration criterion of depressed mood for the diagnosis of Dysthymic Disorderis set at 1 year in children and adolescents. Otherwise, the criteria and diagnostic subdivisions used in children are the same asthat for adults. § s e § s ~ e nand t co~or~idi~
Syrnptom checklists, questionnaires and diagnostic interviews are widely used in child psychiatry and are recommended assess to depressive disorders with desirable levels of reliability (see Table 16.1). Assessment procedures often rely on several informants (i.e. parents, teachers and children) and data from these different sources need to be combined ina final formulation. Self-report questionnaires provide a convenient means to assess the severity of depressive symptoms and are suitable to monitor changes over time in patients. They only take a few minutes to complete and the scoring procedures and interpretation of the scores are generally straightforward. However, despite relatively good psychometric properties, these questionnaires have in general too low a specificity and sensitivityto be used for diagnostic purposes. The assessmentof affective disorders andof comorbid disordersis therefore best conducted with diagnostic interviews. Irrespective of the particular method used, the diagnostic assessment should involve a detailed interview of both the child and the parents. Thelevel of agreement between children and parents on the reporting of symptoms and on the final diagnosis has been typically variable across studies, with the best agreement being generally found for symptoms involving clear-cut behavioural changes such as weight and appetite loss, psychomotor retardation, suicide attempts and sleep ing difficulties. Rules for combining discrepant parent and child reports are a best estimate being studied, and most professionals currently use
rocedure whereby a symptom is considered as present when any informant reports it. Obviously, the weight given to the assessment of in toms varies according to the clinical experience of the inves vidual characteristicsof the child andhis/her family, an also to the severity and pervasivenessof symptoms, Comorbidi~is the rule rather than the exception inboth clinical and com‘ty samples. Anxiety disorders are the most frequent comorbid dia aration anxiety, phobias .andgenera~ed anxiety are found in about of children and adolescents with depression. Conduct disorders are found in 20-30% of children referred for depression. In older children and adolescents, substance abuse and eating disorders are found with high rates too. The sam results have been found in c o ~ u n i t studies y where the comorbidity between affective disorders and other psychiatric disorders was assessed. In most c o ~ u nstudies i ~ at least50% of children and adolescents diagnosed with an affective disorder have a comorbid disorder, with several studies sho ing rates of 10% or less of ’pure’ depressive disorder uncomplicated by comorbid problems. The patterns and correlatesof comorbidi~in community studies have been similar to those found in clinically referred samples.
pidemiological estimates for the prevalence of major depression and dysthymia in children and adolescents vary enormously as a function ofdifferent sampling frames, case findingt e c ~ i q u e scase , definition and instrumentation ( F o m b o ~ e1997a). , Amongst school-a ed, prepubertal children, prevalence rates have been consistently low inCO munity studies. Thus, in et al., 1970)~the prevalence rate for depression the Isle of Wight study (Rutter was 0.1% among lo-year-olds. Similar figures have derived from other studor x2-month prevalence rate is ies (Fleming & Offord, 1990) and the 6-month not in excessof 1% in this age group. Interestingly, more boys than girls have been found suffering from depression in prepubertal children. preponderanceof girls is consistently found in adolescent samp in studies whichhave focused on mid or late adolescence. By the age of 14 or 15, the adult preponderance of females is found with a typical 2 :1 gender ratio. The prevalence rate for adolescent samples is also much higher with, on average, a period prevalence of y”% ( F o m b o ~ eqg7a). , esides age and gender, the assessment of risk factors in relation to the evelopment of depressionin childhood and adolescence has been somewhat difficu~t,due to either the small numberof cases identi~iedin these community surveys or to the design limitationsof these studies. h particular, difficultiesarosefromthecross-sectional n a ~ r of e these i ~ a t i o n s w h ido c~ notallowforaproper iden~ficationof causalfactoreover, due tothe
34
isorders, it has been difficult to assess to which de ression were specific risk posed to risk factors for other ific risk indicators of general psychiatric morin various child and adolescent samples have, om adult studies in showing a trend toincreased risk for affective disorders, sion for the youths born in the last three reasons for this increase ( F o m b o ~ e1995). ,
hich have systematically examined the natural hisdren and adolescents, the study by point to the strong recurrence risk of depression sode. Theseauthors followed up children aged8of a major depression or day s t h y ~ i c disorder with depressed mood. The average length n was7 months for major depression and nearly 4 ers. These two groups had a high risk of subsefourth of children with major depression having up to 40% at 2 years and 72% at a 5figures were obtainedin a follow-up study of ression with a rate of 45% of readmissions at 2 the first investigation, relapse was associated the ~o-occurrenceof dysthymic disorder, ity did not affect the outcome, Relapse rates in follow-upsam~les of de~ressedchildren and adolescents identifiedin comroadly shown similar findings,a l t h o ~ the ~ h absolute to be lower in these unselected samples. Thus, adolesression were followedup at 1 year by L e w i n s o ~ et t about one-fifth of them had relapsed within a year, er than that for other psychiatric disorders.
ave investigated theconse~uencesof childhood ever, one study provided follow-up data on a group of chilsychia~icservices with a depression and reassessed on
average 18 years later (Harringtonet al., 1990,1991). Depressed children had four times the risk of developing major depression in adulthood as compared to that for non-depressed psychiatric controls. Furthermore, the increased ris seemed to apply only to affective disordersin adult life and not toother psychiatric disorders, suggesting a strong specific continuity between childhood and adult depression which cannot be explained by comorbid disorders. This follow-up study also indicated that the depressed group had increased contact with medical services and increased use of medication in adult life. The social and economical impact of child and adolescent depression have similarly been documentedin other studies.
Strong links exist between childhood depression and suicidal behaviours. Thus, depression was found to be the strongest predictor of suicidal ideas, plans and attempts in a large sample of referred adolescents seen at the Maudsley Hospital (Fombonne, 1997b). Findings from follow-up studiesin adult life of depressed children also concurin showing an increased risk of suicide mortality which extends into adult life. Thus, two studies with the longest periodsof follow-up (Harringtonet al., 1990; Raoet al., 1993) provide a rough estimate in the neighbourhood of 4-5Y0 for the cumulative risk of suicidal deathin early adultlife following an episodeof depression in childhood or early adolescence. The strengthof the association between depression and completed suicide is most obviousin recent psychological autopsy studies which have been conducted in the last decade ( M a r t ~ e et n al., 199%; Brent et al., 1993; Shaffer et al., 1996). Thus, in the recent New York study of 20were compared to 147 matched where 120 suicidal deaths under the age c o ~controls, ~ mood i ~ disorders were one of the strongest risk factors. A of them half medium rateof 60% was found for affective disorders with about corresponding to clear-cut episodes of major depression. Thus, depression in children and adolescents predicts suicidal ideas, attempts and completed suicides both in the short and long term.
Controlled studies of first-degree relatives using the family history method have consistently found higher rates of affective disorders among relatives of depressed children than among relatives of psychiatric or normal controls of (Livingston et al., 1985; Kutcher&r Marton, 1991). When direct assessments relatives wereused, similar findings were obtained. Thus, in a study comparung 48 prepubertal depressed children with anxious and normal controls, Pui Antich et al. (1989a) found higher ratesof depression among relativesof the
depressed group as well as a high frequency of alcoholism. Familial loading for affective disorders varied according to clinical characteristics of the depression in the proband child, with increased loading for depression with endogenous features and decreased loading for depression associated with conduct disorder. Similarly the rate of depression m o n g relatives of depressed children was twice as high than of that relatives of non-depressed psychiatric controls matched for non-affective symptoms in the Maudsley study (Harrington et al., 1993)~suggesting that the increased familial loading for depressive disorders was indeed specific to depression and not accounted for by associated psychiatric conditions. Studies of offsprings of affectively ill parents have also pointed towards links between adult and childhood or adolescent depression. Thus, Weissmanet al. (1987) found much higher lifetime rates of depression and other disorders such as substance abuse among offsprings of depressed parents as compared to those of normal controls; a follow-up of these offsprings showed that overofhalf them had developed a first episodeof depression by the end of adolescence (Weissmanet al., 1992). Other studies have broadly confirmed these results. There is therefore little doubt that depression runs in families of depressed children and adolescents. Both genetic and environme~talmechanisms seem to be implicated in the increased familial loading for affective disorders (Hammen, 1991; RadkeYarrow et al., 1992); twin and adoption studies would be needed to tease apart of the interplay between nature and nurture in the intrafamilial transmission vulnerability to child and adolescent depression.
~euroendoc~ and e sleep studies have been conducted in depressed children and adolescents in order to validate the diagnosis of childhood depression against a biological standard or to demonstrate stable biological dysfunction across thelife span. While validating studies have been largely inconclusive dueto between-studies variation in subjects' age, diagnostic criteria, levels of severity, treatment setting, and confounding factors such as concurrent suicidality or comorbid conditions, promising results have been reached linking biological markers of adult and childhood, and especially adolescent, depression. Hypercorticolaemia is less frequently repeated in childhood than in adult depression (Dah1 et al., 1989; Puig-Antich et al., 1989b; Kutcher et al., 1991; Goodyer et al., 1996). Studies of the dexamethasone suppression test (DST) have also yielded disappointing and inconsistent results. On. average, DST results had a low sensitivity and specificity among samples of children, with slightly better performances in samplesof in-patients and with older, adolescent subjects. Similarly, the cortico-releasing hormone (CRH)challenge
tests have not shown the blunted ACT reaction in children, contrary to what is usually observed in adults (Ryan et al., ~99a).The expla~ationsfor these findings arestill u~certain. (G in-induced hypoglycaemia growth hormone The has been shown to be decrea neurotic controls, with follo et al., 1984a,b). maintained after recovery (Puig-Antich Puig-~ntichand colleagues. (1984a,b) also found an increased nocturnal secretion of GH among depressed children. owever, these findings have not been replicated consistently in other stu S, providing an overall mixed re about the amine hypothesis in the biology of childhood depression, tudies of sleep abnormalities in children have shown reduced rapid eye ) latencies in some samples (Emslie et al., 1990) but not in others (Puig-Antich et al., 198a), although abnormalities were found in the subjects of the latter study when they had recovered. In adolescent§, more electro-encephalographic (EEG) abnormalities have been found (Lahmeyer et al., 1983; Dah1 et al., 1990)~especially in suicidal and in-patient subjects. Overall, the findings suggest that EEG changes in depression vary according to both developmental stage and ~haracteristicsof the depressive episode.
By contrast to adult depression, studies of antidepressant medications have largely failed to demonstrate efficacy over placebo in controlled trials. A double-blind placebo-controlled study of the efficacy of imipra conducted on 53 depressed children by Puig-~ntichet al. (1987). children who completed the protocol (mean age 9.6 years), there was no difference in the response rate ifor ~ p r a m i n and e for placebo(56% vs. 68%). A plasma study was conducted forof30these children which suggested that response was more likely to occur with higher plasma concentrations of imipramine and desipramine, and with lower severityof depressive symptoms (Ryan et al., 1986). It was therefore concluded that a possible explanaof this study to showsi~nificant a difference was due to the tion for the failure use of too low doses and also possibly because the washout period was not prolonged enough. Information on comorbid disorders was not available in this study. ~ o m o r b i dwas i ~ taken into account in another study comparing imipramine vs. placebo in young depressed children ( ughes et al., 1990). Repression occurring in conjunction with conduct disorder had a better response to placebo than to the active medication; by contrast, depression occurring in isolation or with anxiety disorders showed a better response to h investigation of the efficacyof nortriptyline imipramine than to placebo. was performedin a double-blindplacebo-con~olledstudy of 56 12-year-old
35
epressive disorder (GeUler et al., 1992.a). onse was poor in both the active treatment lfference betweenthe two. A follow-u tients (Geller et al., 1992b, 1993) suggested that pro~ongeduse dication over a 6-month eriodwas associated with an increase to mania, this risk being more pronounced in those children ive family history for affective disorders. Similar negative findings were obtained in one other and in two studies of amitrip~line( repubertal children with depression. "he small could explain in themselves the lack of statistice across groups. st of these studies reported trends nt conditions, but a similar hi meta-analysis of 12 randomize ell et al. (1995) to reviewthe evldence for efficacy of tricyclic medications. Usingstringent selection criteria for the studies included in their meta-analysis, these authors showed that the effect size for activedrugs ac S these studies was estimated deviationforchangescorespressive symptomatolo~ a ratio of response to medication vs. placebo was 1.08, both indica~ngnonsignificant benefit over placebo. This of lack demonstrated efficacy raises the issues of possible biological differe S underlying childhood depression as compared to the findings in adults particulari~of these studies is that the placebo response rate has been excessively high, therefore obscurin possible benefit from the antidepressant in the actively treated groups; S also may have confounded the results of some trialsand uld certainly be given inthe future to diagnostic criteria used to select subjects in these trials. The task ahead in current research prorammes is to combine t r e a ~ e nstudies t with biological, clinicaland family history data to identifysubgroups of depressed childrenand adolescentswho are likely tobe responders to activeantidepressant medications. ont trolled studies on the efficacy of selective serotonin re-uptake S) are fewer. Onestudy assessed the efficacy of fluoxetine in -year-old adolescents selected on the of basis high depression scores on milton scale(Sim~on et al., 1990). Therewas a trend for better response in the fluoxetine group which failed to reach statistical si~ificance du ample size and a periodof investigation whichdid not excee er, a more recent study based on a larger sample of 96 children and adolescent out-patientswith major"depression, aged between8 and 18, has been completed which indicates a significant improve men^ in the roup over the placebo condition (Emslieet al. 1995). o controlled study of monoamine oxydasei ~ b i t o r (s ~ ~ O Ihas s )been
~ ~ a ~16: tM e ya n a ~ e ~ e in n tc ~ i l ~ yand e n a~oles~ents conducted and there m only anecdotal case reports suggesting their usefulness in depression not responding to tricyclics. However, they r e are c o ~not ended in children and adolescents, mainly because young subjects were to be shown less compliant with the dietary requirements (Ryan et al., 1988a). The more recent reversible MAOIs have not yet been in used children and adolescents. Other compounds such as maprotiline, mianserin, trazodone, l o f e p r a ~ n e and bupropion have notbeen properly tested. Drugcombina~onsalso have been little studied. Lithium augmentation has provided encouraging results although the beneficial effect seems to be less clearcut than with adult nonresponders (Ryan et al., 1988b; Strober et al., 3~99;2). A01 augmentation has been little explored becauseof the dietary complian problems already mentioned. Properly controlled studies are still required on this issue. Based on current evidence and practice, the prediction that a child will respond to antidepressant medication wouldbe enhanced if the depressive symptoms are severe and associated with biological features, if there is evidence for a familial history of affective illnesses and if there are no cornorbid, especially disruptive, disorders. The use of SSNs is likely to increase following the resultsof the most recent trials and because they are relatively safe in overdose. Child psychiatrists, however, continue to prescribe tricyclic antidepressants since most practitioners have had experience of good response to tricyclic medications in individual cases. More clinical trials on their efficacy are clearly needed and,in light of the methodological problemsof available trials, premature conclusions on their lackof effect should be avoided. It is worth noting that clomipraminehas shown a beneficial effectin childhood obsessive compulsive disorder (OCD), and it is therefore unclear why these medications wouldnot have also a positive effect in depression. Care should be taken with prescription of tricyclic antidepressants. Several sudden deaths have been reported in recent years amongst children or adolescents treated with tricyclics (four with desipramine and one with i ~ p r a m i n e(Riddle ) et al., 1993). These sudden deaths, thoughtbeto cardiacrelated, occurred in most instances with drug levels within the therapeutic is not recommendedas a first choice. Baserange. In this context, desipramine line electrocardiograms (ECGs) are recommended before starting atricyclic >0.20 S; which mightbe contraindicated in some instances (PR Changes in cardiac parameters during treatment (increase in or QRS PR intervals, heart rate over130/min, increased blood pressure)call for a reduction of dosage or discontinua~onof the drug. Other classical side-effects of tricyclics have been described among children including switching to mania, c o ~ o n side-effects and withdrawal syndromes. Dosage should be gradually increased, with typical starting doses of 1 mg/kg/day being very progressively increased up to 4mg/kg/day. Dosage will be increased up to thepoint where clinical benefits become evident, or side-effects are too pronounced or plasma levels
354
are at the upper limit of the therapeutic range. The extent to which plasma levels should be used to monitor a child treated with tricyclics many factors; levelsshould, however, be checkedin cases of of toxicity, or concerns about compliance. Similarly,the freich repeatECGs are needed varies with several parameters; it is, however,r e c o ~ e n d e dto repeatan ECG when a stablehi S been reached, or before in the advent of any untoward clinical with SSRIs is easier to monitor owing to fewer cardiotoxicand anticholinergic effects. The prescriptionand monitoring of SSRIs amongst children follows broadly the same rulesas those foradults. Once a response has been obtainedwith one antidepressant, it advised is to continue medication forat least a 6-month period which corresponds to the duration of an untreated episode. ~edicationsshould subse~uentlybe reduced very progressively. As in adult depression, electroconvulsive therapy (EGT) has been used in children and adolescents with severe unipolar depression or bipolar disorder. There has been little systematic research the on efficacy and long-term effectsof ECT in this agegroup. A review of 32 case vertheless, improvementin 80% of the cases of depresrchardt, 1990). ECT is therefore a useful treatment to consider in severe affective disorders resistant to antidepressants (Cizadlo Wheaton, 1995).Finally, the treatment of bipolar disorders in children an adolescents is similarthat to of adults, with lithium and carbama~epinebein the first-line medica~ons.
Y Children and adolescents with depression have been shown to have negative styles of thinking, low self-esteem and cognitive biases whereby they at~ibute negative eventsto internal, globaland stable self characteristics. It is unclear whether these features are found as a predisposing characteristic before young subjects experience a firstepisode of depression, or if they are merelystate markers of depression. owever, together with impaired social relationships and behavioural changes occurring in a depressive episode, these cognitive features provide natura argets for rational interventions based on cognitivebehavioural therapy (C )principles. Several co~itive-behaviouralprogrammes have now been developed which contain components such as activity scheduling, cognitive restructuring, problem-solving and social skills training in various co~binationsof 10-20 sessions. Treatment manuals have been available for severalof these programmes. The first evaluation studies have been conducted in school
355
The post treatment assessment indicated a clear supremacy relaxation on overall a d ~ u s ~ eand n t measures of depressive conditions were not improved by the i for the specificity of the treatment after t r e a ~ e n tCurrent . studie rolonged treatment andof CO
sychotherapy has been modified for useadoles~ent~ with (P eau et. al., 1991).This time limite~ ~sychotherapy eriod with weekly sessionsof 45min each. The tr into three phases. In the initial phase a review of depr rigoro~slyc ~ ~ d u cand t e ~the problem areas are identi ive problem areas can be the focusof the psycho1 include: grief, role transitions, interpersonal role disputes, interpersonal deficits and a problem areasingle-~arent of f a ~ i l i ehas s been a d d e ~ to account for the higher frequency of sin~le-parentfamilies as a context for modern adolescent development. The middle phase of treatment focuses on two
35
these two treatment modalities which did equally better than a waiting-list control group. Formal family therapy has not been assessed in childhood depression.
similarities in the m ~ a g e m e nof t childhood depression with that of the rest of the life span. Some differences are now outlined with respect to asses§ment, treatment setting and t r e a ~ e nchoice. t
esides the depressive disorder, careful attention should be paid to comorbid isorders which will need be toaddressed separately in most cases. Children and adolescents livein families andin group settings, typically a school. Assessment of intrafamilial relationships are therefore a core focus of the initial assessment. The presence of a mental disorder in one parent, parental discord and dish arm on^ quarrels and violent arguments between family members, scapegoating of the child, emotional, physical and sexual abuse, all areassociated risk factors of childhood psycho pa tho lo^ and ought tobe detected. Family assessment can alsobe important in identifying sources of stress in the family life (i.e. a disabled sibling, a recently diagnosed cancer in the parent, lossof a loved grandparent, a recent move or transition to another school t justify specific measures.It is also usefulin identifying, in the familial group, members who have a positive and supportive attitude towards the child, with the aim to build on these protective relationships du ing treatment and in the future. Physical illness in the child can also be a precipitating factor; among biological events, onset of puberty in girls has been consistently associated with increased risk of depressive symptoms, especially amongst early maturers and when puberty occurs with simultaneous social changes such as transition to secondary school. ~imilarly,school factors are crucial especially as parents are not always aware of their existence. Thus, a child might suffer from a specific learning disorderor from a general learning backwardness that could be further assessed with proper psychometry. Teachers' reports on both academic achievement and behavioural descriptions are also routinely requested.as of part the assessment. Bullying in schools and the quality of peer-relationships can exert substantial psyc~ological pressure on children and adolescents which go may ~oticed by adults. "he assessment must therefore rely on several informants (subjects, parents, school).
3
In the majority of cases, depressed children and adolescents are treated as out-patients. ut-patient management allows to reduce the short- and longterm impact on child and adolescent development of missed learning and social opportunities. In-patient admission is, however, required in subjects with severe suicidal thoughts and cognitions, with biological features which massively interfere with school and family life, or with psychotic symptoms. In-patient treatment may alsobe useful if remission has not been obtained after a properly conducted~sychologicaltreatment, or after non-response to conventional antidepressants. In some instances, admission might also be required to start a biological treatment in circumstances where the familial milieu does not offer the necessary guaranteesin terms of managing safely the drug in the home environment, or monitoring side-effects and suicidal ideas. In-patient admission can otherwise be dictated by the treatment needs of comorbid conditions. This might be so when comorbid substance abuse or eating disorder is a concern. A group of children present with affective disorders associated with conduct symptoms or disorders, often arising from a b a c k g r o ~ dof psychosocial adversity. In the first instance, social services might play a decisive role in providing a life setting for the youngster from which treatment is more likely to be accepted and beneficial.
~ u l t i m o d a l t r e a ~ise nthe t rule in child disorders. Each treatment package will therefore involvea combination of individual treatment andof contextual intervention addressing the family or school issues, whether or not these are seenas playing a direct causalin~uenceon the depression. In most forms of mild or moderate depression, the first line t r e a ~ e n t consists of psychological therapies. CBT appears to be the most validated intervention in terms of its efficacy to reduce depressive symptoms; IPT-A, social skills trainin and other forms of time-limited interventions aiming at symptom reduction, behavioural activation and skills learning are also appropriate. There are currentlyno clearcut guidelines to pick up which of these inte~entionsis most likely to work with which patient. To a large extent, theexisting treatment manuals and programes encompass overlapping components. With a young child, for example, the behavioural activation aspects of CBT will be the most important feature of their treatment whereas a sophisticated mature adolescent might benefit more directly from pme cognitive techniques. The main goal is to tailor the treatment to indithe vidual needsand to the context of the child. Other forms of interventions (i.e. counselling, relaxation) have proved ineffective.
359
Chapter 16: Management in children and adolescents Antidepressants are selected for young subjects with biological features, bipolar depression, or for depression not responding to psychological treatment. Given the lack of evidence in favour of tricyclics and their relative toxicity, SSRIs are a good choice to start medication. Their relative safety in overdose, and their fewer side-effectsare particularly suited to teenagers. SSRIs also constitute an acceptable alternative to the treatment of depressions of at least moderate intensity, when no psychological intervention is readily available. Lack of response to the preceding interventions should lead to a referral to specialist services. With the exception of older adolescents whose treatment follows the same lines as for adults, the effective prescription of other biological treatments are best provided in specialist services. No study has so far been conducted on the efficacy of combined psychological and biological treatment as compared to unimodal treatments in this age group. In practice, most practitioners combine these two approaches when they have chosen a drug treatment, together with other psychosocial measures (parent/family work, school liaison, etc.).It is important to keep in mind that most depressed children and adolescents do recover within 1 year and that relapse rates are high. Even if recovery is under way, there is some value in offering psychological intervention along the principles of CBT to foster the recovery and prevent relapse.
References Achenbach, T.M. (1991) Mnriualfor the Child Behni 4 m r Checklist/4-18 and 1991 profile. Burlington, VT: University of Vermont, Department of Psychiatry. Angold, A,, Costello, E.J., Messer, A., Pickles, A., Winder, F. & Silver, D. (1995) Development of a short questionnaire for use in epidemiological studies of depression in children and adolescents. International Journal of Methods in Ps!/c/iiiitric Resenrch, 5, 237-249. Asarnow, J.R., Goldstein, M.J., Carlson, G.A., Perdue, S., Bates, S. & Keller, J. (1988)Childhoodonset depressive disorders: a follow-up study of rates of rehospitalization and out-of-home placement among child psychiatric inpatients. Joiirizal o f A ~ f e c t i v eDisorders, 15,24j-2j3. Bertagnoli, M.W. & Borchardt, C.M. (1990) A review of ECT for children and adolescents. Journnl of the American Acndemy of Child and Adolescent Psychiatry, 29, 302-307. Birleson, P., Hudson, I., Buchanan, D.G. & Wolff, S. (1987)Clinical evaluation of a self-rating scale for depressive disorder in childhood (Depression Self-Rating Scale). Journnl of C h ~ l dPsychol-
ogy and Psychiatry, 28,4340, Brent, D.A., Perper, J.A., Moritz et al. (1993) Psychiatric risk factors for adolescent suicide: casecontrol study. Journal ofthe Americnn Academy of Child and Adolescent Psychintry, 32, 521-529. Cizadlo, B.C. & Wheaton, A. (1995) Case study: ECT treatment of a young girl with catatonia. Joiirnal of the Aniericaiz Academy of Child and Adolescent Psychintry 34,332-335. [See also letters to the Editor, Journal of the Americnn Academy oj Child and Adolescent Psychtatry, 34, 1256-1257.1 Dahl, R., Puig-Antich, J., Ryanet al. (1989)Cortisol secretion in adolescents with major depressive disorder. Actn Psychiatrica Scaiidinnz~ica,So, 1826. Dahl, R., Puig-Antich, J., Ryan, N.D. et nl. (1990) EEG sleep in adolescents with major depression: the role of suicidality and inpatient status. Journal ofAflective Disorders, 19,63-75. Elander, J. & Rutter, M. (1996) Use and development of the Rutter parents’ and teachers’ scales. lnternntionni Journal of Methods in Psjychintric Research, 6, 63-78. Emslie, G . J . , Rush, A . J . , Weinberg, W.A.,
demy of Child and Adolescent Psychiat~,31,43-50. Rintelmann, J.W. & Roffward, H.P. (1990) ChilGoodman, R. (1994) A modified version of the dren with major depression show reduced rapid Rutter Parent Questionnaire including extra eye movement latencies. Archives of General Psyitems on children’s strengths. ~ournalof Child chiatry, 47,119-1q. psycho lo^ and Psychiat~, 35,1483-1494. Emslie, G.J., Kowatch,R., Costello, L., Travis, G. & Pierce, L. (1995) ~ o u ~ l e - ~ study l i n dof~uoxetine Goodyer, I.U., Herbert, J. & Altham, P.M.E. et al. (1996)Adrenalsecretionduringmajor in depressed children and adolescents. Symposium depression in 8-to 16-year olds. Psychological at the Annual Conference of the American Academy of Child and Adolescent Psychiatry, medic in^, 26,245-256. Hammen, C. (1991) Depression runs in families. October 1995, Toronto. In: The Social Context of Risk and Res~liencein Fine, S., Forth, A., Gilbert,M. & Haley, G. (1991) Group therapy for adolescent depressive disor- Children of Depressed Mothers. Springer-Verlag, der: a comparison of social skills and therapeu- New York. Harrington, R., Fudge, H., Rutter,M,, Pickles, A. tic support. ~ournalof the A~zericanAcademy of & Hill, J. (1990) Adult outcomes of childhood Child and AdolescentP s y c h i a ~30~79-85. ~, and adolescent depression. I. Psychiatric status. Fleming, J.E.& Offord, D.R. (1990) Epidemiology Archives of GeneralPsychiatr~,47,465-473. of childhood depressive disorders: a critical review. ~ournalof the American Academy of Child Harrington, RC., Fudge, H., Rutter, M., Pickles, A. & Hill, J. (1991) Adult outcomes of childhood 29,571-580. and AdolescentPsychiat~, Fombonne, E. (1994) Increased rates of depression:and adolescent depression: 11. Risk for antisoof the American Academy of update of epidemiological findings and analyti- cial disorders.~ournal Child andA~olescent Psychiat~, 30,434-439. cal problems.Acta Psychiatrica Scandinavica,90, Harrington,R.C.,Fudge, 145-156. H., Rutter, M., Fombonne, E. (1995) Depressive disorders: time Bredenkamp, D., Groothues, C. & Pridham, J. trends and possible explanatory mechanisms. In: (1993) Child and adult depression: a test of British Psychosocial iso orders in Young People: Time trends continui~eswith data from family study. uses (eds M. Rutter & D. Smith) pp. 162,627-633. ~ o ~ r nofaPsychiat~, l 544-4515. Wiley& Sons, Chichester. Harrington, R., Rutter, M. & Fombonne, E. (1996) Fombonne, E, (1997a) Theepide~ologyof child Developmental pathways in depression: multiple meanings, antecedents and endpoints.Deand adolescent depression.Psycho~ogicalMedicine (in preparation). velop~ent and psycho pa tho lo^, 8,601-616. Fombonne, E. (1997b) Suicidal behaviours in Hazell, P., O’Connell,D.,Hearthcote,D., young people: time trends and their correlates. Robertson, J. & Henry, D. (1995) Efficacy of (submi~ed). tricyclic drugs in treating child and adolescent Fombonne, E. (1998) Time trends for affective dis- depression: a meta-analysis. British Medical orders. In: Where and en:~ n ~ u e nofc ehistorical ~ o u r n a310, ~ , 897-901. time and place on aspects ~ s yofc h o p a t h o l(eds o ~ P. Hodges, K. (1993) Structured Interviews for Cohen, L. RobinsPS C. Slomkowski) Lawrence Assessing Children. ~ournalof Child psycho lo^ Erlbaum Associates, Mahwah, New Jersey. and P s y c h i a ~34 ~ ,(l), 49-48. Geller, B., Cooper,T.,Graham, D.L., Fetner, H.H., Hughes, C.W.,Preskorn,S.H.,Weller,E., Marsteller, F.A. PS Wells, J.M. (1992a) Hassanein, R. PS Tucker, S. (1990) The effect of Pharmacokinetically designed double-blind concomitant disorders in childhood depression placebo-controlled study of nortriptyline in on predicting treatment response. Psycho6-12-year-olds with major depressive disorder. pharmacolo~Bulle~in,26,235-238. ~ournalof the American Academy of Child and Kahn, J.S.,Kehle, T.J., Jenson, W.R. & Clarke, E. doles scent Psychiatry, 3 ~ ~ 3 4 - 4 4 . (1990)Comparison of cognitive-behavioural, Geller, B., Fox, L.W., Cooper, T.B. & Garrity K. relaxation, and self-modeling interventions for (1992b) Baseline and 2-to 3-year follow-up chardepression among middle-school students. acteristics of placebo-washout responders from School psycho lo^ Review, 19, 196-211. the nortriptyline study of depressed 6-to x2-year Kashani, J.H., Shekin, W.O. & Reid, J.C. (1984) olds. ~ o u ~ nofa lthe American Academy of Child and Amitriptyline in children with major depressive Adolescent Psychiat~,31, 622-628. disorder. A double-blind crossover pilotstudy. Geller, B., Fox,L.W. & Fletcher., M. (1993) Effects ~ournalof the American Academy of C h i l ~and of tricyclic antidepressants on switching to Adolesce~tPsychiatry, 23,348-351. mania andon the onset of bipolarity in depressed Kovacs, M. (1982) The Children’s Depression In6-to 12-year-olds. ~ournalof the Amerjcan Acaventory. ~anuscript.University of Pittsburgh.
Kovacs, M., Feinberg, T.L., Crouse-Novak, M., ~ournalof the American Academy of Child and Paulsuskas, S.L. & Finkelstein, R. (1984a) Adolescent P s y c h i a ~30, ~ , 642-651. Depressive disorders inchildhood. I. A longitu- Mufson, L., Moreau, D., Weissman, M.M. & dinal prospective study of characteristics and Klerman, G.L. (1993) ~nte~ersonal psychot~rap~ , recovery Archives of General Ps y c h i a t ~41,229” for Depressed A~olescents.Guilford Press, New York. 237. Kovacs, M., Feinberg, T.L., Crouse-Novak, M., Mufson, L., Moreau, D., Weissman, M.M., Paulsuskas, S.L., Pollock, M. & Finkelstein, R. Wickramaratne, P., Martin, J. & Samoilov, A., (1984b) Depressive disorders in childhood. II. (1994) Modification of Interpersonal PsychoA longitudina~ study of the risk fora subsequent therapy with depressed adolescents (IPT-A): major depression.Archives of General Ps y c h i a t ~ / Phase I and I1 studies. ~ournalof the American A c a ~ e of ~ yChild and Adolescent P s y c h i a ~33, / 411 643-649* Kramer, A.D. & Feiguine, R.J. (1981) Clinical ef695-705. fects of amitriptyline in adolescent depression: Petersen, A.C., Compas, B., Brooks-Gunn, J., A pilot study. ~ournalof the American Academy of Stemmler, M., Ey, S. & Grant, K.(1993) DepresChild psycho lo^ and P s y c h i a t ~20,636-644. / sion in adolescence.American Psychologist,Feb., 155-168. Kutcher, S. & Marton, P. (1991) Affectivedisorders in first degree relatives of adolescent onset bi- Petti, T.A. & Law, W, (1982) III.Imipramine treatment of depressed children: A double-blind polar, unipolars, and normalcontrols.~ournalof the American Academy of Child Ps y c h i a t ~30~75, pilot study ~ournalof Clinical P s y c h o p ~ ~ c o l o ~ , 2,107-110. 78. Kutcher, S., Malkin, D., Silverberg, J. et al. (1991) Poznanski, E., Grossman, J.A., Buchsbaum, Y., Nocturnal cortisol, thyroid stimulating horBanegas, M., Freeman, L. & Gibbons, R. (1984) Preliminary studies of the reliability and validmone, and growth hormone secretory profiles ~ ity of the children’s depression rating scale. Jourin depressed adolescents. ~ o u of~thea American Academy of Child Ps y c h i a t ~3,0,407-414. nal of the American Academy of Child Ps y c h ia t ~, Lahmeyer, H.W., Poznanski, E.O. & Bellur, S.N. 231 191-197(1983) EEG sleep in depressed adolescents. Puig-Antich, J., Goetz, R. & Hanlon, C. et al. (1982) A ~ e r ~~ournal c a ~ of P s y c h i a t ~140, , 1150-1153. Sleep architecture and REM sleep measures in Lang, M. & Tisher, M. (1987)Children‘s ~ ~ r e s s ~ o nprepubertal children with major depression: a controlled study. Arch~vesof General P s y c h i a ~ ~ , a ~ American edn), Consulting Scale ~ a n u (North PsychologistsPress, Palo Alto, CA. 391 932-939* Lewinsohn, P.M., Clarke, G.N., Hops, H. & An- Puig-Antich, J., Novacenko, M.S. & Davies, M. et al. (1984a) Growth hormone secretion in drews, J. (1990)Co~itive- behaviour^ treatment ~ a p y , prepubertal children with major depression, I. for depressed adolescents.g e ~ v i o u r ~ h21, Final report on response to insulin-induced 385-401. ’ hypoglycemia during a depressive episode. Lewinsohn, P.M., Hops, H., Roberts, R.E., Seeley, Archives of General Ps y c h ia t ~, 41, 455-460. J.R. & Andrews, J.A. (1993)Adolescent psychopathology: I, Prevalence and incidence of de- Puig-Antich, J., Goetz, R. & Davies, M.et al. (1984b) Growth hormone secretion in prepubertal chilpression and otherDSM-III-R disorders in high dren with major depression. 11. Sleep-related a~ lo^/ school students. ~ournalof A ~ n o r m psycho 102,133-144. plasma concentrationsduring a depressiveepis41, 463-466. ode. Archives of General Ps y c h ia t ~/ & Smith, R.G. Livingston, R., Nugent, H., Rader, L. Puig-Antich, J., Novacenko, H., Tabrizi, M.A. (1985) Family histories of depressed and et al. (1984) Growth hormone secretion in severely anxious children. Americun ~ournaZof prepubertal major depressive children. 111:ReP s yc h ia t ~142,1497-14g9. , sponse to insulin induced hypoglycemia in a Marttunen, M.J., Aro, H.M.,Hemiksson, M.M. & drug-free, fully recovered clinicalstate.Archives Lonnqvist, J.K. (1991)Mental disorders in adoof GeneralPs y c h ia t ~,41, 461-475. lescent suicide: DSM-III-R axes I and I1 diagW. e f al. (1987) noses in suicides among 13- to 19-year-olds in Puig-Anti&, J., Perel, J.M., Lupatkin, Imipramine in pubertal major depressive disor/ Finland. Archives of General Ps y c h i a t ~48,83444, 81-89. ders. Archives of General Ps y c h ia t ~, 839. et al. (1989a) Moreau, D., Mufson, L., Weissman, M.M. & Puig-Antich, J., Goetz, D., Davies, M. A controlledfamily history study of prepubertal Klerman, G.L. (1991) Interpersonal psychomajor depressive disorder.Archives of General therapy for adolescent depression: description 46,406-418. Ps y c h ia t ~, of modification and preliminary application.
Archives of G e ~ e r a l P s y c h(in ~ a press). t~ Puig-Antich, J., Dahl, R. & Ryan, N. (1989b) & CopCortisol secretion in prepubertal children with Simeon, J.G., Dinicola, V.E., Ferguson, H.B. ping, W. (1990) Adolescent depression: A plamajor depressive disorders.Archives of General cebo controlled fluoxetine treatment study and Psychiatry,46,801-809. follow-up. Progress in Neuropsychophurmacolo~ Radke-Yarrow, M., Nottelmann, E., Martinez,P., and ~iologicalPsychiatry,14, 791-795. Fox, M.B. & Belmont, B. (1992) Young children Stark, K.D., Reynolds, W.M. & Kaslow, N. (1987) of affectively ill parents: a longitudinal study of psychosocial development.Journal of the AmeriA comparison of the relative efficacy of self-control therapy and a behavioural problem-solving can Academy of Child Psych~atry,31, 68-77, Radloff, L.S. (1977) The CES-D Scale: Aself-report therapy for depression in children. Journal of A ~ n o r ~Child a l psycho lo^, 15,91-113. scale for research in the general population. Stark, K.D. (1990) Childhood Depression:School-~ased Applied PsychologicalMeasure~ent,I,385-401. ~nterven~ion. Guilford Press, New York. Rao, U., Weissman,M.M.,Martin, J.A.& Hammond, R.W. (1993) Childhood depression Strober, M., Freeman, R., Rigali, J.,Schxnidt, S. & and risk of suicide: a preliminary report of a Diamond, R. (1992) The pharmacotherapy of depressive illness in adolescence: 11. Effects of longitudinal study. ~ o u r of ~ athe~American Academy of Child and Adolescent P s y c h i ~ t32, ~, lithium augmentation in nonresponders to I, 21-27. imipramine. J o ~r n aolf the American Academy of Reynolds, W.M. (1986) A model for the screening Child and Adolescent Psychiatry, 31, 16-20. and identification of depressed children and Tisher, M. & Lang, M. (1983) The Children's Deadolescents in school settings. pression Scale: review and further developProfessional School ments. In: A~ectiveDisorders in Childhood and PSycholo~,1,117-129. (eds D.P. Cantwell & G. A. Carlson) Ado~esce~ce Reynolds, W.M. & Coats, K.I. (1986) A comparison ofc o ~ t i v ~ b e h a v i otherapy ~al and relaxa- pp. 81-203. MTP Press, Lancaster. tion training for the treatmentof depression in Vostanis, P. & Harrington, R. (1994) Cognitive beadolescents. Journal of Consulting and Clinical havioural treatment of depressive disorder in child psychiatric patients-rationale and deP s y c h o ~ o54, ~ ,653-660. scription of a treatment package. European Child Riddle, M., Geller,B. & RyanI N. (1993) Another sudden death in a child treated with desipramine. and Adolescent P s y c h ~ a t3,~ 111-123. , Weinberg, W.A., Rutman, J., Sullivan, L., Penick, ~ournalof the Am e r i c a ~Academy of Child and E.C. & Dietz, S.G. (1973) Depressionin children Adolescent P s y c h i a t~32,792-797. , referred to an educational diagnostic center: Rutter, M., Tizard, J. & W1tmore, K. (1970) Educadiagnosis and treatment. Journal of Pediatr~cs, tion, health and ~ehaviour.Longmans, London. Ryan, N.D., Puig-Antich, J.,Cooper, T.et al. (1986) 83,1065-1072. Imipramine in adolescent major depression: Weissmn, M.M., Gammon, D., John, K. et al. (1987) plasma level and clinical response. Acta PsychiChildren of depressed parents: increased psyatrica Scandinavica,93, 275-288. chopathology and early onset of major depresRyan, N.D., Puig-Antich, J., Rabinovich,H. et al. sion. Archives of General Psychiatry, 44, 847(1988a) MAOIs in adolescent major depression 853. unresponsive to tricyclic antidepressants. JourWeissman, M.M., Fendrich, M., Warner, V. & nal of the A ~ e r i c a nAcademy of Child and AdolesWickramaratne, P. (1992) Incidenceof psychiatcent P s y c h i a t~29 , , 755-758. ric disorder in offspring at high and low risk for Ryan, N.D., Meyer, V,, Dachille,S., Mazzie, D. & depression. ~ournalof the A ~ e r i c a nAcademy of Puig-Antich, J. (198813) Lithium antidepressant Child Ps yc h ia t ~, 31, 640-648. augmentation in TCA-refractory depression in Wood, A., Kroll,L., Moore, A.& Harrington, R.C. adolescents. ~ournalof the A~zericanAcademy of (1995) Properties of the mood and feelings quesChild and Adolescent P s y c h i a t ~2,9, 371-376. tionnaire in adolescent psychiatric outpatients: Ryan, N.D., Birmaher, B. & Perel, J. et al. (1992) a research note. Journal of Child psycho lo^ and Neuroendocrine response to L-5 hydroxy36,327-334. P s y c h i a t ~, tryptophan challenge in prepubertal major Wood, A., Harrington, R.C. & Moore, A. (1996) depression. Archives of Genera2 P s y c h ~ a t49, ~, Controlled trial of a brief co~itive-be~avioural 843-85 1. intervention in adolescent patients with depresShaffer, D., Gould, M.S., Fisher, P.et ai. Psychiatric sive disorders. Journal of Child psycho lo^ and diagnosis in child and adolescent suicide. ~ s y C h i a t ~ l 3737-746. 9,
Several case control studies have compared of rates depression inparturient women with those in non-childbearing women and, in contrast to the psychoses, have found no increased re valence of depressions after childbirth, at least of those severe enough to be categorized as according to Research et al., 1988; OIHaraet al., 1990; iff fen Diagnostic Criteria (RDC) cases (Cooper Gotlib, 1993).Cox et al.(1993a),however, noted an increase in ~ n c ~ ~ e(i.e. nce new cases) of depression at around 1 month postpartum.
There is little in terms of clinical presentation and psychiatric history to distinguish postpartum and non-postpartum depressions, apart from the confounding of the normal physical correlates of parturition with depressive tiredness, sleeping difficulties) and differences in sympstnatal depressions are relatively mild and short-lived (Cox et al., 19g3a; Whiffen Gotlib, 1993) but a significant proportion become severe and intransigent, frequently lasting a year or longer (Cox et al., 1982; Watson et al., 1984) and sometimes requiring admission. It is important to differentiate between postpartum depressions, which are usually non-psychotic, and postpartum psychoses. The latter are a heterogeneous groupof illnesses defined largelyby their temporal relationshipto parturition (see Brockington& Cox-Roper, 1988) with different aetiolgy, incidence and timing of onset from non-psychotic postpartum depressions (Kendell et al., 1987; Marks et al., 1992a). There is, however, some overlap: 35% of admissions to the Mother and Baby Unit at the Bethlem Hospital are for depressive illnesses, about half of which have psychotic features (Kumar et al., 1995a). Overall, the riskof maternal suicidein the first postnatal year is low, one sixth of the expected rate (Appleby, 1991). When suicides do occur, the woman is likely to be psychotic (Appleby, 1996).
The most important indicatorof risk is psychiatric history. There is a threefold increasein the rateof depression after childbirth in women with histor& Murray, 1990; Marks ies of depression cornpared to those without (Carothers et al., q92a; Appleby et al., 1994; Cooper& Murray, 1995). Sub-clinical symptoms of depression and anxiety during the pregnancy are also precursors (Pitt, 1968; Watson et al., 1984; Markset al., 1992a). In a meta-analysis of 59 prospective studies which investigated risk
factors for postnatal depression,
ra and Swain (1996) found that past hislargest effect size. They concluded that tory of psychopathology produc there is ac o n t i n u i ~ of psychiatric disturban~ethat extends back
only after childbirth (Cooper for these women issuesCO women may reveal important differences in aetiology, t r e a ~ e nand t outcome.
Caring for a newbornbaby requires enormous reservesof energy as well as c o ~ i t m e nand t resilience. There is also the need to manage anxieties inherent in the developmentof a new relationship. This takesits toll on even the most robustof parents but for those with pre-existing psychological fragilities the experience of early parenthood is particularly fraught. The existing evidence suggests that thevulnerabi1i~-stressmodel of depression applies to artum illness, as with depressions occurri at other times. Adverse life and social difficulties (Paykelet al., 1980 ’Hara et al., 1984; Watson et al., 1984; Marks et al., 1992. ,poor social support and/or marital problems (Paykel et al., 1980; Kumar Robson, 1984; O’Hara et al., 1984; Watson et al., ra & Swain, 1996) andchi1 th p o s ~ a r ~depression. m retrospective studies of depressed women admitted to hospital1 month within p o s ~ a r t u mfound either a marginal effect (Martin et al., 1989) or no effect (Dowlatshahi & Paykel, 1990) of social adversity suggesting that social factors maybe of less importance in the more severe pos artum depressions.
For some women issues c o n c e ~ i n gchildbirth and mothering are more important than social stress perse. Such difficulties may stem from the woman’s erience of parenting andher consequent i~entification with the mothering role (Ruble et al., 1990; Gotlibet al., 1991) or from more recent problems or changes in.her life. Reproductive difficulties suchas infertility probunplanned pregnancy, scarriage lems, or stillbirth obstetric complications and ( Hara, 1986; Edwards
3
et al., 1994; Yoshidaet al., 1997) haveall been linked to subsequent postpar~m depression. Then there is the inevitable emotional upheaval associated with the transition to parenthood, for example, the change in status from that of employment in the workforce with its daily social contact and financial rewards to the unremarked isolation and strapped financial circumstances of m a t e r n i ~(Mc~ntosh,1993). The woman’s relationship withher partner also inevitably cause changes Lovestone, can this and1995).
Disturbances in the woman’s relations~p with her infant have also been linke to maternal depressions, althoughit is difficult to disentangle cause andefso is likely fect in these cases. Clearly a mother who is depressed or becoming to have an impact on her infant’s mood and behaviour, and also is likely to perceive her infant more negatively. However, thereis evidence to suggest h addithat the infant’s temperament can influence maternal mood directly. tion, the mother’s perception of an infant may lead to subsequent maternal depression. For example,poor motor scores and highirritabilit~assessed in the neonate, are predictors of the onset of maternal depression by 8 weeks postpartum (Murray et al., 1996) as is mother’s perceived satisfaction with infant feeding (§andersonet al., 1996). Not all mothers feel immediately after delivery that they love their babies and sometimes feelings of affection and a t t a c ~ e nemerge t slowly over days or weeks. However, many mothers who are depressed also express conce~s about their relationship with their infant. These may include of feelings inadequacy and guilt for not coping well as a mother, excessive worries about the infant’s physical health, fears that they may harm the infant in some way or persisting feelings of either indifference ordislike for their baby (Margison, 1982; Kumar, 1997).
There is an increasing body of evidence that maternal depression may sometimes have long-term adverse impact on the psychological developmentof the new child(Cogill et al., 1986; Murray, 1992; Sharp et al., 1995) especially if the depression occurs during the infant’s first year (Sharp et al., 1995). These effects are thoughtto be related to the way depressed mothersinteract with their infants. Depressed mothers have been shownbe to distracted and preoccupied, expressionless and ~ s t i m u l a ~org irritable , and intrusive. The form, content and emotional tone of their speech with the infants is also
different. They focus on their infants less and they tend to be more critical urray et al., 1993). There can be disturbances in the qualityof mothers’ reactions to cues from their infants, for example their synchrony and contingency with them, and these disturbances can result in patterns of behaviour in the infant which mirror the mother (Field et al., 1988). Such disturbances may lead to impairmentsin the infants’ memory and learning of insecure attachprocesses (SingerPS: Fagan, 1992) and to the development ment patterns (Ainsworthet al., 1978). There may alsobe more disturbing consequences,A New Zealand study itchell et al., 1992) found that mothersof infants who had cot deaths were three times more likely to have been depressed at 1 month postpartum than other mothers.In this study, depression had been assessed prospectively. The E~inburghPostnatal Depression Scale had been administered by health visitors, at 4 weeks postpartum, toall mothers in a numberof districts who had delivered during the year of the study. Forty-seven per cent of mothers whose infants subsequently had cot deaths scored as cases of depression at this time, compared to16% of control mothers.Got death mothers were also more likely to have been treated for depression in the past and also more likely to have a tal than control mothers. A more recent family history of ~ ~ s t ~ adepression et al., 1996). British studyhas replicated the New Zealand findings (Sanderson Sociodemographic and medical data were collected from all mothers who delivered in Sheffield during 1988-93, at birth and at 1 month pos~artum, including scores on the Edinburgh Postnatal Depression Scale (EPDS). There were subsequently 42 cases of unexpected infant death. The relative riskof sudden infant death was found to be almost four times greater among women with high scores on the EPDS compared to the low-scoring group. Explanationsof these effects are highly speculative at present but are like to invoke at the very least certain characteristic features of the depressed mother-infant relationship. It may be, for example, that depressed mothers are less attuned to their infants’ ’state of health and well-being, andless able to respond appropriately to any changes in the infant itchel ell et al., 1992).
There is no increase in the prevalenceof depression in fathers following childbirth (Kendellet al., 1976; Ballardet al., 1994)~although men may have more subclinical depressive symptoms during the first few weeks after their wives have had ababy (Ballard PS: Davies, 1996). Fathers, however, influence and are influenced by their partners’ depresrown et al. (199oa, 199ob) have demonstrated the importanceof the quality of close personal relationships, particularly the marital relations~p, in women with children. in the occurrenceof, and recovery from, depressions
orwomensufferin artum depression,both the onset and course are influenced bythe woman%relationship with her partner ceptions of poor marital support or marital problems have enhanced rates of postpartum pression in community sa Robson, 1984; r at so^ et al., 19 Boyce et al., 1991)and th ship has also been implicated inthe more severe depressions. For example, women without partners are more likely to be admitted to psychiatric hospital after deliv endell et al., 1987). e omen with histories of severe ly to have a reoccurrence of illness after childers who are uncommunicative ( S who express appreciationof their to protect womenat risk frompostpartum relapse (Markset al., 1996). Men whose wives experience~sychiatricillness afterc ~ l d b i r t h have enhanced rates of depression ( McGrath, 1988; Lovestone & Kumar, 1993). Lovestoneand Kumar showedthat partners of women admitted to a psychiatric mother and baby unit were more likely to be depressed than either partners of women admitted outside the postpartum period or control partners of postpartum women who are well. The men whose wives had been admitted were also more likely to have had a depression in the past than the other men. These data suggest there may be couples who are at risk of maternal depression. In these couples,if the mother is depressedthen the father is also likely to have some mood disturbance. Treatment strategies involving both parents, for example marital or family therapy may beof benefit. At present there has been nostudy of the efficacy of this kindof help.
Treatments are those used for depressions generally and have been covered in detail in other chapters. owever, an important but difficult aspect inthe management of maternal depression is that decisions about t r e a ~ e nhave t to take into account both the mother’s and her infant’s well-being. These include judgements about immediate risk of harm to the infant and also long-term ones concerning the mother’s ability as a parent which have to be balanced with the problems which might result from over-intrusive care and which may impair the mother’s natural and healthy relationship with her infant. Another difficulty concernsthe reluctance of mothers who are depressed to recognize that this is so, and then to feel it is safe to seek help for their condition. Many mothers feel they are not coping, that they have failed as mothers, and sometimes fear that should their mental state become public they could lose careof their child. These concerns, as wellas more general
fears about 'going mad' and being sti~mati~ed, mean that women are often sychiatric services. Offersof care therefore ptability of the form of care delivery to the S of t r e a ~ e n are t also influenced by the fact that p e r i p a r ~ m ve close contact with obstetric, health visitor and primary care serS that detection of cases and delivery of care areincreasin~ly out at a primary care level, and that there is more psychiatric ese groups in the provision of care to pos~atallyde
t probostnatal ~epressionsare mildand self- limit^^ and t r e a ~ e ncan limited tosupportive counselling (Holdenet al., 1989). However, about one-third (Oates,19 9) may require more active interven~on,either phamachological. A very small proportion (I in a 1000 of all may require in-patient admission. Thus, as for depresside the postnatal period, most mothers can be cared for in the communitywith regular General ~ractitioner(GP) and health visitor contact. A. variety of services have been developed for those mothers whose lnesses are more severe. These include psychiatric out-patient contact, ita1 care,and in-patient motherand baby units. ie
rturient women are in repeated contact with obstetric services, and one way of targeting womenwho are depressed orat risk of becoming so is via these services. For example, at Kin 'S College Hospital, London, there is an c-psychiatricliaisonservicewhichvidesapsychiatricserviceto erinatal women identified at antenatal bo g as having histories of mender. Under this scheme, patients with histories, or current, significant ic illness (screened by midwivesat antenatal booking) are offered tment with a psychia~istand are then monitored by the psychigular intervals during the p r e ~ a n and c ~ postpartum. However, any omen are reluctantto accept this kind of psychiatric contactand takee low-about one-quarter of women referred do not attend their initial ~sychiatristappointment and a further quarter of those who do, failto attend later a p p o i n ~ e n t (Appleby s et al., 1989).
ss
As long as the motheris not breastfeeding, indications for the use of anti
treatment of depressions occurring pressant medication in the ssions occurring at other ti y which has compared the relative effic
atment must involv itably, therefore, ther psychotropic medi wledge about the to the infant t h ~ o ~ g breastmilk h and the possible impact on the infant is limited. on antidepress~t medica~on and evidence currently available concerns the tricyclic antidepressants (TGAs) although there are few a case studies of selective ser re-uptake i ~ i ~ i t o r s .Essentially the data suggest that these and their principal lites :plasma ratios are generally a ~ o ~ n pass freely into breast milk and that milk
unity. Reports of drug or drug metabolites detected in infant have been rare (TCAs, 3 instances; SSRIs, 1 instance) as have reports o adverse effects in infants (doxepin, 1 instance; ~uoxetine, 1 instance) (see et al., 1990; Yoshida& Kumar, 1996; for reviews). Thus, the dose ingestedby infants appears tobe minimal and reportsof harm rare. However, the total numbers of studies and subjects studied are very small (n = 44 for tricyclic compounds,n = 5 for SS S) m d o ~ s e ~ a t i o n s of infants have not always been systematic. In light of the present state of owlege, there seems to be no reason to prevent a mother who is takin tricyclic antidepressants from breastfeeding if she wishesto, with ble exception of doxepin. Less is known about the §SRI group of dr they are particularly indicated and provided that the infants are caref~lly monitored, breastfeeding should notbe prohibited. b re-term, i m a t u r e infants should not be exposed to psychotropic drugs, and the baby should always be monitored for any adverse effects and healthy before allowinga medicated mother to breastfeed.
There has long been the idea that postpartum depression maybe linked to physiological changes occurring during and immediately after pregnancy, This view, largely anecdotal and until recently untested, has lead to various proponents of sex steroids as either prophylaxis or treatment for postnatal depression. There arevery large changes in steroids over the course of parturition. For example, cortisol, prolactin and testosterone ~rogesteroneand oestrogen reach plasma levels of 100-200 times p r e p r e ~ m c y levels and then drop steeply immediately after delivery, returning to p r e ~ r e g n ~values cy within days. Changes in the levelsof some of these hormones have been linked to alterations in mood, although the evidence is sc and sometimes conflicting. For example, there is evidence that physiolo & Gel doses of oestrogen may improve mood (Sherwin al., 1991) and that doses of progesterone may depress Sherwin, 1991); that reductions levels in of either oestr 1996) or progesterone(see Harris, 1996) may depress mood. The mechanism for effects such as these is thought to be related to an impact of steroid hormones on neuronal neurotrans~ittersystems. cise description of their action will not be covered here except to say that interactions with various neurotransmitter systems have been implicated. These include dopaminergic and serotonergic function. (Interested readers are referred to Wieck, 1996, for a summary of this research.) There is some evidence that increasedsensitivity of the dopamine system maypre~ispose
women at risk of depression becauseof their historiesof psychiatric illness, to a depressive relapse after childbirth(Wieck et al., 1991; McIvor et al., and Wieck et al. (1991)have suggestedthat rapid falls in oestrogen at pa tion may over-sensitizethe dopamine systems of women who are alreadyat risk. Further evidence for a with link oestrogen comes from its successful use as an antidepressant in postnatally depressed women. In a placebo-controlled trial, Gregoire et at. (1996)a d m ~ s t e r e d oest uffering frompostnatal de ression and found that p showed significantly hanced mood withi treatment. Thisstud needs to be replicatedand the mechani oestrogen to be eluc Evidence for the efficacy of other hormonal treatments, forexa esterone (Dalton, 1980) atispresent mostly anec properly cont~olledstudies have yet to be ~ndertaken.
ealth visitors have become increasingly involved in the dete tal depressionand in caring for women identified as depress I 6), either in support groups (Pitts, 1995; Romaine, 1995) or individually. den et al, (1989)demonstrated that weekly counselling sessions by trained health visitors is a successful treatment for postnatal depression. WO tified as having a postnatal depression were allocatedto a con up. Women in the treatment group received eight consec 3 months, 69% ofwomen in the health visitors. After group had recovered compared to 38% of the controls popular with patients: 92% of the women identified so, 91% completed the trial.
Ina randomi~edcontrolled trial, Cooper and urray (1997) assess benefits of different psychotherapies. An unselected sample of dep women, with infants aged between 8 and 18weeks, were randomly allocated to either non-directive counselling, cognitive behaviour therapy, psychodynamic psychotheraortoacontrolcon primary care. Sessions re carried out at home and lasted aboutan hou mean numberof sessions was eight (range, fiveten). to Mothers and infants were followed up at intervals after treatmentended. All therapy cond were associated both with improved maternal mood and with mothers
373
reporting improvements in their relationshipswith their infants. There were no differences between treatments. While treatment had no direct impact on any of the infant measures taken, early remission in the motherwas associated with security of attachment in the infant. These data highlight the importance of early detection and treatment of maternal depressions, not only for the mother but also her infant. They confirm that the more quicklya mother's depression is curtailed, the less likely it will have lasting negative effects on her infant.
Appleby et al. (19977) assessed the benefits of combining antidepressant medication with brief counselling for the treatment of women depressed at 6-8 weeks after delivery. Ina randomized controlled treatment trial, depressed women received either fluoxetine or placebo with either one or six sessions of cognitive behavioural counselling. Fluoxetine and counselling each produced improvements in maternal mood but there was no additional a~vantagefrom receiving both forms of treatment together. The authors suggest that the choice of treatment therefore could be left up to individual women.
Many women find it easier to seek help from non-professionalssuch as can be obtained from self-help support groups developed and run by their peers. In Britainsuch organizations exist locally (for example, Newpin in London) and at a national level (for example, National~ h i l d b i rTrust; t~ Association for Postnatal Mental Illness). Help provided includes befriending, wherea member, usuallya woman who has herself sufferedpostnatal depression, is introduced to the depressed women and becomes available to her for help and support as required.
Where possible, prevention is preferable to treatment, p r and o ~ a ~aimed e s at doing so have been developed for women identified during pregnancy as being at risk.
In an open clinical trial, Wisner and Wheeler (1994) tested the efficacy of antidepressant medication administered during the postpartum period to prevent a recurrence of postpartum depression in womenwho had at least one
374
~ h a ~ t e17r:~ e ~ r e s s iafter o ~ child~irth previous postpartum episode that met DSM-111-R (American Psychiatric Association, 1987) criteria for non-bipolar major depression without psychotic features. Twenty-three women took part in the study,15 ofwhom chose medication. Patients were given the antidepressant medication that had been effective for theirprevious episode, or nortriptyline,within 24h after delivery. They were then monitored for 3 months and their psychiatric outcome during this follow-up period was compared with the eight women who elected to be monitored without medication. Only7% of women who received medication had a recurrence of depression (DSM-111-R criteria) whereas 63% of untreated women becamedepressed. While this study needs replicationin a randomized, placebo-controlled trial, the findings suggest that for women who havehad a previous, severepostpartum depression, judiciouslyadministered antidepressant medication may be of prophylactic benefit.
These have been shown to be effective but unpopular. Elliott et al. (1988)have shown that monthly group meetings with a health visitor or psychologist during pregnancy halvedrates of subsequentpostnatal depression. However, many women were reluctant to participate: 35%of women identifiedas being at risk did not attend any of the meetings offered. a In sixnilar study only 31% of mothers attended the support groups and there was no effect on rates of subsequent postnatal depression (Stampet al., 1995).
Another response to the difficulty in providing accessible care has been the development of specialist midwifery services for high-risk women (Kumar et al., 1g95b). The key featureof such services is continuity of midwifery care rather than specialized training, for example, counselling training, as a main aim is to deploy existing midwifery resources to good effect. Continuity of care is thought to have a preventative role by providing social support to mothers, enhancing the likelihood of detection of prodromal symptoms and improving take-up rates if referrals are made to other professionals Briscoe, (cf. 1986). Thereis evidence, too,that continuity of midwifery care may have direct beneficial effects. It has beenshown to reducepostnatal 'blues' (Odent, 1984)~ labour times and obstetric complicationsin both mother and baby (Sosa ;Klaus et al,, 1986; Kennel1et al., 1991)and it isthought that it may ibute to enhanced maternal postnatal mood. Indications are that this form of support is very popular with high-risk women.The impact on syc~opathologyafter delivery is currently being assessed.
ipolar affective illness in the postnatal period raises issues which are different from bipolar illness elsewhere in the life cycle and unipolar illness following childbirth. First, the month after delivery is theof the timehighest risk for manic illness (Kendell et al., 1987) and for women with stories of bipolar affective disorder, the risk of a further episode of illness after subsequent deliveries is40--50% (Reich & ~ i n o k u r1970; , Schopfet al., 1984; Markset al., 1992a). The risk of recurrence is further influenced by the time which has elapsed since thelast episode of illness and women who have been admitted are particularlyat risk (Markset al., xgg;?a). in the 2 years precedingd e l i v e ~ For women with historiesof bipolar disorder, the riskof a purely depressive episode p o s t p a r ~ mis similar to that of women with histories of unipolar disorder (Marks et al., 1991). However, many puerperal manic episodes are followed by depressions which are sometimes very severe.
8o-g0% of manic episodes begin within ; ?weeks It is estimated that r roc king ton & Cox-Roper, 1988;~ l o m p e ~ o u&~van e r Hulst, 1991; Kumar et al.,1995a). Thereafter the rate of relapse fallsexponentiall~None theless, it remains elevated until at least 2 years p o s ~ a r t u m (Kendell et al.,1987). m onsets which are more lar illnesses occurringin the ~ u e ~ e r i uhave therwise the phenomenology appears to besimilar to those o c c u r r ~ goutside the postnatal period (Schopf & Rust, 1994). OS
ile the mechanisms responsible for the ~ u l n e r a ~ of i lbipolar i~ patients to the p u e ~ e r i u mare not yet known,it is thought that the physiologicalconcomitants of pregnancy and parturition are probably implicated. One hypothesis is that there isan enhanced sensitivityof dopamine functionto oestrogen and progesterone withdrawal. Neuroendocrine evidence in supportof this hypothesis has been published (Wieck et al. 1991, 1997). However, the critical testof the hypothesis willbe whether or not puerperal relapse can be reduced by lowering the rateof fall of brain concentrationsof oestrogen and esterone, and this is currently being investi~ated. There is some evidence to suggest that psychosocial factors, for example, life events, are implicatedin bipolar depressive episodes after childbirth,as with unipolar puerperal depression, but that they are less important in manic episodes (Markset al., 1991). The qualityof the marital relationship, however, may have an impact on manic brea~downarks et al., 1992a,b).
ei There are very few studies of the impactof early exposure to maternal bipolar illness on the infant and the findings are inconsistent. One study found that children whose mothers were currently depressed but had had a previous manic illness were more likelyto have insecure attachment patterns and to have more disturbed peer relationships and more behavioural disturbance than childrenof mothers who had suffered unipolar depressions (De & Radke-Yarrow, 1991). This contrasts with other studies which suggest that children of mothers with bipolar disorder may be at decreased risk compared to those of unipolar mothers, in terms of social competence and disturbed behaviour (Gonnors et al., 1979; Winters et al., 1981)~security of attachment and persistencein play tasks (Hipwellet al., 1997)~and that there is no longterm impact on the infant’s cognitive and social development (Davies et al., 1996). It may be that in contrast to mothers who are depressed, when a mother becomes manic sheis more obviously not safe with her infant, and that the care of the infant is therefore more likelybetosupervised and taken over by others when necessary; thereby protecting the child from the full impact of maternalillness.owever,thehandful of studiescompletedto date have consisted of very small and unrepresentative samples. Future research, using methodology designed to test more precise hypotheses about bipolar mothers’ relationships with their infants, may yet demonstrate a specific impact of maternal bipolar illness on the infant.
In general, women who become manic after delivery will require psychiatric admission and usually mother and infant are admitted together, either to a general psychiatric ward or to a specialized psychiatric mother baby unit. and Initially all contact between mother and baby needs to be supervised by nursing staff, andas the mother’s condition improves contact with her baby to be increased accordingly. Physical treatment is dictated by the clinical picture, as with bipolar illnesses occurring at other times, and neuroleptics, antidepressants, mood stabilizers and electroconvulsive therapy are all used.
The useof lithium as a prophylaxis has been tested inopen clinical trials two which showed some reduction in relapse rates (Stewartet al., 1991; Austin, 1992). Transdermal oestrogen administered after delivery and continued for et al., in the first two postpartum weeks hasno had impact on relapse (Kumar prep.). Forboth these treatments, breastfeeding is contraindicated and therefore unwelcome to women who are currently well and wish to breastfeed.
~
Most postnatal depressions are mild and short lived, andbe can cared forin the community with minimal intervention. It is important, however, that carers are alert to the possibility that sometimes the condition can become severe and intransigent, and thatit may have a long-term, possibly fatal impact on the infant. At present, treatment strategies are similar to those used fordepressions occurring at other times although different modes of delivery of care, particularly those involving health visitors and midwives are currently evolving. Research into the efficacy of different typesof pharmacological treatment for this condition has yetbeto done as have investigationsinto women whose depressions occur only after childbirth.
Ainsworth, M.D.S., Blehar, MC. &Wall, S. (1978) Bebbington, P.E., Dean, C., Der, G., Hurry, J. & P a t t e ~ofs Attackment. Erlbaum, HiIlsdale,NJ. Tennant, C. (1991) Gender, parity, and the prevaAmerican Psychiatric Association (1987) Diagnoslence of rninor affectivedisorder.gri~isk Jo~rnal tic and Statistical Manual of Mental Disorders, reof Psyckiatry, 158,40-45. vised 3rde&. American Psychiatric Association, Bowlby, J.(1980) Attackment and Loss, Vol. 3: Loss. Washington, D C . Basic Books,New York. Appleby, L. (1991)Suicide during pregnancy and Boyce, P., Hickie, I. & Parker, G. (1991) Parents, in thefirst postnatal year. gritisk ~ e d i c a l J o u ~ a l ,partners or personality?Risk factors for postnatal depression. Journal of Afective Disorders, 3O2,137-140. Appleby, L. (1996) Suicide behaviour in child2% 245-255. bearing women. ~nternationalReview of Psycki- Briscoe, M. (1986) Identificationof emotional probUtry, 8,107-115. lems in postpartum womenby health visitors. Appleby, L., Fox, H., Shaw, & M.Kmar, R. (1989) Britisk Medical ~ o u ~292,1245-1247. ul, The psychiatrist in the obstetric unit. Establish- Brockington, I. & Cox-Roper, A. (1988) The British Journal of Psyckiatry, ing a liaison service. nosology of puerperal mental illness. In: M o ~ k ~ k oand o d Mental Illness, Vol. 2: Causes and 154, 510-515. Appleby, L., Gregoire, A., Platz, C. & Kumar, R. con^^^ (eds R. K ~ m aand r I. F. Brockington), (1994) Screening women for high risk of postpp. 1-16. Wright, London. & Andrews, natal depression. ~ournulof Psyckoso~aticRe- Brown, G.W., Bifulco, A., Veiel, H.O.F. B. (199oa) Self esteem and depression.U[:Social search, 38,539-545. Appleby, L., Warner, R., Whitton, A. & Faragher, correlates of self esteem. Social Psyckiatry and B. (1997) A controlled study of fluoxetine and Psyckiatric E p i d e m i o ~ o25,225-234. ~, cognitive-behavioural counselling in the treat- Brown, G.W., Bifulco, A. & Andrews, B. (q9ob) Self-esteem and depression. 111: Aetiological ment of postnatal depression. g ~ i t i s kMedical Journal, 3x4, 932-936. issues. Social Psychiatry and Psychiatric E p i d ~ i Austin, M.P. (1992)Puerperal affective psychosis: OlOgY, 25,235-243. is there a case for lithium prophylaxis?British Buist, A., Norman, T.R. & Dennerstein, L. (1990) Journal of Psyckiatry 161,692-694. Breastfeeding and the use of psychotropic mediBallard, C. & Davies, R. (1996) Postnatal depres~ournal o~Afective Disorders,19, cation: a review. ~l of Psyck~try, 197-206. sion in fathers. I n t e ~ a t i oRmiew 8,65-71. Carothers, A.D. & Murray, L. (1990) Estimating Ballard, C.G., Davies, R., Cullen, R.N., Mohan, psychiatric morbidity by logistic regression. R.N. & Dean, C.(1994) Prevalence of postnatal Psyckological Medicine, 20,695-702. psychiatric morbidity in mothers and fathers. Cogill, S.R., Caplan, H.L., Alexandra, H., Robson, British Journulof Psychiatry,164,782-788. K.M. & Kumar, R. (1986) Impact of maternal
postnatal depression on cognitive development Apilot study of postnatal depression following ~ 292, in young children. Brit~shM e d ~ c aJournal, caesarian section using two retrospective self1165-1167. rating instruments. Journal of Psychosomatic Research, 38,111-117. Comers, C., Himmelhock,J., Goyette, C., Ulrich, R. & Neil, J. (1979) Children of parents with Elliott, S.A., Sanjak,M. & Leverton, T. (1988) Paraffective disorders.Journal of Academy of Child ent groups in pregnancy: a preventative interPsychology,18,600-607. vention for postnatal depression? In: MarshalCooper, P.J., Campbell, E.,Day, A., Kennerley, H. ling Social Support: Formats, Processes and Efects & Bond, A.(1988) Non-psychotic disorder after (ed. B. H. Gottlieb), pp. 87-110. Sage, Beverly childbirth: a prospective study of prevalence, Hills, CA. incidence, course and nature.British ~ournalof Field, T.M., Healy, B., Goldstein, S. et al. (1988) Psychiatry,152, 799-4306. Infants of depressed mothers show ‘depressed’ Cooper, P.J. & Murray, L.(1995) Course and recur- behaviour even with nondepressed adults. rence of postnatal depression: evidence for Child ~evelopment, 59,1569-1579. the specificity of the diagnostic concept. British Gater, R.A., Dean, C.& Morris, J.(1989) The conJournal of Psyc~iatry,166,191-195. tribution of childbearing to the sex difference Cooper, P.J.& Murray, L.(1997) The impact of psy- in first admission rates for affective psychosis. chological treatments of postnatal depression onPsychological Medicine, ig,71g-724. maternal mood and infant development. In: Gotlib, I.H., Whiffen, V.E., Wallace, P.M.& Mount, Postpartum ~ ~ r e s s iand o nChild ~ ~ e l ~ m(eds e n t J.H. (1991) Prospective investigation of postL. Murray& P.J. Cooper), pp.201-220. Guilford, partum depression: factors involved in onset New York. and recovery. Journal of Abnormal Psychology, Cox, J.L., Connor, Y. & Kendell, R.E. (1982) Proloo, 122-132. spective study of the psychiatric disorders of Gregoire, A. J.P., Kumar, R., Everitt, B., Henderson, childbirth. British Journal of Ps y c h i a t ~140, , 111A.F. & Studd, J.W.W. (1996) Transdemal oestro117. gen for severe postnatal depression. Lancet, 347, Cox, J.L., Murray, D. & Chapman, G. (q93a) A 93-33, controlled study of the onset, duration and Harris, B. (1996) Hormonal aspects of postnatal prevalence of postnatal depression.British Jourdepression. lnternational Review of Psychiatry, 8, nal ofPsychiatry, 163, 27-31. 27-36. Cox, J.L., Gerrard, J., Cookson, D. & Jones, J.M. Harvey, I. & McGrath, G. (1988) Psychiatric mor(199313) Development and audit of Charles bidity in spouses of women admitted to a Street Parent and Baby Day Unit, Stoke-onmother and baby unit.British Journal of PsychiTrent. Psychiatric Bulletin, ~7,711-713. atry, 152,506-510. Cutrona, C.E. (1984) Social support and stress in Hipwell, A.E., Goosens, EA., Melhuish, E.C. & the transition to parenthood.Journal of AbnorKumar, R.(1997) Severe maternal psychopathomal Psyc~ology,92,161-172. logy, joint hospitalisation and infant-mother Dalton, K. (1980) ~epresssionafter Chil~birth.Oxattachment. ~evelopmentand Psychopathology. ford University Press, Oxford. (In press.) Davies, L., Hipwell, A. & Kumar, R. (1994) Holden, J.(1996) The role of health visitors in postSevere postnatal mental illness and the psynatal depression.ln~ernationalReview of Psychichological development of the 4 year old: a atry, 8,79436. controlled follow-up study. lnfant Be~aviour and Holden, J.M., Sagovski, & R.Cox, J.L. (1989)Coun~evelopment,17,594. selling in a general practice setting: controlled De Mulder, E.& Radke-Yarrow, M. (1991) Attachstudy of health visitor intervention in treatment ment with affectively ill and well mothers: con- of postnatal depression.British medica^ Journal, current behavioural correlates. ~evelopmentand 298,223-226. Psychopathology,3,227-242. Kendell, R.E., Wainwright, S., Hailey, A. & Ditkoff, E.C.,Crary, W.G.,Cristo, M. & Lobo, R.A. Shannon, B. (1976) The influence of childbirth (1991) Oestrogen improves psychological func- on psychiatric morbidity.Psycholo~cal Medic~ne, tion in asymptomatic post-menopausal women. 6,297-302. Obstetricsand Gynuecolo~, 78,991”yyj. Kendell, R.C., Chalmers, J.C. & Platz, C. (1987) Dowlatshahi, D.& Paykel, E.S. (1990) Life events Epidemiology of puerperal psychoses. B r ~ ~ i s ~ and social stress in puerperal psychoses: absence Journal of Psychiatry, 150,662-673. of effect.PsychologicalMedicine, 20,655-662. Kennell, J., Klaus, M. & McGrath, S. et al. (1991) Edwards, D.R.L., Porter, S.A.M. & Stein, G.S.(1994) Continuous emotional support during labour in
l
379 \
a US hospital: a randomised controlled trial.
R. (1991) Lifestress and postpartum psychosis: a preliminary report. British Journal of Psychi2197-2201. atry, 158 (Suppl. 10),45-49. S.S. & Sosa, Marks, M.N., Wieck, A., Checkley, S.A. & Kumar, Klaus, M.H., Kennell, J.H., Roberson, R. (1986) Effects of social support during R. (1992a) Contribution of psychological and parturition on maternal and infantmorbidity. social factors to psychotic and non-psychotic British Medical Journa~,6,585-587. relapse after childbirth in women withhistories Klompenhouwer, J.L. & van Hulst, A.M. (1991) ~ournulof A~ectiveDisorders, of affective disorder. Classification of postpartum psychosis: a study 29,253-264. of250 mother and baby admissions in the Marks, M.N., Wieck, A., Seymour, A., Checkley, Netherlands. Acta Psyckiatrica Scandinavica, 84, S.A. & Kumar, R. (1992b) Womenwhose mental 255-261. illnesses recur after childbirth and partners Kumar, R. (1997) Anybody‘s child: 44 mothers with levels of expressed emotion during late pregpost-partum onsetof disturbances of maternal 163,211-216. nancy. British Journal of Ps y c h ia t ~, affection and of the mother-infant relationship. Marks, M.N., Wieck, A., Checkley, S.A. & Kmar, British Journal of P s y c ~ i ~(Int ~press.) . R. (1996) How does marriage protect women Kumar, R,& Robson, K. (1984)A prospectivestudy with histories of affective disorder from of emotional disorders in childbearing women. postpartum relapse? British Journal of Medical British Journfflof P s y c h i a ~144,35-47. ~, P s y c h o ~ o69,32cp-342. ~, Kumar, R., Marks, M.N., Platz, C.& Yoshida, K. Martin, C.J., Brown, G.W., Goldberg, D.P. & (1995a) Clinical survey of a psychiatric mother Brockington, I.F. (1989) Psychosocial stress and and baby unit characteristics of 100consecutive puerperal depression. ~ o u r n aof~A~ectiveDisorders, 16,283-293. admissions. J o u ~ aofl A~ectiveDisorders, 33,1122. Mitchell, E.A., Thompson, J.M.D.,Stewart, A.W. et al. (1992) Postnatal depressionand SIDS a proKumar, R., Marks, M.N.&Jackson,K. (zgggb)Prespective study.J o u r ~aofl Paediatric Child ~ e a l ~ h , vention and treatmentof postnatal psychiatric 28 (Suppl. I), S13-16. disorders: therole of the midwife. Brit~shJournal of M i d ~ ~3, e314-317. ~ , Murray, L. (1992)The impact of postnatal depresl Child sion on infant development. J o ~ r n a of Kumar, R., Marks, M.N., Campbell, I.C. & Checkley, S.A. Transdermal oestrogen and psycho lo^ and Psychiatry,33, 543-561. Murray, L., Kempton, C., Woolgar, M. & Hooper, postpartum biploarrelapse. (Inpreparation.) R. (1993) Depressed mothers’ speech to their Lovestone, S. & Kumar, R. (1993) Postnatal psyinfants and its relation to infant gender and chiatric illness: the impact on spouses. British cognitive development.Journal of psycho lo^ and Journal of P s y c h i u ~68,157-168. ~, Ps y c h ia t ~, 34,1083-1102. McIntosh, J. (1993) Postpartum depression: woR., King, F. & Fiorimen’s help-seekingbehaviour and perceptions Murray, L., Stanley, C., Hooper, Cowly, A. (1996) The role of infant factors , of cause. J o u ~ aofl Advanced ~ u r s i n g18,178-184. in postnatal depression and mother infant McIvor, R.J.,Davies, A., Wieck, A.et al.(1996) The al and Child interactions. ~ e v e l o p ~ e n tMedicine growth hormone response to apomorphine at ~ e u r o l 38,~09-119. o~ 4 days postpartum in women with histories of developmentof an integrated major depression. Journal of A~ectiveDisorders, Oates, M. (1988) The community orientedservice forsevere postna40,131-136. tal mental illness. In: Motherhood und Mental Magos, A.L., Brewster, E. & Sing, H. et al. (1986) The effectsof norethisterone in post-menopau- Illness (eds R. Kumar & I. F. Brockington), pp. 133-158. Wright, London. sal women on oestrogen replacement therapy: a model for the pre-menstrual syndrome.Brit- Oates, M. (1989)Management of major mental illness in pregnancy and the puerperium. In: ish Journal of Obstetrics and G y n a e c o l o ~93, , Psyc~ologicalAspects of Obstetrics and Gynaeco1290-1296. Clinical logy (ed M Oates), pp. 905”po. Balli~re’s Margison, F. (1982) Thepathology of the motherObstetrics and Gynaecolo~,3. infant relationship. In: Motherhood and Men~al ~ l l n e s s(eds I.F. Brockington & R. Kumar), Odent, M. (1984) Birth Reborn. Souvenir Press, London. pp. 191-222. Wright, London. support, life events and Marks, M.N. & Lovestone, S. (1995) The role of OHara, M.W. (1986) Social the father in parental postnatal mental health. depression during pregnancy and the puerper~ ,569-573. ium. Archives of General P s y c k i f f ~43, Britisk Journfflof Medical psycho lo^, 68,157-168. & Kumar, OHara, M.W. & Swain,A.M. (1996)Rates and risk Marks, M.N., Wieck, A., Checkley, S.A. Journal of the A ~ e r i c a nMedical Association, 265,
Allen, H. & Kumar, R.(1995) The impact of postof postpartum depression-a meta-analysis. natal depression on boys’ intellectual developlnternatio~al Review ofPsychiaty, 8,37-54. ment. ~ournalof Child psycho lo^^ and Psychiatry, OHara, M.W. & Zekoski, E. (1988) Post-partum 36,1315-1336. depression: a comprehensive review.In:MotherSherwin, B.B. (1991) The impact of different doses hood and Mental Illness (eds R. Kumar & I.F. of oestrogen and progestin on mood and sexual Brockington), pp-17-63. Wright, London. behaviour in post-menopausal women.~ o ~ r n a ~ O’Hara, M.W., Neunaber, D.J. & Zekoski, E.M. (1984) Prospective study of postpartum depres- of Clinical EndocrinoZogy and Me t a bo lis ~,7 sion: prevalence, course and predictive factors. 334-343. Sherwin, B.B.& Gelfand, M.M. (1985) Sex steroids ~ournaZof A ~ n o r ~psycho ai lo^, 93, 158-171. and affect in the surgical menopause: a doubleO’Hara, M.W., Zekoski, E.M., Phillips, L.H. & Psychone~roendocrinolo~, Wright, E.J. (1990) Controlled prospective study blind cross-over study. of postpartum mood disorders: comparison of 10, 325-335, Singer, J.M. & Fagan, J.W. (1992) Negative affect, childbearing and non-childbearing women. emotional expression and forgetting in young ~ournalof A ~ n o r ~ a l P s y c99,3-15. hol~, infants. ~ e v e l o p PsychoZo~, ~ ~ ~ a ~28,48-57. Paffenbarger, R.S., Jr (1964) Epidemiological aspects of postpartum mental illness. British Sosa, R., Kennell, J.H., Klaus, M.H.& Urruttia, J. (1980) The effect of a supportive companion on ~ournalof Preventative and Social ~ e d i c ~ n18, e, perinatal problems, length of labour and 189-195. mother-infant interaction, New England ~ournaZ Paykel, E.S., E m s , E.M., Fletcher, J. & Rassaby, of Medicine, 303, 597-600. E.S. (1980) Life events and social support in puerperal depression.~ r i t ~~ournal sh of Ps y c h i a t ~ , Spitzer, R.L., Endicott, J. & Robbins, E. (1978) Research diagnostic criteria: rationale and reli136, 339-346. ability.Archives of General Psychiaty, 35, 773-782. Pitt, B. (1968) Atypical depression following childStamp, G.E., Williams, A S . & Crowther, C.A. birth. brit is^ ~ournalof P s y c h i a t ~114,1325-1335. (1995) Evaluation of antenatal and postnatal Pitts F. (1995) Comrades in adversity: the group support to overcome postnatal depression: a approach. Health Visitor,68,q4-145. randomized, controlled trial. Birth,22 (3), 138-143. Pugh, T.F., Jerath, B.K., Schmidt, W.M. & Reed, R.B. Stewart, D.E., Klompenhouwer, J.L., Kendell, R.E. (1963) Rates of mental disease relating to & van Hulst, A.M. (1991) Prophylactic lithium childbearing. New England ~ournalof Medicine, in puerperal psychosis: the experience of three 268,1224-1228. centres. British ~ o u ~of a Psychiat l y 158,393-397. Reich, T. & Winokur, G. (1970) Postpartum psyWatson,JP., Elliott, S.A., Rugg, A. J. & Brough, D.I. choses in patients with manic depressive (1984) Psychiatric disorder in pregnancy and the disease.~ournal of Nervo~s Mental Disorders,151, first postnatalyear. British ~ournalof Psychiaty , 60-68. Romaine, S., Jones, A. & Watts, T. (1995) Postnatal 1441 453-462. Wieck, A. (1989) Endocrine aspects of postnatal depression: facilitating peer group support. mental disorders. In: Psychologica~Aspects of ~ e a l t Visitor, h 68,153. Obstetrics and GynaecoZo~(ed. M. Oates), pp. Ruble, D.N., Brooks-Gunn, J., Fleming, A.S., 857-877. BalliereTpdall, London. Fitzmaurice, G., Stangor, &C.Deutsch, F. (1990) Transitions to motherhood and the self: meas- Wieck, A. (1996) Ovarian hormones, mood and neurotransmitters. ~nternationalReview of Psyurement, stability and change.~ o u ~of a Personal l chiaty, 8,17-25. and Social psycho lo^, 58,450-463. Sanderson,CA., Cowden, B.& Hall, D.M.B. (1996) Wieck, A., Kumar, R., Hirst, A.D., Marks, M.N., Campbell, I.C.& Checkley,S.A. (1991) Increased Is ~ostnatal d~ression a riskfactor for’cot death’? sensitivity to dopaminergic stimulation prePaper presented to the biennial meeting of the cedes recurrence of affective psychosis after Marce Society, September 4-7, 1996. Schopf, J.& Rust, B. (1994) Follow-up and family childbirth. British Medical ~ournal,303,613-616. study of postpartum psychoses. Part I: Over- Wieck, A., Davies, R.A. & Kumar, R. et al. (1997) Increased D2 hypothalamic receptor sensitivity: view. E u r o ~ e Archives a~ Psychiatry Clinical a trait marker for puerperal manic-depressive Neuroscience,244,101-111. illness? (Submitted for publication.) Schopf,J., Bryois, C., Jonquiere, & M. Le, P.K. (1984) Whiffen, E.V. & Gotlib, I.A. (1993) Comparison of On the nosology of severe psychiatric postpostpartum and nonpostpartum depression: partum disorders. €uro~eanArchives of Psychiclinical presentation, psychiatric history and atry, and NeurologicalSciences, 234,54-63. psychosocial functioning.~ o ~ r nofa Consulting l Sharp, D., Hay, D., Pawlby, S., Schmucker, G.,
and Clinicai P s ~ c h o i o61,485-494. ~, Yoshida, K. & Kumar, R. (1996) Breastfeedingand Winters, K.C., Stone, A.A., Weintraub, S. & Neale, psychotropic drugs.lnternational Review of PsyJ.M. (1981) Cognitiveand attentional deficits in chiatry, 8,117-124. children vulnerable to psychopa~olog~ ~ournal Yoshida, K., Marks, M.N., Kibe, N., Kumar, R., of A ~ n o ~Child a l P s ~ c k o ~p,o435-453. ~, Nakahano, H. & Tashiro, N.(1997) Postnatal Wisner, K.L. & Wheeler, S.B. (1994) Prevention depression inJapanese women who have given of recurrent postpartum major depression. birth in England. ~ournalof A~ective ~isorders, ~ o s p i ~and a l C o ~ ~ uPsny ci h~ iat~ 45, 1191-1196. 43169-77.
L U C Y C. A ITK E N A N D ROBERT C. B ~ L ~ W I N
The objectof this chapteris to review those aspects of the treatmentof depression in old age which are different from the treatment of depression in younger people. The main differences concern diagnosis, which be more can uncertain in the elderly, the assessment of memory loss, the presenceof physical illness and social isolation, and the implications of these both for physical and psychosocial management.
ie
S
There are factors which may modify the expression of depressive illness in elderly people. Some are due to generational differencesin the perceptionof psychological and physical health, others occur because of the frequent overlap of depressive symptoms and physical illness. Sometimes these factors may accentuate certain aspects of the clinical picture whilst others may obscure the diagnosis. Factors which may alter the presentation of depression in older people are summarizedin Table 18.1. Table 18.1 Factors which modify the presentation of depression in old age. Overlap of medical and psychiatric symptoms Minimal expressionof sadness Somatization or disproportionate complaints associated with physical disorder Neurotic symptomsof recent onset Deliberate self-harm (especially medically 'trivial') 'Pseudodementia' Depression superimposedupon dementia Conduct disorder Accentuation of abnormal personality traits Late onset alcohol dependency syndrome
Associated physical ill-health may lead to an overlap of symptoms. DSMIV criteria (American Psychiatric Association, 1994) requirethat symptoms due to associated physical illness should be discounted when diagnosing
383
practice this is not always easy An older indiv rways disease may experience insomnia, appetite equally from the chest disease or from an associated depression.If aid to the following points, difficulties can often be resolved. First, the history may indicate on enquiry closer that some symptoms are new: for example, early morning wakening may emerge from a back generali%edsleep disturbance due to dyspnoea. Second, for those with limited mobility or exercise tolerance, questions about e n e r ~ yand activity riate: 'Do you feel tired even when resting?', not patients who dismiss depressionas purely 'un A hisbecause of ill-health can easily wrong-foot an inexperienced cli~cian, tory not only from thepatient but also froman forma ant is thus essential, It may demons~ate,for example, a recent decline i nterestsat a timeof static physical health. Cohen-Cole and §toudemire (1 ) provide a useful overview of the problems of diagnosing depression in the presence of major physical illness. Essentiallythe clinician can opt to be restrictive, that is to allow only thosesymptoms which are unequivocally due to epression rather than physical ill-health; or over-inclusive,in which case S er to depression or physical ill-health are counted as depressive in would agree with Cohen-Cole and §tou~emirethat whereas strict criteria are necessary for research, clinicians may wish to take the less restrictiveoption in order to avoid overlooking depression in complex cases. Another difficulty arises because of the well-reco elderly people to minimize feelings of sadness (Georgotas et al., 1983)~ presumably reflecting a cohort of people brought up not to bother their doctors with emotional difficulties. It is essential to askabout two areas: ~ ~ ~ e ~ O ~ the inability to experience pleasure, m d ~ ~ ~ et ~5 O 5~ g~especially ~ ~~ 5e , reduced self-estIt,worthlessness and suicidalideation.Even if dep deniedtized,as it oftenisinolderpeople,some of these S will be present if the patient has a depressive illness. An obscure pain may occasionally have its origin inundia~noseddepression whichan accurate history will usually uncover. Similar to this is Post's observation that occasional cases of severe tinnitus are relieved when an underlying depression is treated (Post, 1982). ore c o ~ m o nthou presentation of an unreassurable elderly person with physical CO ate to known pathology. Again the key is a good history, espe recent changes in pain threshold, plus a careful physical review. A serious error isthat of taking at face value neurotic sym~toms of recent onset. The sudden emergence of severe anxiety, obsessional compulsive
3
phenomena, hysteria or hypochondriasis an inelderly person not previously prone to such disorders should leada to careful search for depressive illness, which is the usual cause (Baldwin, 1988a). Likewise, at all levels of severity of depression, anxiety isa common accompanying symptom(Post, 1972). If it dominates the clinical picture the unwary may miss the underlyin epression. noth her trap is to dismiss an act o eliberate se~f-harmin an older person because the medical conse~~ences ear to be trivial. There is no re of deliberate self-harm and any ascorrelation b e ~ e e the n medical se sychiatric illness; nor isit necessarily the most severely depress arraclough, 1971). Elderly patients who kill themsel ’manipulative’ overdoses. have depressive illness (Pos 1987) as ’a number of condimentia presents for attention sponsible’. In the settin o as ’dementiaof depression’ (Pearlsonet ntia’ and is thought tobe more common n several different ways. when an elderly person, usually clearly reening test for dementia. Since demenbasis of single test results, this usage is lication of the term at dramatic clinical an older person. Post (1982) has sumare always aware of memory imet accurately, unlike the onset of dementia which is much more insidious. The patients often complain vociferouslyabouttheirmeory;andoccasionallyothertivedifficulties,again dementia most ike sufferers. e ~ n c t i olead often n to ’don’t know’ responses, sometimesacco~paniedby marked irritation,in contrast to those with dementia who try their best but are inaccurate. Patients with pseudodementia convey a great dealof despair, often non-verbally, during the interview. The history is often short and sometimes a previous personal or family history of depression is uncovered. Nursing observation may reveal insomnia, diurnal mood change and the chance comment to s~~gest depressive ideation, for example,’ h e nowhere to live’; ’I’m ruined’. These patients are often perplexed and inaccessible. Post argues that depressive can be tested, rather pse~dodementiashould onlybe diagnosed in patients who than those who appear confused but are inaccessible and hence not really testable.
Third, the term 'pseudodementia' is sometimes applied to a clearly depressed patient who is found to have subtle but measurable cognitive impairment. Sometimes this may be global impairment although more usually it is specific, for example an amnesia. Reynolds et al. (1989) have reviewed the literature and found that cognitive impairment was reported in lo-zoo/o of depressed elderly patients. However, deficits in higher cortical function (aphasia, agraphia, etc.) are rarely present.Yet if 'pseudodementia' were to be dein fined purely on the basis of any degree of cognitive dysfunction occurring a depressed patient, then most elderly depressives could be deemed to be cases of 'pseudodementia', since most have subtle deficits (Abas et al., 1990). The term then becomes redundant. Post (1982) argues that depressive 'pseudodementia' is aterm with limited clinical utility: for severe depressive symptoms patient in a require treatment, whether or not it is thought that dementiais present. Perhaps then the term serves its purpose best by acting as a reminder that sometimes apparently demented patients may be suffering from severe depression. 'Depression on dementia', in contrast, refers to the superimpositionof depression on an established dementing disorder. Elderly people sometimes present their depression as a disorder of behaviour. Pitt (1982) points to certain patterns. These include presentations with food refusal,'incontinence' (for example, a perverse ability to eliminate in almost any place other than thetoilet, unlike the incontinenceof dementia), screaming and outwardly aggressive behaviour. Often the location is a a resentful dependency. residential or nursing home facility and the context A good historyis crucial. This is best given either by the resident's key nurse or a care assistant who knows the resident well. A family memberand/or close friend can amplify this and give details of any previous psychiatric history, inforrnation whichis often not known to those caring for the individual. Besides the usual neurovegetative symptoms of depression, a history of a relatively recentchange in behaviour canbe very illuminating.This may take the form of social withdrawal and loss of interest inself and the environment, as well as the more obviously disruptive behaviours described by Pitt. Depressive illness may also lead to an accentuation of prernorbid personality traits. The adventof depressive illness may lead to dramatic theatricality and ceaseless importuning to several agencies simultaneously-for example, social services and primary care. The key is an accurate appreciation of change in the light of known premorbid personali~traits. Also within the areaof behaviour disturbanceit is worth recalling that sometimes a conduct disorder, such as shoplifting, may be a behavioural correlate of depressive illness in an older person, as is an alcohol dependence syndrome acquired forthe first time. Finally, a complaint of loneliness from an individ-
T ~ 18.2 ~ Causes l ~ of organic depressions. Occult c a r c i n o ~ a
Lmg Pancreas ~e~aboZic/endocrine Hypothyroidism Hypercalcaemia Cushing’s disease B,, deficiency
~nfections Post-viral infections Myalgic encephalomyelitis Brucellosis Neurosyphilis Organic brain disease Space-occupying lesion Dementia
Drugs Steroids Beta-blockers Methyldopa Clonidine Nifedipine Digoxin L-dopa Tetrabenazine
ual who has hitherto coped quite well alone, usually accompanied by a request to be re-housed, should lead to a search for depressive illness.
The laboratory investigation of depression is particularly important in the elderly in view of the associated medical causes of depression. These are listed in Table 18.2. ~vestigationshould include haemoglobin and red blood cell indices which may point to possible B12deficiency or alcohol excess. B12estimation should be undertaken in a first episode. Where depression has been present for many weeks folate levelsshould be checked, as subclinical nutritional deficits may have developed.For similar reasons ureaand electrolyte estimations are important. A.low serum potassium will lead to electroconvulsive therapy (EGT) being delayed and an elevated calcium is occasionally associatedwith depression, as in primary h~erparathyroidism(Peterson, 1968)and metastatic cancer. One principle toguide investigation isan awareness that elderly people have less physiological reserve. Severe depression in a 75-year-old may lead to quite serious metabolic derangement which would be unlikely in a fit 40-year-old. Thyroid function testing should be performed because of the well-known association of depression with h~othyroidism,which may be overlooked in the elderly, and because ‘apathetic h~erthyroidism’can be mistaken for depression. Table 18.3 summarizes investigations appropriate
387
.3 Investigations for depression inlater life.
Investigation
First episode
Recurrence
Full blood count Urea and electrolytes Calcium Thyroid function
Yes Yes Yes Yes Yes Yes Yes Yes If clinically indicated If clinically indicated
Yes Yes If indicated If indicated, ormore than 12 months previously If indicated, ormore than 2 years If indicated by nutritional state If indicated (e.g. alcoholabuse) Only if not already done Only if neurologically indicated Only if neurologically indicated
B12
Folate Liver function Syphilitic serology CT (brain) EEG
for a first episode of depression and a rec~rrence. alogram (EEG), there are dia no standard 12-lead EEG. The ma entiate depression from an organic brain syndrome. proach tod i a ~ o s iiss the useof screenin ting i n s t ~ ~ ein~depression ts are t of its cut-off for f these was designeds~ecificallyfor U ltiple choice format whichso son the ~eriatric ~epre§sion Scale t al., 1983). It focuses almost ent aspects of depressive illness, rather than physical dep and has a simple'yes/no' format. Even so, the full version has30 ~ ~ e s ~ o n s .4) and the original vali ation was carried out o depressives which cont ned ~dividual§as you cut-of€ scoreof 11 or above indicates prob ive satisfactory sensitivity ith concurrent medical A shorter version is available a score of 5 and above indicate Whether the GDSis a use€ul m
sociated with~ l ~ h e i m'sedisease r reasonably well( ~ ' ~ o r d aetnal., 2 tia. New scales-for example, the ulos et al., 1988)-have b inco~orates i~ormation from a carer.
~ ~ s ~ ~ Choose ~ c ~ ithe o best ~ s answer : for how you have felt overthe past week. 1 Are you basically satisfied with your life? 2 Have you dropped many of your activities and interests? 3 Do you feel your life is empty? 4 Do you often get bored? 5 Are you hopeful about the future? 6 Are you bothered by thoughts you can’t getout of your head? 7 Are you in good spirits most of the time? 8 Are you afraid something bad is going to happen to you? 9 Do you feel happy most of the time? 10 Do you often feel helpless? 11 Do you often get restless and fidgety? 12 Do you prefer to stay athome, rather thangoing out and doingnew things? 13 Do you frequently worry aboutthe future? 14 Do you feel you have more problems with your memory than most? 15 Do you think it iswonderful tobe alive now? 16 Do you often feel down-hearted andblue (sad)? 17 Do you feel pretty worthless the way you are? 18 Do you worry alot about the past? 19 Do you find life very exciting? 20 Is it hardfor you to starton new projects (plans)? 21 Do you feel full of energy? 22 Do you feel that your situation i s hopeless? 23 Do you think most people are better off (in their lives) than you are? 24 Do you frequently get upset over little things? 25 Do you frequently feel like crying? 26 Do you have trouble concen~ating? 27 Do you enjoy getting up in the m o r n ~ g ? 28 Do you prefer to avoid social gatherings (get-togethers)? 29 Is it easy for you to make decisions? 30 Is your mind as clear as itused to be?
Yes Yes Yes NO
Yes Yes Yes Yes Yes
Yes Yes Yes
Yes Yes Yes Yes o Yes
‘Yes’ or ‘no’ responses are those which score ’l’. Phrases in parentheses refer to alternative ways of expressing the questions to a U po~ulation, as used by the authors.
0
diagnostic criteria outlined earlier for major depression is not necessarily a reason to withholdantidepressant treatment. Patientswith fewer symptoms than are specified within diagnostic criteria such as ICD-10 (World Health nization, 1992) and DSM-IV may nevertheless benefit from treatment. -10partly recognizes this by introducing categories such as 'mild' depressive episode. Factors to consider are: subjective distress, degree of social and functional impairment, and extent of loss of interests. If one is unsure about whether to prescribe an antidepressant, and there often is more uncertainty when assessing older people, it is often helpful to the seepatient two or even three times over a few and weeks to reassess. proportion A of depressive s ~ d r o m e resolve s spontaneously.h alternative to this 'wait and see' approach is to conduct a therapeuticof trial an antidepressant.Involving a community psychiatricnurse (CPN) to monitorsymptoms and response can be helpful. The patient's moodshould be reassessed3-4 weeks after being on a therapeutic dose of an antidepressant. At that oint, if there has been. Some response, the antidepressant should be continued. If the patient remains unchanged, then a re-evaluation will be required,with special attention to dosage of the antidepressant, compliance and possible ~aintaining psychosocial factors. For example, an elderly lady who lived alone developed low mood, insomnia, pervasive anxiety and poor concentration following a burglary. Her sleep pattern improved after a trial of an antidepress~t, but her mood remained low. However, with support from a CPN, she moved to sheltered housingand this resultedin rapid improvement in her mood.
Building apartnership with the patient is therefore crucial. This begins with an e ~ ~ thatZ depression ~ ~ which ~ ~warrants ~ otreatment ~ with tablets is an illness, and that it is common, treatable and not a sign of moral weakness, Many patients need ~ e ~ s sthat ~ the ~ ~tablets ~ c ethey will be asked to take are not addictive, and that depression is neither 'senility' nor a harbinger of dee ~ thet r e a ~ e n tThey . need to be told, in lay mentia. Theyshould be i ~ v o Z vin terms, why they should not expect immediate results. They and the clinician should agree on a plan of treatment. For example, a patient should be informed that antidepressantswill relieve their depressive symptoms, but that further help to manage anxiety address or a berea ment, for example, is also indicated. Nowadays, almost all psychiatrists specializing in of theolder care people work as part of a multidisciplinary team.A 'key worker' from this team should be appointed who can coordinate the care delivered and act as a point of contact forthe patient and family.
~ h a ~ t18: e r~ a n a g e ~ine the ~ tel~erly
This requiresan understanding of the age-associated changesin the distribution, metabolismand excretion of therapeutic drugs. The followingpoints are important. First, in elderly people levels of circulating serum albumin are lower, reducing the capacity for protein binding and resulting in elevated unbound drug concentrations. The ratesof both hepatic metabolism of the drugs and renal excretionof the metabolites will be lower than among younger patients, Any resultant accumulation can lead to exaggerated side-effects. Second, as a resultof normal physiological changes, the pharmacokinetics of tricyclic antidepressants differ in the elderly from those in youngeradults. Consequently, considerably higher steady-state plasma levels may develop in older patients. Third, an increase in inter-individual variabilityfound is in the elderly, making prediction of the therapeutic dose of tricyclics particularly difficult. Some require dosages one-third orofhalf those recommended for younger adults, others the full dose. Time can thus be lost 'titrating' tricyclics in this way. Fourth, some of the selective serotonin re-uptake inhibitors (SSNs) (fluoxetine, fluvoxamineand paroxetine) have proved to be inhibitors of hepatic enzymic oxidation (cytochromeP450 2D6-debrisoquine hydrochloride (Creweet al., 1992)). These drugs can alter the pharmacokinetics of other, hepatically oxidized drugs, leading drugtointeractions, Drugs metabolized by this enzyme,and hence likely to be affected by these new antidepressants are: tricyclic antidepressants and trazodone; neuroleptics; lipophilic beta-blockers (e.g. metoprolol); some antiarrhythmics, and triazolobenzodiazepines, suchas alprazolam. Lastly, because depressionand physical illness are closely linked in oldand age, because the burden of physical morbidity is invariably the greatest in older people, there are considerable risks of adverse interactions with drugs for medical disorders. Particular problems may arisewith coadministration of diuretics, antiarrhythmics, calciurnchannel blockers, some ulcer-healing drugs, hypnotics and other sedative drugs and, of course, alcohol, the abuse of which is frequently unsuspected in an older person. Response times to recovery are longer than younger patients (Schneideret al., 1994). Whilst mostpatients respond within 6 weeks, a sizeable minority may takeup to 9 weeks (Georgotas & McCue, 1989).
This rarely remits with an antidepressant alone. As with other age groups, either a combination of an antidepressant plus an antipsychotic will be required, or electroconvulsivetherapy (ECT) (Baldwin,198813).
393
of class~specific contraindica~ons to, and side-effectsof, s is covered elsewhere an tension is a common and dangerous proble nts and those with reduced left ventricular o e this, a case-control study inA cant increase in hip fracture in S. This is clearly an avoi ments against the us e. ~ e l i r i u m is also a problem witht occurs most often in patients who are medically ill; a offender. ~ a r d i o t o x i cisi ~a risk in the face of a cardiac rhythm disturbance or a clinically relevant conduction disorder. These effects are much more likely lasma levels are allowed to rise iotoxicity of tricyclicdruhasbeenoverstat ever, is und~rstatedand, at a time tients with major depression shoul
why it is a popular choice in the effects are worse with tertiary amine tricyclics (amitriptyline and S,
in contrast, have hardly any action on cholin ittle effect on alphal adrenoreceptors. 15%);diarrhoea ( ~ r o ~ ation (;?--lg%), headache and wei ausea is the most fre nger and older patients. nsion also limits the use d the traditional monoamine For moclobemide, a reve r of monoamine effect profile seems relativ~ly tion and insomni of a benzodiazepine fo eks of treatme~t. ficant anticholi ic effects and is non-cardiotoxic, blood tests, limitsits use. sedative and lacks adverse cardiovascular nlafaxine and nefazadone is too limited effects to make anyCO
ehavioural toxicity (affe
be taken more seriously. She five c o ~ m o n l yprescribed
dothiepin.
~ l d e~atients ~ l ~ ~ i t dementia k
eifler et al. (1989)r orted improvement in depression with underlying A1 imer ' S disease in a trial of imipra
d with moclobemide alsohad meas
,although this was only significant roup without dementia. ~ e ~ ~ e sassociate^ s i o ~~itk
~ esystemic n e ~ a l ise ease
Intolerance toanti essants occurs in out orte-third of patie medical problems (Lipsey et al., 19 pression and com et al. (1989) had to abandon a proposed drug trial usin e in p~ysicallyill older patients because virtual~lyall ha contraindicatiorts or were intolerant of the drug. Physically ill have been treated in an open trial forup to 1year with fluoxetine (Evans, 1993). The g was tolerated well in a range of serious medical disorders, many grav ore recently Roose et al. (1994)found that atiertts with heart disease with a mean ageof 73 years responded less well to uoxetine than 42 comparable atients treated with n o r t r i p ~ l 70). Indeed, only fiveof the 22 fluoxetine-treated patients re
3
with melancholic subtype of major depression (the majority)did especially poorly. The numberof dropouts in each group was similar, So, althou~hthe newer a~~depressants appear to have a niche here, the evidence that they are more appropriate than established antidepressants is not totally convincing and it is surprising howlittle evaluation of them there has been in this context.
EGT is well tolerated in elderly patients (Benbow, 1989). All contraindications are relativeif it is given as a life-saving procedure. It is best avoided within3 months of a stroke or myocardial infarct, or if pulmonary function is poor, Since most complications of EGT in older people are cardiovascular (Benbow, 1989), with a higher risk of asystole (Abrams, 1991)~this system should be examined carefully, andpast a history of an arrythmia accurately determined, plus a cardiac opinionif there is any concern or doubt.~nilateralelectrode placement is often the choice in elderly patients to minimize memory disturbance. However, whilst this is measurably worse with bilateral ECT, it may not be of any practical sipphcance, and some argue that bilateral placem is more effectivein older depressed patients( enbow, 1989). A pacemaker is not a contraindication provided the patient’s bed is insulatedfrom the ground, the patient is not touched and them o ~ t o r equipment ~g properly earthed. Symptoms which predict a good response are similar at allbut ages, one area of difference is that older patients with marked anxiety, often regarded as oor prognosis in younger adults,do just as well (Benbow, 1989).
There havebeen four prospective studies of lithium a~gmentationin elderly et al. (1988) and Flint and Rifat (1994)found response patients to date. Zimmer of younger rates of20% and 23%, respectively, much lower than open studies adults. Parkeret al. (1994) compared a group treated with a single antidepressant (n = 23) with a lithium augmented group (n= 21)in a prospectivestudy. The latter group were foundto have si~ificantly lower depression severity scores on the ~ontgomery-AsbergDepression Scale at follow-up and a trend for lower scores on the Geriatric Depression Scale. Although all but one (LA) group experiencedside-effects,the patient in the lithium augmentation total burden of side-effects was greater in the antidepressant alone group. The results can only be viewed as encouraging but preliminary. In elderly patients, LA is associated with a higher rate of lithium side-effects than in younger patients. Therefore the necessary doselevel is usually one-third to one-half thatof younger patients.
394
Georgotas et al. (1983) gave phenelzine to 20 elderly refractory patients with a 65% recovery rate. However, many older patients are intolerantof the older monoamine oxidase inhibitors. A fashionable strategy; that of combining a tricyclicwith a serotonin re-uptake inhibitor, is reportedly beneficial to older patients with resistant depression (Seth et al., 1992),but it has not been properly evaluated.
There is no agreed strategy for the drug management of an older patient with depressive illness. The following points should be considered. First, a trial of an antidepressant, chosen by tailoring the drug to the patient, should be given for a minimum of 6 weeks provided,of course, it is safe and humane to do so. Second, if this fails,then assuming compliance and adequate dosaging, and after a reviewof psychosocial maintaining factors (see below), then two options appear possible. One is to change to an antidepressant from another class for afurther 6 weeks. Thisstrategy is popular in generaladult psychiatry but there has been no evaluation of it among older patients. Third, if these strategies fail the or patient’s condition is deteriorating, then either EGT or lithium augmentation can be given. These are the only strategies with evidence for efficacy in resistant depression in older patients. In olderpatients continuation therapy should be for a minimumof 12 months (Flint, 1992) or longer for younger patients.
Services for older people with mental illness aroseas a response to the growing challenge of dementia, and social managementhas always been a major component in old-age psychiatry. Depression no less than dementia requires a multidisciplinary approach, ableaddress to social and psychological as well as medical needs. Ideally; a team serving the needs of older peoplewith depression will include a specialist psychiatrist, GPNs, an occupational therapist, a social worker and a psychologist. Most parts of the UK now have such services, although not all have a full compliment of disciplines. In fact there is little research evidence that manipulation of social factors of isbenefit in managing old-age depression, but clinical experience is compelling enough to encourage services toaddress these factors wherever possible. Here, social factors are discussed under the headings: predisposing, precipitating, and maintaining factors, albeit with some overlap between these categories. Although the first two seem more relevant to aetiology than management, an appreciation of them isimportant in planning social care. Accordingly, they will be briefly discussed first.
e
ace
Early life experience
Parental loss, childhood poverty and sexual assault have been shown to predispose to depressionin the elderly (Kaminsky, 1978; McMordie & Blom, 1979) as well as younger people.
Lower levels of education have been shown to be related to greater risk of depression in later life in age heterogenous samples. However, a study by George (1992) suggests this relationship may be weaker in the old. Social s~pport
Brown and Harris (1978) proposed that lack of social support a number is one of of vulnerability factors which may increase the risk of depression in the presence of a provoking agent. Older people generally have fewer social contacts andsources of support. urphy (1982) suggests that amongthe elderly a confidant may act as a buffer against social losses and therefore against depression. Older people witha life-long inability to form intimate relationships are more susceptible to depression given an adverse life event. Emmerson et al. (1989) found the absenceof a confidant tobe a striking risk factor for depression, particularly in men. Pahkala (1990) did not find such an association for men, but did for women. Lastly, Henderson et al.(1986) found depressed elderly patients at home did not areport lack of confiding relationships, but did report a lack of more diffuse relationships. Studies of older people suggest that those with greater social support have lowerofrates depression (Dimondet al.,1987; Holahan& Holahan, 1987; Krause, 1987a; Norris & Murrell, 1984). There is a suggestion that tangible support as well as perception of support is important in the elderly (Krause, 1986a, 1987c; George, 1992). A few studies have found that participation in formal aspects of social s t ~ c ~ rfor e ,example religious or voluntary organizations, protects against psychiatric disorderin the elderly (Idler, 1987; Koenig et al., 1989).
Theoretically, at least, addressing one of these predisposing factors, social support, may reduce the risk of depression developingin an older person. A social worker can assess the financial circumstances of the patient, ensuring that the correct benefits are received. Social support be improved may by introducing the patient to a day centre or luncheon club, or voluntary organizations such as Help the Aged or Age Concern. There may be
397
n ~ ~ e ~ine the n te l ~ e ~ l y opportunities for promoting new social roles and increasing self-esteem through these organi~ations.They may offer an opportu~tyfor confiding relationships to develop. Similarly, aftercare visits by community psychiatric nurses or social workers attached to the old-age psychiatry team may serve increase perceived social support and allow the development of a supportive relationship. In contrast to younger depressed patients, this will oftenbe a -term c o m m i ~ e ~given t , the high rate of recurrence of depression in older patients. Lastly, in theory at least, improved public awareness of the potential benefits of social supports in ameliorating or even preventing depression could lead to improved social networks within the c o m m u ~The ~ . Royal College of Psychiatrist’s ’Defeat Depression’ campaign has resulted in leaflets aimed at the general public; one of these offers advice specifically for older depressed people and their carers. Although many of these interventions are most likely to act as a buffer against relapse and/or recurrence they may also he with management of the acute episode. ~ a t e r r eet~al. s (1994) describe the positive impact of interventions by a CPN to elderly patients with depression in an inner London study. The group with CPN involvement(n=47) had significantly lower depression scores at the end of the study than those who were followed up by primary care services(n= 49). The interventions describedin this study are informative and included: personal supportive therapy, behaviour therapy and teaching relaxation, encouragement to seek medical advice regarding physical health problems, family and marital work and bereavement counselling. However, the biggest impact in this study appears to have been via improved medication education.
Usually in the formof sudden, undesirable, losses these have been shown to have a s i ~ i f i c a ntemporal t relations~pto the onset of depression. In most studies, depressed patients report an excess of life events in the months preceding the onset of depression (Holmes& Rahe 1967; Paykel et al., 1969; Brown Harris, 1978; George, 1992). This has also been people (Holahanet al., 1984; Norris& Murrell, 198 types of loss event are common in old age, for example, loss of significant others through bereavement, loss of health, lossof independence,loss of role, loss of status and lossof income. These events are particularly likely to provoke depression (Paykel et al., 1969). Murphy(1982) found that almost half of a sample of elderly depressed patients had experienced at least one severe
adverse life eventin the preceding year. However, many older people experience such events without becoming depressed. Interestingly an inAustralian study by Emmerson and colleagues (1989), a depressedc o ~ u n i t sample y reported serious life events no more than a healthy control group. Rrown and Harris (1978)and Rrown et al. (1986)proposed that long-ter~ major difficulties (or chronic stressors) such as marital and family conflict, physical ill health, housingand financial difficulties may also act as precipitating factors for depression inthe same wayas adverse life events. Similar findingshave come from studies of older people (Aneshensel et al., 1984; & Huber, 1985; Krause, xg87b; Georgeet al., 1989). As chronic physical illness is more likely in old age, income is generally low, housing ofteninadequate and family conflict no less likely, this research of is particular concern.
~anagementof depression in an older personshould aim to reducethe impact of adverse life events wherever possible. These in the main involve loss and it is essential address to such triggersas soon as practically possible,that is after depressive symptoms have remitted sufficiently. Specific areas to address are bereavement (see next section) and help to adjust to altered (lost) physical health. The strategies employed are no different from adopted those with younger patients. Change in physical health should be thoroughly assessedand t r e a ~ e n t optimized. Referral to other medical specialists should not be ignored because of age. Many old-age psychiatric services have specific types of liaison with colleagues in geriatric medicine. For example, our own practice in chester consistsof a weekly joint clinic attended by both a senior psyc~atrist and a geriatrician. It may be possible to increase independence physical with aids for activitiesof daily livingand a full assessmentof functional abiliti~s and disabilities byan occupational therapist from within the team is helpful. Chronic housing or financial difficulties may be eased following social assessment by a social worker. Services required may range from finan~ial help with heating or cleaning, to help in finding suitable alternative accommodation. However,although many older depressed people request a housing move, often this preoccupation diminisheswith successful treatment of the underlying depressive illness. Anecdotally, it is not infrequent to find a w e l l - m e a ~ grelative who has found alternative accommodation for a depressed older person,who then later regretsthe move. It isimportant to try and understand the dynamics of family and marital conflict where this may have served to precipitate depression. This requires interviews with family membersas well as the patient, both separatelyand together, and is discussedin the next section.
399
Most patients canbe managed in the c o ~ ~ i tFor y some . an in-patient admission may be required, for others a period of increased support and supervision maybe best delivered through a psychiatric day hospital. again, whatever the setting, the key to managing psychosocial factors in late-life depression is multidisciplinary teamwork.
These are factors which are implicated in delaying recovery froman episode of depression orin precipitating a relapse of symptoms during the course of recovery. There is a need for much more research on the ma~tainin factors ~ of old-age depression.
There are few longitudinal studies exploring the relationship between age and recovery from depression, but the prognosis seems poorer for older people (Post, 1972; Mann et al., 1981; Cole, 1983; Murphy, 1983; ICeller et al., 1986). Sex
~ o ~ seem e nmorelikely to recover than men (George et al., 1989). Marit~lstatus
George (1989) found that among the old, married people were less likely to recover from depression than unmarried. Although counter intuitive, the explanation probably lies in the quality of the marital relationship, which may not, in some cases, withstand severe depression.
It has been shown that living with a hostile or critical relative and therefore with a high level of expressed emotion increases the risk of relapse in depression (Vaughn & Leff, 1976; Hooleyet al., 1986). Clinical experience suggests this is also true for the elderly depressed.
Adequate social support has been found to predict recovery from depression & Moran, 1983). In the study by George, in the old (George, 1989; Henderson
social factors were more strongly related to recovery a number than of clinical urphy (1983), however, did not that findone formof social support, a confiding relationship, protected against relapse. Life events
Recovery from depression in old age may be influenced adversely by ~ontinuingstressful life events, as foundby Murphy (1983)~although these findings were not confirmed by George (1989). Chronic stressors
These may alsoplay a part in maintaining depression. In the case of chronic physical ill health, several studies have found this factor reduces thelikelihood of recovery from depression in the elderly (Mam et al., 1981; Murphy, 1983; Baldwin& Jolley 1986; Schulberget al., 1987; Wells et al.,1989; Keitneret d., 1991). Carer stress
Another sourceof chronic stress whichis particularly relevantto the elderly population is the role of a carer which many people (particularly women) have to take on. The dependent relative is usually a parent or spouse, often with physical and/or cognitive problems. Studies have shown high rates of et al.,1985; Cohen& Eisdorfer, 1988; Gallagher et depression in carers (Coppel so that interventions aimed at relieving their stress al., 1989; Schulzet al., 1990)~ are commonly taken on by the old-age psychiatric team. Hinrichsen and Hernandez (1993) have shown that three factors, psychiatric symptoms in the carer, reported difficulties and poorer carer health, were all associated with poorer outcome at one year from major depression in a cohort of patients with a mean age of just over 70 years. Attention to carers may thus enhance the p r o ~ o s iof s the designated patient.
Many of the psychosocial factors implicated in increasing the risk and precipitating the onset of depression in the elderly also seem to be associated with maintaining it. Continued attendance at a day hospital to reduce the time spent with a hostile or critical relative, to encourage involvement in activities, or to increase social support or the opportunity for confiding relationships may be required. In many circumstances attendance at a day
centre or involvement in voluntary activities can provide similar benefits. y c o ~ u n i t psychiatric y nursesor social workers to monitor patient’s progress may also be perceived as increased social support and foster confiding relationships/as already discussed. Such visits are articularly important at critical times for relapse of depression suchas the anniversary of a bereavement. The burdenof care on families looking after a chronically depressed elderly person can be considerable and regular visits may help by educating families about how to manage particular behaviours, or how their own behaviour may be altered to reduce the patient’s symptoms. Very occasionally a patient may require re respite care in a residential facility to reduce the level of expressed emotion at home and may eventually decide on a permanent move.
Paucity of research in this area is disappointing/ particularly in view of the roblems associated with drug t r e a ~ e nin t the elderly, such as increased sensitivity to side-effects and drug interactions, more medical contraindications and frequent voluntary non-compliance. Proven effective alternatives to drug ~ e a ~ ewould n t be very welcome. Some psychological therapies be will briefly reviewed.
ince bereavement is such a common precipitant of depression, it is useful to ensure that at least one memberof an old-age psychiatric teamis skilled in ereavement counselling.
It is not u n c o ~ o for n patientsrecover in^ from late-life depression to suffer from residual anxiety symptoms. They can usefully be addressedby relatively low key interventions, such as teachin relaxation techniques either individually or in a group. fter a prolonged episode of depression, and older patients do take longer to recover (Schneider et al., 2994, art avoidance fuelled by anxiety may develop to everyday activities. Simple graded tasks carried out with a suitably trained member of the team/ or a ’graded discharge’(that is increasing the amount of time at homegradually) if in hospital, can’be extremely effective.
Y elderly patients, this is the most extensively researched psychological treatment.Thetheoreticalmodelunderoposes that cognitiveandbehaviour disturbances of particular a ipitate and maintain depression. e-behavioural therapy (C ent and comparableto younger adults( Clinical reports patients (for example, Yost et al. older patients is, however, li flaws. Thompson (1982; ~ 9 8 3example, for ) ~ who have made thecontribution m to date, used data from a selected middle-class elderly depressed, who are not representative of the general po~ulationof elderly depressed. Therefore results should be viewed with caution, but generally they point to findings comparable with research in younger depressed people. S treated with short-term cognitive thera relational therapy, Gallagher and Tho decrease inde ression inall three groups,but only the cognitive and behaviad maintained treatment gains at %-yearfollow-u 1982). In another study comparing three similar treatbehaviour andbrief dynamic therapy),but with an addi~onaldelayed t r e a ~ e ncontrol t group, aninitial s i ~ i f i c a ndecrease t in & Gallagher, depressive symptoms inall four groups was noted (Thompson 1984).Atx-yearfollow-up, 27% metdiagnosticcriteriaformajor de but there wasno significant difference between the groups. Similar et al., 1987, 1988; Gaston et al., were found at 2-year follow-up ( ~ o m p s o n ,1989; Marmaret al., 1989; Gallagheret al., 1990). Thompson and colleagues.(1991) compared drug therapy ( T, and combined drug and CBT. The combined thera proved more than the drug-only group, but not more than the C using two measures of outcome (Beck Depression Inventory and Hamilton o significant differences were found between the CO and drug therapy-only groups at post-trea~entfollow-up third outcome measure (Research Diagnostic Criteria) there cant differences between the three groups from prepost-trea~ent. to )also compared drug~eatrnentto CBT and found45% re12% atrnent group( ~ p r aor~doxepin) e compared to owever, the patients were not randody S have examined the application of depressed patients. Evans et al. (1982) found problem-solving to be more *
effective than no treatment in a group of elderly depressed blind patients. Steuer et al. (1984) found that CBT and psychodynamic therapy were equally as effective in reducing depressive symptoms in patients with chronic physical illness. Lovett and Gallagher (1988) reported that depression in caregivers was significantly reduced by CBT or behaviour therapy, and both were significantly more effective than being placed on a treatment waiting list. Results of other studies suggest factors which may influence the outcome of treatment with CBT. Those with more endogenous depressions respond less well (Gallagher &z:Thompson, 1983)’as do patients with personality disof treatorders (Thompsonet al., 1988), or with severe symptoms at the onset ment (Beckham, 1989). Also important are patients’ expectancies of chan et and the role they play in the developmentof a therapeutic alliance (Gaston al., 1988,1989; Marmar et al., 1989).
tudies by Gallagher and Thompson mentioned in the previous sectionon GBT suggest that psychodynamic therapy is more e€fective than no treatment €or elderly depressed patients andis as effective as GBT, par~cularlyin the acute phaseof treatment. One study has found interpersonal therapy (IPT) to in the treatmentof the elderly depressed be equally as effective as nortriptyline (Sloane et al., 1985). Klermanet al. (1984) suggests IP treatment adapted for A specific adapelderly peopleis useful in combination with drug treatment. tation of IPT for elderly patients exists (Frank et al., 1993). XPT focuses on interpersonal factors that are highly relevant to the elderly. They include role transitions, interpersonal role disputes, interpersonal deficits and grief. in the treat~entof eynolds et al. (1992) reports evidence for greater efficacy depression by combining IPT and nortrip~line.
Most old people are part of a family network, and €or many elderly mentally ill people the familyis the main source of support. Families continue todevelop and relationships change as the members age. Family relationships can, of course, have negative as well as positive aspects, and there is evidence that family relationships can affect the course of mental illness. Ratna and Davis (1984) found 60% of referrals to a community psychogeriatric service were precipitated by events which mightbe seen to represent a changeof some sort in a significant relationship, such as retirement, illness in a carer or a bereavement. Clinical practice suggests that conflict in relationships often plays a significant part in the course of mental illness in the elderly, although researchin this areais scarce.
Further progress has been made in this fieldin recent years. describes the development of a family therapy clinicin the old-age psychiatry anchester. Two members of a multidisciplinary team interviewed the family while theof rest the team observed from behind a two-way assessment ortherapy session of about 45 min, the team e and the sessionwas then reconvened to feed-back to ough depression was not the problem in all the families nbow and colleagues, the perceived uses of the clinic, for example promoting ~derstandingof a problem within the family context and coordinating support between different family membersand between family members and services, suggests in that, the managementof ~epression in the elderly, family therapy could be useful in certain cases. However, there has been no systematic evaluation of family therapy in older depresse patients.
Abas, M.A., Sahakian, B.J. & Levy, R. (1990) Beckham, E.E. (1989) Improvement after evaluation in psychotherapy of depression: evidence Neuropsychological deficits and CT scan of a placebo effect? Journal of Clinicalpsycho lo^, changes in elderly depressives. Psychological Medicine, 20,507-520. 451 945-95O. Abrams, R. (1991) Electroconvulsive therapy inthe Benbow, S.B. (1989) Therole of electroconvulsive therapy in the treatment of depressive illness in Psychiatric Clinmedically compromised patient. ~al 155,147-152. old age.~ r i t i s h ~ oof~Psychiat~, ics of North America, 14, 871-4385. therapy inthe elderly. Alexopoulos, G.S., Abrams, R.C., Young, R.C. & Benbow, S.M. (1988) Family In:Current Approac~esin Aflective Disorders in the Shamoian, C.A. (1988) Cornel1 Scale depresfor Elderly (eds E. Murphy & S. W. Parker) pp. sion in dementia.Biological Psychiatry, 23,27144-48. Duphar Laboratories Ltd,Southamptom. 284. life. In: American Psychiatric Association (1994) Diagnos- Blazer, D. (1989) Affectivedisorders in late Geria~ricPsychiatry (eds E. Busse & D. Blazer), tic and StatisticalManual of Mental Disorders,4th pp. 369-401. CambridgeUniversity Press, Camedn. American Psychiatric Association, Washbridge. ington, DC. heshensel, C.S., Frerichs, R.R. & Huba, G.J. (1984) Blazer, D., George, L.K. & Landerman, R. et al. (1985)Psychiatric disorders:a ~ r a l / u r comb~ Depression and physical illness: a multiwave, ~ ~Social e a l ~ h parison. Archives of General Psychiatry, 42,651non-recursive model.~ o u ~ n a l oand B e ~ v ~ o u25,350-371. r, 656. Baldwin, R.C. & Jolley, D.J. (1986)The prognosis Blazer, D.,Hughes, D.C. & George, L.K. (1987) The aZ epidemiologyof depression in an elderly comof depression inold age. ~ r i t i s h ~ oouf ~PsyAmerican~ournaZof Epidemimunity population. chiatry,~49,574-583. OlOgY, 11,521-37. Baldwin, R. (1988a) Late life depression-underh ~ o u ~ 296,519. al, Brown, G.W. & Harris, T.O. (1978)Social Originsof treated? ~ r i ti s Med~cal Depression: A Study of Psychiatric Disorder in Baldwin, R. (198817)Delusionaland non-delusional Women. Tavistock, London. depression in late life: evidence for distinct Brown, G.W., Bifulco, A. & Harris, T. et al. (1986) t~, subtypes. ~ r i t i s~h o u ~of aP l s ~ c h ~152,39-44. Life stress, chronic subclinical symptoms and Barraclough, B M . (1971) Suicide in theelderly. In: vulnerability to clinical depression. ~ o u r n aof~ Recent Developments in Psychogeriatrics (eds Aflective Disorders, ~ 1 ~ 1 - 9 . D.W.K. Kay & A. Walk).Headley Bros., Kent. J.E. Burke, W.J.,Houston, M.J., Boust, S.J. Beck, A.T.,Ward, C.H., Mendelson, M., Mock, & Roccaforte, & Erbaugh, J. (1961) An inventory for measurW.H. (1989) Use the of Geriatric Depression Scale ing depression.Archives of General Psychi~try,4, in dementia of Alzheher type. ~ournalof American Geriatric Society, 37, 856-860. 561-571.
~ h a ~ trr8: e r ~ a n a ~ e ~ine the n telderly Cohen, D. & Eisdorfer, C. (1988) Depression in Gallagher, D. 8:Thompson, L.W. (1982) Differenfamily members: caring for a relative with tial effectivenessof psychotherapies for the treatAlzheimer’s Disease. Journal of the American ment of major depressive disorder in older adult Geriatric Society, 36,885-889. populations.psycho the rap^; Theory, Research and Cohen-Cole, S.A. & Stoudemire, A.(1987) Major Practice, 19,482-490. depression and physical illness: special con- Gallagher, D.& Thompson, L.W. (1983) Effectivesiderationsin diagnosis and biologic treatment. ness of psychotherapy for both endogenous and Psychiatric Clinics of ~ o r t America, h xo,1-17. nonendogenous depression in older adult Cole, M.G. (1983) Age, ageof onset and courseof outpatients. J o ~ r n foff ~~ e r o n t o ~ o g38, y , 707primary depressive illnessin the elderly.Cana712. at~, dian J o ~ of~P s a~ c~h ~ 28,102-104. Gallagher, D., Rose, J. & Rivera, P. et al. (1989) Coppel, D.B., Burton, C., Becker,J. et al. (1985) The Prevalence of depression in family care-givers. relationshipsof cognitions associated with cop- Gerontologist,29,449-456. ing reactions to depression in spousal caregivers Gallagher, D., Hanley-Peterson,P. & Thompson, of Alzheimer’s Disease patients. Cognitive L.W. (1990) Maintenance of gains vs. relapse T h e r a and ~ Research, g, 253-266. following brief psychotherapy for depression. Crewe, H.K., Lennard, MS,, Tucker, G.T., Woods, J o ~ ~ofaConsulting l and Clinical P s ~ c h o l o58, ~, ER. & Haddock, R.E. (1992) The effect of selec371-374. tive serotonin re-uptake inhibitors on cytochrome Gaston, L., Marmar, C.R.& Thompson, L.W.et al. P4502D6 (CYP2D6) activity in human liver (1988) Relation of patient pretreatment characmicrosomes. British J o ~ r n aof l CiinicaZ P ~ r m a teristics to the therapeutic alliance in diverse cology, 34,262-265. psychotherapi~.Journal o f c o n s ~ ~ t iand n g CliniCrowell, B.A., Jr, George, L.K. & Blazer, D. et al. cal psycho lo^, 56,483-489. (1986) Psychosocial risk factors and urban/ Gaston, L., Marmar, C.R. 8r Gallagher, D. et al. rural differences in the prevalence of major (1989)Impact of confirming patient expectations depression. British Journal of Psychiatry, 149, of change processes in behavioral, cognitive, and brief dynamic psychotherapy. Psychothera~y, 26, 307-314Dimond, M., Lund, D.A. & Caserta, M.S. (1987) 296-302. The role of social supportin the first two years George, L.K.(1988) Social partici~ationin later We: of bereavement in an elderly sample. Gerontoloblack/white differences, In:The Black Amer~can ~syc~osocia~ Elderly: Research on Ph ~s ic aand l gist, 27,599-604Emmerson, J.P., Burvill, P.W., Finlay-Jones, R. & HeaZth, (ed J.S. Jackson), pp. 99-126. Springer, Hall, W. (1989) Life events, life difficulties and New York. confiding relationships in the depressed elderly. George, L.K.(1989)Social-economic factorsin geriBritish Journa~of P s y c h ~ a t155,787-792. ~, atric psychiatry. In: Geriatric Ps ~c h ~a t(eds r y J. Evans, R.L., Werkhoven, W. & Fox, H.R. (1982) Brody & G.L. Maddox), pp.203-234. American Treatment of social isolation and lonelinessa in Psychiatric Press, Washington, DC. sample of visually impaired elderly persons. George, L.K.(2992) Social factors and the onset and outcome of depression. In:Aging, ~ e a l t h ~ e h a v i P s ~ c h oReport, ~ o ~ 51,103-108. Evans, M.E. (1993) Depression in elderly physiours and Health Outcomes (eds J.W. Schaie, J.S. House & D.G. Blazer), pp. 137-160. Lawrence cally ill inpatients: a 12-month prospective study International Journal of Geriatric Psychiatry, 8, Erlbaum, Hillsdale, NJ. George, L.K., Blazer, D.G., Hughes, D.C. et al. 587-592. Flint, A.J.(1992) The optimum duration of antide(1989) Social support and the outcome of major pressant treatment in the elderly ~nternational depression. ~ r i t i ~ournal s~ of Psyc~iatry,154, Journal of Geriatric Ps y c h i a t ~7,617-619. , 47&-485. Flint, A.J.& Rifat, S.L. (1994) A prospective study Georgotas, A.& McCue, R. (1989) The additional of lithium augmentation in antidepressantbenefit of extending an antidepressant trial past seven weeks in the depressed elderly Internaresistant geriatric depression. Journal of Clinical 14,353-356. Psychiatry,4, 191-195. Psychopharmaco~o~, t ~ o n a ~ J oofuGeriatric ~al Frank, E., Frank, N.,Comes, C. et al. (1993) Inter- Georgotas, A., Friedman, E., McCarthy, M. et al., personal psychotherapy in the treatment of late-(1983) Resistant geriatric depressions and therapeutic response to monoamine oxidase inhilife depression. In:New Applicatio~sof I n t e ~ e r bitors. Biological P s y c h i a ~18,195-205. sonal Psychotherapy (eds G.L. Merman & M.M. ~, Weissman), pp. 167-198. American Psychiatric Hebenstreit, G.F., Baumhackl, U.,Chan-Palay, V. et al. (1991)Proceedingsof the F@h Congress of the Association, Washington,DC.
~ h ~ ~18:t ~e r~ n ~ ~in the e ~e ~e ~ ne rt ~ y depression and co-morbid psychiatric or mediInternational Psychogeriatric Association,p.31, July cal illness (compound depression). Amer~can 1991, Rome. Journalof Psychiatry,148,345-350. Henderson, AS. & Moran, P.A.P. (1983) Social reKeller, M.B.,Lavori, P.W. & Rice, J. et al. (1986)The lationships during the onset and remission of persistant risk of chronicity in recurrent episodes neurotic symptoms: a prospective community of nonbipolar major depressive disorder: a prostudy. B r i f i sJ~o u ~ aolf P s y c h i a f ~143,467-472. , spective follow-up. American ~ o u ~ofaPsychil Henderson, AS., Grayson, D.A., Scott, R. et al. (1986)Social support, dementia and depression atry, 143'24-28. & Rounsaville, B.J. among the elderly livingthe in Hobart Commu- Merman, G.L., Weissman, M.M. et al. (1984)I n fe ~e r s o n aPsycho~herapy l of Depresnity. PsychoZogicalMedicine, 16, 379-390. sion. Basic Books, New York. Hinrichsen, G.A. & Hernandez, N.A. (1993) FacK.G., Cohen, H.J. & Blazer, tors associated with recovery from and relapse Koenig, H.G., Meador, D. (1988) Depression in elderly hospitalised into major depressive disorder in the elderly. patients with medical illness.Archives of InterAmerican J o u ~ aofl Psychiat~,150, 1820-1825. nal Medicine, 148,929936. Holahan, C.K.,Holahan, C.J. & Belk, SS. (1984) Koenig, H.G., Goli,V.,Shelp, F.et al. (1989)AntiAdjustment in aging: the role of life stress, depressant use in elderly medical patients: leshassles, and self efficacy. ~ e a lPs~chology, f ~ 3, sons from an attempted clinical trial. J o u ~ a Zof 315-328. e, Holahan, C.K.& Holahan, C.J. (1987) Self efficacy, General Infernal~ e d i c i ~4,498-505. Koenig, H.G., Siegler, I.C. & George, L.K. (1989) social support and depression in aging: a longiReligious and non religious coping: impact on tudinal analysis. JournaZ of Gerontology,42,65adaptation in later life. Journal of ReZigion and 68. Holmes, T.H. & Rahe, R.H. (1967) The Social ReAging, 51 73-94. adjustment Rating Scale. J o u ~ aofl Psyc~osomaf~c Kovacs, NI., Rush, A.J., Beck, A.T. et al. (1981) Depressed outpatients treated with cognitive Research, 11,213-218. therapy or pharmacotherapy. Archives of General Hooley, J.M., Orley, J. & Teasdale, J.D. (1986) Psychiatry,38,33-39. Levels of expressed emotion and relapse in depressed patients.British ~ournaZof Psychiatry, Krause, N.(1986a)Social support, stress and well being among older adults. Journalof G ~ o n ~ o l o g y , 148,642-647. Idler, E.L. (1987) Religious involvement and the 4% 512-519. health of the elderly: some hypotheses and an Krause, N. (1986b) Stress and coping: reconceptualizing the role of locus of control beliefs. initial test. Social Forces, 66,226-238. Jackson, R.& Baldwin, B. (1993) Detecting depres- ~ournuZof GeronotoZogy,41, 617-622. sion in elderly medicallyill patients: the useof Krause, N. (1986~)Stress and sex differences in depressive symptoms among older adults. Jourthe Geriatric Depression Scale compared with nal of Geronotology,41,727-731. medical and nursing observations. Age and AgeKrause, N.(1987a)Chronic financial strain,social ing, 22,349-353support and depressive symptoms among older Jarvik, L.F., Mintz, J.M. & Steuer, J. et al. (1982) Treating geriatric depression: a 26-week interim adults. Psychology of Aging, 2,185-192. analysis. Journalof the American Geriatric Society, Krause, N. (1987b) Chronic strain, locus of control, and distress in older adults. Psychology of 30, 713-717Kaminsky, M. (1978) Pictures from the past: the Aging, 2,375-382. uses of reminiscence in case work with the Krause, N. (1987~)Life stress, social support and self-esteem in the elderly population. Psychoelderly. Journal of GerontoZogical Social Work, I, logy of Aging, 2,349-356. 19-31. Kastenbaum, R. (1987) Prevention of age related Lipsey, J.R., Robinson, R.G., Pearlson, G.D.,Rao, K. & Price, T.R. (1984) Nortriptyline treatment problems. In: ~ a n d ~ofo Clinical o~ Gerontology of post-stroke depression: a double-blind study. (eds L.L. Carstensen & B.A. Edelstein), pp. Lancet, 333,297-300. 322-353. Pergamon, New York. Katz, I.R., Simpson, G.M., Curlik, S.M., Parmelee, Lishman, W.A. (1998)Organic Psychiatry,3rd edn. Blackwell Science, Oxford. P.A. & Muhly, C. (1990) Pharmacologic treatLovett, S. & Gallagher,D.(1988)Psychoment of major depression for elderly patients in educational interventions for family caregivers: residential settingsJ o u ~ aolf Clinical Psychiatry, preliminary efficacy data. Behavior 'Therapy, 51 (SUPPl. 4); 41-47. Keitner, G.I., Ryan, C.E.& Miller, J.W. et al. (1991) 19,321-330. Twelve-month outcome of Datients with maior Mam. A.H., Tenkins, R. & Beslev, E. (1~811The
407
twelve month outcome of patients with neurotic illnessin generalpractice. Psycholo~ca~ Medicine,
P s y c h i a of ~~ Late Life (eds R. Levy& F. Post), pp. 176-196. Blackwell ScientificPublications, Ox-
ford. 1% 535-550, Marmar, C.R., Gaston, L., Gallagher,et al. D.(1989) Post, F. & Shuhan, K. (1985)New views on old Alliance and outcome in late life depression. age affective disorders. In: Recent Advances in Journal of Nervous and Mental Disorders,177, 464Psychogeriatr~cs,Vol. 1 (ed. T. Arie), pp. 119-140. Churchill Livingstone,Edinburgh. 472. McMordie, W.R. & Blom, S. (1979)Life review Postma, J.U. & Vranesic, D. (1985)Moclobemide therapy: psychotherapy for the elderly.Perspecin the treatment of depression in demented tives in Psychiatric Care, 4,162-166. geriatric patients. Acta Ther~peutica,11,1-4. Murphy, E. (1982)Social origins of depression in Ratna, L. & Davis, J. (1984)Family therapy with . old age. British Journalof Psychiat~, 141,135-142. the elderly mentally ill some strategiesand techMurphy, E. (1983)The prognosis of depression in niques. BritiskJourna~of Psychiatry, 145,311-315. 142,111-119. old age.BritishJou~alof Psychiat~, Ray, W.A.,Friffin, M.R., Schaffner, W., Baugh, U.K. Myers, J.K., Weissman, M.M. & Tischler, G.L.et al. & Melton, L.J.(1987)Psychotropicdrug use and (1984)Six-month prevalence of psychiatric disthe risk of hip fracture. New England ~ournalof orders inthree communities.Archives of General Medicine, 316,363-369. P s yc h i a t ~41, , 959-970. Reynolds, C.F., Perel,J.M.,Cornes,C. & Norris, EH. & Murrell, S.A.(1984)Protective funcKupfer, D.J. (1989)Open-trial maintenance tion of resources related to life events, global pharmacotherapy in late-life depression: surstress and depression in older adults. Journal of vival analysis. Psychiatric Research, 27,225-231. ~ e a l t and h Social Behaviour, 25,424-437. Reynolds, CF., Frank, E. & Perel, J.M.et al. (1992) ORiordan, T.G., Hayes, J.P., O’Neill, D., Shelley, Combined pharmacotherapy and psychoR., Walsh, J.B. & Coakley, D, (1990)The effect therapy in the acute and continuation treatment of mild to moderatedementia on the Geriatric of elderly patients withrecurrent depression: a Depression Scale and on the General Health preliminary report. American Journal of Psychiatly, 149,1687-1692. Questionnaire. Age and Ageing, 19,5741. Old Age Depression Interest Group (1993)How Reifler, B.V., Teri, L. & Raskind, M. et al. (1989) Double-blind trialof imipramine in Alzheimer’s long should the elderly take antidepressants?A disease patients with and withoutdepression. double blind placebo-controlled study of continuation/prophylaxis therapy with dothiepin. A ~ e r i c a nJournal of Ps y c h ia t ~, 146,45-49. British J o u ~ aofl Ps y c h i a t ~162,175-182. , Roose, S.P., Glassman, AH., Attia, E. & Woodring, Pahkala, K. (1990)Social and enviromental factors S. (1994)Comparative efficacy of selective serotonin reuptake inhibitors and tricyclics in Journal and depression in old age. ~nternatio~al of Geriatric Ps y c ~ z i a t2,119-123. ~, the treatment of melancholia. A ~ e r ~ c a n J o ~ r n a ~ Parker, K.L., Mittmann, N. & Shear, N. W.(1994) of Psyck~a~ry, 151,1735-1739. Lithium augmentation in geriatric depressed Ross, C.E. & Huber, J. (1985)Hardship anddepresoutpatients: A clinical report. ~ n t e ~ n a t i o ~ a lsion. Journal of ~ e a l t hand Social ~ehaviour,26, 312-327. ~ o u r nofa ~Geniatric P s y c h ~ a t9, ~ , 995-1002. Paykel, E.S., Myers, J.K. & Dienelt, M.N. et al. Roth, M,, Mountjoy, C.Q. & Amrein, R. (1996) Moclobemide in elderly patients withcognitive (1969)Life events anddepression: a controlled decline and depression: an international doustudy. Archives of GeneralPsychiatry, 21,753-761. ble-blind placebo-controlledtrial. British Jou~naZ Pearlson, G.D., Rabins,P.V., Kim, W.S.et al. (1989) of Psychiat~,168,14p157. Structural brainCT changes and cognitive deficits with and without reversible dementia Schneider, L.S., Reynolds, CF., Lebowitz, B.D. & Friedhoff, A.J. (1994)Diagnosis and ~ r e a t ~ eofn t (’pseudodementia’). PsychoZogical Med~cine,19, Depression in Late Life. American Psychiatric 573-584. Press, Washington,DC. Peterson, P. (1968)Psychiatric disorders in primary l Clinical Endo- Schulberg, H.C., McClelland, M. & Burns, B.J. hyperparathyroidism. J o u r ~ aof (1987)Depression and physical illness: the c r ~ ~and oM ~ eot a~~ o ~28,1491-1495. ~s~, prevalence, causation and diagnosis of co-morPitt, B. (1982)Psychogeriatrics/2nd edn. Churchill bidity. Clinical psycho lo^ Review, 7,145-167. Livingstone, Edinburgh. Post, F. (1972)The management and natureof de- Schulz, R., Visintainer, P. & Williamson, G.M. (1990)Psychiatric and physical morbidity: efpressive illnesses in late life: a follow-through fects on caregiving.Journal of Gerontolo~caland ~, study. British Journal of P s y c h i a ~121,393-404. Psychological Science, 45, 181-191. Post, F. (1982)Functional disorders. In: The
~ h a ~ t18: e r~ a n a ~ e ~ine the n tel~erly Proceedings of the 99th meeting of the American Seth, R., Jennings, A.L., Bindman, J., Phillips, J. & PsychoZogical Association, San Francisco, Calif., Bergmann, K. (1992) Combination treatment August. with noradrenalin and serotonin reuptake inhibitors in resistant depression. British Journal Vaughn, C.E. & Leff, JP. (1976) Influence of family and social factors on the course of psychiatof Psychiatry/ 161,562-565. ~ l lag, ric illness. British ~ o u r of~ Psychiatry, Sherwood, N. & Hindmarch, I. (1993)A compari125-137. son of five commonly prescribed antidepressWarheit, G.L.,Holzer,C.E. & Arey, S.A. (1975) Race ants with particular reference to behavioural and mental illness: an epidemiologic update. toxicity,Human Psychopharmacology,8,417-422. JournaZ of Health and Social Behaviour, 16,243-256. Sloane, R.B., Staples, F.R. & Schneider, L.S. (1985) Interpersonal therapy vs. nortriptyline for de- Waterreus, A., Blanchard, M. & Mann, A. (1994) Community psychiatric nurses for the elderly: pression in the elderly. In: Clinical and Pharmafew side-effects and effective in the treatment (eds G.D. cological Studies in Psychiatric Disorders , of depression. ~ournalof Clinical ~ u r s i n g3,299Burrows, T.R. Norman & L. Dennerstein), pp. 306. 344-346. John Libbey, London. Steuer, J.L., Mintz, J. & Hammen, C.L. et al. (1984) Wells, K.B., Stewart, A. & Hays, R.D. et al. (1989) The functioningand well-being of depressed paCognitive-behavioral and psychodynamic tients: results from the MedicalOutcomesStudy. group psychotherapy in treatment of geriatric depression.Journal of Consulting and Clinical Psy- Journal of the American Medical Association,262, chology, 52,180-189. 914-919. Teri, L. & Lewinsohn, P.M. (1986) Individual and World Health Organization (1992) Tenth Revis~on of the Int~national Class~cation of Diseases. World group treatment of unipolar depression: comHealth Organization, Geneva. parison of treatment outcome and identification of predictors of successful treatment outcome. World Health Organization (1993) The ICD-10 Class~cationof Mental and Behavioural Disorders: Behaviour Therapy,17,215-228. Research Criteria. World Health Organization, Teri, L. & Umoto, J. (1991) Reducing excess disGeneva. ability in dementia patients: training caregivers to manage patient depression. Clinical Yesavage, J.A. (1988)Geriatric Depression Scale. Psychophar~acology Bulle~in/ 24,709-710. Gerontologist, 10,4943. Yesavage, J.A., Brink,T.L., Rose, T.L. & Lum, 0. Thompson, L.W. & Gallagher, D. (1984) Efficacy (1983) Development and validation of a geriatof psychotherapy in the treatment of late life ric depression screening scale: a preliminary depression. Advances in Be~avioralResearch and report. Journal of Psychiatric Research,17'37-49. Therapy, 6,127-139. Thompson, L.W., Gallagher,D. & Beckenridge, J.S. Yost, E., Beutler, L., Corbishley, M.A. et al. (1986) Group Cognitive Therapy: A Treatment Approach (1987) Comparative effectiveness of psychofor Depressed old^ Adults. Pergamon, NewYork. therapies for depressed elders. Journal of ConZung, W.W.K. (1967) Depression in the normal 55,385-390. sulting and Clinical Psychology, aged. Psychosomatics/ 8,287-292. Thompson, L.W., Gallagher, D. & Czirr, R. (1988) Personality disorder and outcome in the treat- Z i m e r , B., Rosen, J., Thornton, J.E., Peral, J.M.& Reynolds, C.F. (1988) Adjunctive low dose ment of late life depression. ~ournalof Geriatric lithium carbonate in treatment resistant Psychiatry and ~eurology/21,133-146. depression: a placebo-controlled study. Journal Thompson, L.W., Gallagher, D., Thompson, D. et o ~ C l ~ n ~Psychopharmacology, ca1 8(2), 120-124. al. (1991) Treatmentof late life depression with cognitive/behavioural therapy or desipramine.
MATTHEW HOTOPF AND S I M O N WESSELY
Depression in patients with physical illness is common, fre~uently mrecognized and leads to dilemmas in diagnosis and treatment. Depression may worsen disability caused by the physical illness, accentuate pain, affect prognosis and reduce compliance with physical treatments. Physical illness may restrict treatment options for, and worsen the prognosis of, depression. In this chapter we shall give an overview of the literature on prevalence and predictors of depression in the physicallyill and someof the diagnostic difficulties implicit in the relationship.We shall then review the literature on physical, social and psychological treatments and their efficacy in the management of depression in the physically ill. This review is based on aM E D L I ~ E search of depression in thecontext of specific diseases and physical illness in general, plus extensive useof cross-referencing from previous reviews.
iolo~y of ~ e ~ r e s si i o ~
There is an extensive literature on the prevalence of depression in physical & Hawton, 1986; Lishman, illness, reviewed in more detail elsewhere (Mayou 1987; Fann& Tucker, 1995). The followingis a brief resum&. Prevalence studies have used two broad designs, single disease studies studies of mixed clinical populations. Oneof the most studied diseases is cancer, reviewedby 1.5-50% over 20 studMcDaniel et al. (1995). Rates of depression ranged from ies, with a median value of 22%. Most of these estimates were based on interviews and relatively stringent definitions of depression. Similarly, Cummings (1992) has reviewed the relationship between depression and Par~nson’s 4 to 70%. There are similar reviews for disease and estimated prevalence from & Roose, 1990) and stroke (Robinson et depression and heart disease (Dalack al., 1990). Studies of single diseases and depression allow links between disease process and depressionto be explored, and may cast light on our understanding of ‘primary’ depression. For example, the finding of very high rates of deof evidence for the role of pression inCushing’sSyndrome was an early piece corticosteroids in the ~athophysiologyof depression (Jeffcoate et al., 1979).
Likewise, stroke is strongly associated with depression and there is some evidence to suggest left hemisphere stroke is more strongly associated with & Price, 1982), although depression than right hemisphere stroke (Robinson this has not been replicated (Sharpeet al., 1990; Schwartzet al., 1993). These examples might suggest specific mechanisms in the aetiologyof depression, but say less about the overall burdenof depression in the physically ill. We h o w of no physical illness which not is associated with depression!This indicates that whilst specific biomedical mechanisms may play a role, the effect of illness and its resulting disability are likely to be important non-specific stressors. The second approach to measuring prevalence of depression is to use a heterogeneous sampleof physically-ill patients among clinical populations. These studies are often motivated more for public health ends than a specific interest in aetiological research, and aim to inform planners and policy makers of prevalence as well as consider issues such as recognition and management. Again, high rates of depression among clinical populations have been found.In broad terms thereis a tendency for rates to increase from primary care, to secondary-care (Katon & Sullivan, 1990). From Mayou and Hawton’s (1986) review, rates range from13-61% in medical in-patients and x4-52% in medical out-patients. These estimates depend on the case definition used, the higher estimates being based simply on scoring above threshold on the General Health Questionnaire (GHQ) (Goldberg, 1972). When studies are restricted to those which used the Clinical Interview Schedule (CIS) (Lewis et al., 1992) or Present State Examination (PSE) (Wing et al., 1974) prevalence falls to 13-39% for medical in-patients and 20-45% for outpatients, indicating higher rates than seen in c o m h t y samples. The prevalence of studies on which these estimates are based use mixed populations of patients, and it is often unclear what proportion havea defined physical disease. Physical illness is not always distinguished from somatization, which is also associated with high rates of depression and followsa similar pattern of increasing prevalencein different clinical settings. Van Hemert et al. (1993) addressed this issue ina survey of newly referred medical out-patients. Just under one halfof the patients seen hada clear medical explanation for their symptoms and their prevalence of psychiatric disorder was 15%. The remainder with doubtful or no medical diagnosis rates had of 40%. High rates might also be a result of different referral patterns among depressed and nondepressed physically ill individuals. Depression may modify illness behaviour in individuals withphysical disease and increase their chanceof consulting. In other words, there may be an interaction between physical illness and depression leading to increased consultation. One recent prevalence surveygot around some of these problems using DSM-IV (American Psychiatric Association, 1994) criteria to diagnose psychi-
atric disorder in 313 acute medical in-patients (Silverstone, 1996). Psychiatric disorder was diagnosed in27.2% and only 5.1% had major depression. This relatively modest figure (which is still higher c than o studies)is ~ prob- ~ ably due to the inclusion of acute patients admitted only, who were then interviewed 7 days following admission,and the useof a stringent modified criteria for depression (Endicott, 1984).
Severity of physical illness is associated with degree of psychiatric disorder, but not as strongly as some might think. h predicting depressionin the context of physical illness there is a need to take note of premorbid factorssuch as personality and past psychiatric history; aspectsof the illness, including its p r o ~ o s i and s chronicity; particular symptoms suchas pain; t r e a ~ e n t s , especially those which are painful or mutilating; and social factors such as social support or the effectsof the illnesson employment prospects. and Hawton(1986)identified a few general themes regarding those d risk. They suggest that younger patients, women and those with revious psychiatric illness have higher risk. Social problems such as housing, social isolation and unemployment also appear important. These risk factors are similar to those seen in community studies of depression. In a review of risk factors for depression in cancer patients, Harrison and Maguire (1994) identified similar demographic factors, premorbid psychiatric illness or neuroticism,and lack of social support as factors which appear to predict depression independent of disease severity There is a risk of circularity in determining who is at greatest risk. For example, Godding et al. (1995) in a cross sectionalstudy found poor quality of life wasa predictor of depression in patients with cancer. Since qualityof life is assessed using highly mood dependent ratings, it is not clear whether this is a causal relationship.
There isan extensive literature on the detection of psychiatric disorder in primary care. There has been less extensive research in hospital samples, but there aregrounds to suspect that even fewerpatients are detected.~ a g u i ret e al. (1974) screened medical in-patientswith a clinical interview and used casenote evidence to determine whether the clinical team were of aware cases of psychiatric disorder, They found approximately half the cases had been recognized. Factors which increased the likelihood of recognition were uncooperative noisy patients and those witha past psychiatric history.ilki ins on et al. (1987) applied a two-stage screening method to detect psychiatric ‘cases’ on the Clinical Interview Schedule (CIS) among dia~eticpatients and
~ ~ a ~~ t9~e: ~r r e s s i and o n ~ ~ ~ illness s i ~ a l compared this to physician’s ratings. Using these criteria they found that approximately two-thirdsof cases would have been undetected. Similar finds have been reported by Brid es and Goldberg (1984) who found72% of depressed neurological patients were not diagnosed and P4rez-Stableef al. (1990) who foundthat among medicalout-patients only 37% of cases on the Diagnostic I n t e ~ i e wSchedule ( obin et al. 1981) were reco hysicians. This study also de nstrated quite frequent physicians. Finally, Silverstone(1996) demons~ated that DSM-W cases were more likely to be reco ized by nurses (61%) than doctors(41%). In our clinical experience there is considerable variation between medical specialities in their ability to detect depression. Traditionally, oncologists and palliative care physicians, and more recently IV physicians, have been unable to ignore the devasta~ngpsychological impactof the illnesses they treat. This interest is demonstratedby the predominance of these illnesses in the psychotherapy literature. s
e reason for the low detection rates of depression maybe that it is seen as an understandable consequence of physical illness.‘The psychological aspects of depression are ’understandable’ in the contextof what maybe a devastating physical illness. Thusit could be argued that we are not detecting high rates of depressionas an illness, but high ratesof distress as a normal partof the process of adjustment to disability, handicap or impending death. Furerm more, symptoms of depression and physical illness overlap. Fatigue, sleep ain, loss of appetite and loss of weight are fea ses as well as being symptoms of depression ( rlap in these symptoms could lead to artefac h rates of depression in the physicallyill. House (1988) has argued that the diagnosisof depression, based on subjective symptoms, isa phenomeno~ogicalone. Depression isa syndrome, not a disease entity which can be reliably diagnosed on the basis of objective investigations. Psychiatric epidemiologists and liaison psychiatrists require criteria to base the diagnosis on, and if there is overlap between the symptoms of physical illness and depression this does not alter the fact that the individua~still meets criteria for depression. Thus studies which haveaused standardized assessment suchas the General Health~uestionnaire(GH Beck Depression Inventory(BDI) or CIS, are valid, because the diagnosis is reached in a consistent manner, If we ignore the problem of possible misclassification from cross sectional studies in physically-ill populations, the main question forcli~iciansand researchers becomes: ‘Doesthe presence of depression matter in patients with physical illness?’ The next section will attempt to address this question.
4=3
It has long been known that patients with depression have a shorter life expectancy than the general population and that this is not due simply to higher suicide rates (Malzberg, 1937). There is an extensive literature on copi styles in cancer and survival (Greer, 1991). These studies indicate that either a 'fighting spirit' or denial are protective and more depressive coping styles are associated with a poorer outcome. Furthermore, there has been growing interest in depression itself as a risk factor for poor prognosis of physical illness. For example, Silverstone (1990) demons~ated that acutely ill patients admitted to hospital with myocardial infarction, pulmonary embolus, gastro-intestinal haemorrhage or stroke, were more likely to die if they reported symptoms of depression. This finding did not take into account possible confounders; however, the results were striking.We calculated the risk ratio of death in those with depression to be 5.8 (95% Confidence Interval: 4.4-7.6). Subsequently, Frasure-Smith et al. (1993) have shown that major depressionat the timeof a myocardial infarction is an independent risk factor for death at 6 months (Hazard Ratio 4.3; 95% GI 3.1-5.4). Given the high prevalence of depression in this study (15.7%)~this result implies that depression couldbe an important and theoretically preventable of cause death in patients suffering myocardial infarctions. It is unclear whether these results generalize to other physical illnesses. Similar findings have been reported for patients with stroke (Morris et al., 1993). Our researchinto chronic fatigue syndrome indicates the importance of a past history of depression as an independent risk factor for poor prognosis (i.e. prolonged fatigue) following trivial (Wessely et al., 1995) and serious (Hotopf et al., 199613) viral infections.It has been shown that there is a modest correlation between depressive symptoms and hospital of length stay for stroke patients and amputees (Schubertet al., 1992), but we are not awareof stringently designed studies which have found an increase in ~ o r t a l i t yamong those diagnosed with depression in other illnesses. There are many possible reasons why depression might be a poor prognostic indicatorin physical disease. Social class could confound the relationship since individuals from lower socioeconomic groups are likely to have both a worse prognosis of physical disease and a higher prevalence of depression. Depression may lead poor to compliance and motivation to change lifestyle. Thisis supported by Ladwig and colleagues(1994) finding thatpatients with depression were less likely to stop smoking following a myoc infarction, and the finding that depression reduces participation in cardiac that rehab~l~tation programes (Ades et al., 1992). Similarly, there are reports
C h a ~ ~;rg: e r ~ e ~ r e s sand i ~ ~hysical n illness compliance is poor in depressed renal dialysis patients (Rodin et al., 1981). It seems probablethat a web of other psychosocial risk factors are responsible, for example low social support appears to bea risk factor for dying following a heart attack (Caseet d . , 1992). Alternatively, specific organic processes may account for the relationship. It is well documented that depression is associated with a wide variety of changes involving catecholamines, corticosteroids and the i m u n e system. These changes could modify the disease process in physical illness. Nevertheless, fashionable though such theories are at present, we areunaware of any compelling evidence support to them. If the relation between depressionand poor outcome for some physical illnesses were genuine,what would this imply for management? In the absence of clear understanding of the causal pathway, it is difficult to sure be whether more aggressive treatment and rehabilitation of depressed patients with physical illness would improve their outcome. Althougharethere some studies examining the efficacy of identifying depressed physically-ill patients in hospital settingsand using proximal outcomessuch as duration of hospital stay to assess the efficacy of any intervention chosen by the psychiatrist, unfortunately these studies have not given encouraging results (Levenson et al., 1992). This may be because the psychiatric intervention has been minimal and the outcome determined by powerful additional factors relatedto physical disease. It remains to be seen whether aggressive identificationand treatment of depression in physically ill people affects their prognosis. A more proximal and probably more realistic aim should be to improve quality of life among those suffering from physical disease. For example,it is possible to have cancer and a good quality of life. It is barely possible to have cancer, depression and a good quality of life. hysical illness in depression
Coexistent physical illness has been linked a with worse prognosisof depression especiallyin elderly patients (Lyness et al.,1993; Cake et al., 1993).If the onset of the physical illness was responsible for the depression, it is probable that a chronic illness will acta as chronic stressor,so the relationship between physical illness and poor prognosis of depression is perhaps unsurprising. This viewpoint may be overly pessimistic: depression in these patients often may go unrecognized and untreated.
Whilst we agree with House’s comments on the need to maintain a phenomenological approach in research, in clinical practice it is sensible to give more weightto the symptomsof depression which are not likely to a be
direct result of the physical disease. Aswe shall see, there is relativelylittle information on efficacy of t r e a ~ e n t sin this group, so much of the advice on assessment and treatment is little different than that for depression in o settings. he re~erru2.Liaison psychiatrists often emphasize the i sitive handling of the referral, usually in the contextof p cally unexplained symptoms. Even in physically ill patients, referral to a psychiatrist maybe interpreted as a sign that the physician trivial~sing is the physical illness. It is useful to acknowledge the unpleasant nature of the physical illness and then go on to mention how this could be stressful or depressing. Good liaison between the physician and psychiatrists is important, ideally with the psychiatrists based within the general hospital and ttached to specificclinical teams to aid prompt referral. Sturting the u s s e s s ~ e ~Ittis . often best to start the assessment by reviewing the history of the physical illness. This servestwo purposes: reassuring the patient that the physical complaints are being taken seriously ~ d e r sand tanding the impactof the illness.If loss is central to the psychology of depression, there are plentyof opportunities for patients with physical illness to experience it. These may includeloss of ability to work; financialloss; loss of their role in the family; sexual dysfunction and loss of ability to have children,as well as the more obvious threatof loss of life. The psychiatric history should seek to draw an understanding of these life changes. It is also useful to gain an understanding of the impact of the physical illness on day-to-day activities. Asking how the patient spends a 'typical day' may be helpful to assessthis, and gives an ideaof the amountof social support available, and thelimits illness may put on them.It is often surprising how little a patient with a serious physical illness may know of its pro or even why they are in hospital. Sensitive questioning about their ing of the implicationsof illness may reveal bewilderment and profound ance or denial. Patients may have fantasies based on misunderstando ings of diagnosis or prognosis, or may not have had the cuss aspectsof diagnosis with physicians. If this is the case this back to the medical team. Finally, there maybe aspects of the physical illness which contribute to depression directly, for example the use of epressogenic drugs such as corticosteroids. Diagnosis of ~ ~ r e s s i oSince n . physical illness may cause many biological symptoms of depression, there are specific cognitive featuresof the mental state examination which should be given special weight.It is for this reason that the Hospital Anxiety and Depression Scale (HADS) was developed (Zigmond & Snaith, 1983). Features of the mentalstate which shouldbe given more weight include anhedonia, failure of the mood to react to positive events, hopelessness, worthlessness and guilt. The view that such emotions are
~ ~ a19:~~ ~~r ees sri and o n ~ ~ y s i cillness al understandable and therefore cannot be treated is u ~ e l p f u lPhysical . illness is a risk factor for suicide, it sois important to assess suicidal risk carefully in the normal way, and also to be aware of additional means to this, such as access tolethal dru Aspects of the~ ~ y s i cillness a l ~ h i c alter h treat~ent of ~epression.Physical illnesses may limit or contraindicate the of use antidepressants and electroconvulsive therapy T), as discussed later in this chapter. The patient may be on medication W the potential to interact with antidepressants. Cogniti~e~ s s e s s ~ eDementia n~. is an important differential diagnosis of epression in this setting, especially in neurological diseases such as stroke, motor neurone disease, or multiple sclerosis. Alzheimer’s disease is not infrequent in many of the older patients with physical illness and depression may e an early sign. ~ ~ e coft the s ill~ess onf a ~ ~ l y Close ~ e family ~ ~ e members ~ s . are often intimately involved in the care of patients especially if they are severely disabled. The illness may profoundly change relationships, with the patient ng dependent. This may lead to resentments and tenin the relationship and spouses are at increased risk of emotional disorder.It is, therefore, useful to see the spouse or other close family members or carers.
The approach to treatmentof depression in the physicallyill need not differ dramatically from treating depression in other settings. The complications of suffering at least two illnesses increases the need to an take inter~isciplinar~ approach and good liaison between psychiatrists, physicians, general practic o ~ u n i t y )occupational , thertioners (GPs), nurses (bothin hospital and the apists, psychologists and social workersis desirable, but probably rare. ~e shall concentrate on treatments which have been systematically evaluated, but before doing so a few generalc o ~ e n ton s the evaluations published are necessary. Randomized controlled trials (RCTs) of intervention for depression in physical illness suffer from several common faults: they areall small, some with as few as10 patients in each group. This lack of statisticalpower is compounded by their tendency to compare several treatments simultaneously. Another problem for researchers is the temptation to examine multiple outcomes, such as measuresof depression, quality of life, specific outcomes of relevance to the physical disease (such as physiological measure and inventories of specific symptoms), behavioural changes (such as smoking habits) and
4-17
so on. Multipleendpoints, whilst desirable from the clinician’s point of view, lead to serious statistical problems and raise the likelihood of type I error, whereby the null hypothesis is rejected incorrectly. As if this was not bad enough manyof the trials (both antidepressant and psychotherapy) compared several treatments simultaneously. This combination means there are often more endpoints compared between treatments than patients randomized into the trial! Future trials would benefit by specifying a ~ ~outcomes i ~ such as quality of life.
~
A n t i ~ ~ r e s s a nand t s ~ ~ ~ s illness ical
The trials we identifiedof antidepressantsin physical illness are summarized in Table 19.1. Placebo-controlled trials of antidepressants in this group indicate in the mainthat active medication is superior to placebo in terms of recovery from depression. The majority of placebo-controlled RCTs have used tricyclic antidepressants, and found that most patients could tolerate side-effects. Some studies have attempted to demonstrate improvements in physical disease with active medication. Where subjective measures of disease activity are reported there is a relationship between remission from depression and improvement in physical symptoms. However, these changes are not usually paralleled by improvements in objective measures of disease status. We were unable to find any placebo-controlled trialsof newer antidepressants in treating this patient group, although descriptions of non-randomized SSRIs (Huyse et al., 1994) and methylphenidate (Katon &I case series exist for Raskind, 1980). There have been few direct comparisons between different antidepressants. The well-documented side-effects of the tricyclics potentially restricts their usein patients with a wide range of physical illnesses. There has been onetrial of a tricyclic vs.a SSRI in patients with heart disease which, perhaps surprisingly, found both drugs were equally well toleratedbut the tricyclic (nortriptyline)was more efficacious (Roose et al., 1994). If antidepressantswork in RCTs how easy are they to give in clinical practice? In a retrospective study, Popkinet al. (1985) reviewed 1649 psychiatric consultations ina general hospital setting. Seventy-eight cases, were selected, as they metthe criteria of being consultations for depression in which antidepressants were prescribedand a primary medical illness had been diagnosed. Outcome was assessed using case-note information on the success of treatment. h all, theyfound only 40% had a good response (i.e. a clinical recovery was noted), 28% had no response and 32% had to be withdrawn from antidepressant treatment due to side-effects. Half of the withdrawals due to side-
~ ~ a19:p ~ ~ p ~r ~ rs sand ion p ~ y s i ~illness al effects were for delirium. The authors concluded this disappointing result was at odds with patterns of efficacy observed forprimary affective disorder. This study illustrates someof the problems of inferring from the results of observational data. Without a comparison group it is unclear how manyof the side-effects were directly attributable to antidepressant therapy, although the overall levelof treatment withdrawals is remarkably similar tothat observed in RCTs (Hoto f et al., 1997). Nor is it possible to determine the number of remissions which would have taken place spontaneously. None the less thisstudy points to the importanceof proper supervision of antidepressants. Choice u ~ a ~ t i ~ ~ r e s s a ~ ~ s
SSRIs have theadvantage of greater safety in overdose and being marginally better toleratedthan tricyclics. It may, therefore, be tempting to conclude that these should be first-linet r e a ~ e n for t thisgroup of patients. The SSRIs are considerably more expensive, however, and the disadvantagesof the tricyclics may have been overstated (Hotopf et al. 1996a). Tricyclics also have analgesic properties which may be useful in many settings wherepain is hard to control, for example diabetic neuropathy (Max et al., 1992). We would recommend that tricyclics are used first line, except where there is a clear medical contraindication, high perceived suicide risk or they were not tolerated in a previous episode.We would not recommendthe use of monoa~ne-oxidase inhibitors inthe setting of medical illnessdue to their interactions with other drugs. Lithium is contraindicated orshould be used with caution in many illnesses (see Table 19.2). The usual indications forlithium apply. The specific conditionsin which antidepressants must be usedwith caution have been reviewed extensively (Cunningham, 1994; Series, 1991; Cavanaugh, 1991; Fava &: Sonino, 1996) and are shown in Table 19.2. In general terms the metabolism of antidepressants in the physically ill may be altered by poor renal or hepatic function, low plasma protein and pharmacokinetic drug interactions. Whilst some patients require full doses of antidepressantsit isprudent to use a lower starting dose and build it up more gradually. Cavanau h(1991)recommends taking tricyclic levels5 days after the patient has been maintainedon the lower therapeutic dose.
The traditional importance of ECT in the management of depression in the physically ill may have diminished, as theadvent of the SSRIs has allowed safe treatmentof depression in patients with heart disease. EGT is usefulin certain situations where there is urgency to treat the depression quickly or in
a,
v
v)
h
2
5
5 U
M)
hl
L,
cn U]
m v)
a;
~
~19:~ ea ~ ~ e ~s sand i o~ n~ ~ ~ eillness s i ~~ a l le 19.2 Physical illnessand physical treatmentof depression. Disease ~ a r d i o v a s c udisease: l~~
Recent myocardial infarction Heart block Congestive cardiacfailure
Aortic or cerebral aneurysm Hypertension
Eye disease: G l a u c o ~ a
Tricyclics contraindicated Tricyclics contraindicated MAOIs contraindicated Tricyclics: risk postural hypotension-avoid Lithium excretion lowered byACE inhibitors and diuretics ECT contraindicated Lithium excretion reduced by diuretics Neurotoxicity of lithium with methyldopa MlAOIs enhance antihypertensives Tricyclics contraindicated
Geni~o-urinarydiseases
Prostatic hypertrophy Renal failure
Tricyclics worsen symptoms-risk of retention of urine Risk of toxicity from tricyclics and lithium
~eurologicaldisease
Epilepsy
Intracranial spaceoccupying lesion Raised intracranial pressure Cerebrovascular accident Parkinson’s disease Migraine Liver failure Gast~o-in~estinal diseuse
Upper GI tract disease Lower GI tract disease Severe diarrhoea
All antidepressants lower seizure threshold Maprotiline contraindicated Interactions betweenSSRIs and anticonvulsants (raised levels of phenytoin, carbamazepine) Avoid carbamazepine withMAOIS ECT Contraindicated ECT contraindicated ECT potentially dangerous MlAOIs contraindicated Interaction betweenfluoxetine and selegiline (confusional state) Interaction between lithium and sumatriptine (CNS toxicity) Decrease dose of all antidepressants If severe, tricyclicscontraindicated SS€& may worsen nausea
Tricyclic levels raised bycimetidine Tricyclics may worsen constipation Lithium contraindicated
Endocri~e disease
Addison’s syndrome Hypothyroidism Phaeochromocytoma Hyperthyroidism
Lithium contraindicated Lithium contraindicated MAOIs and moclobemide contraindicated Tranylcypromine and moclobemide contraindicated
~ l o o diso orders
Agranulocytosis Sickle celldisease
Tricyclics and mianserin contraindicated Caution withECT, anaestheticrisk
Others: Porphyria
Avoid tricyclics
resistant depression.ECT is contraindicatedin patients with a recent cerebral bleed, raised intracranial pressure, brain tumour or recent ventricular dysrhythmia.There is some evidence to suggest that EGT is safeand effective even in poststroke depression (Murray et al.,1986),which had previously been considered a contraindication for EGT. Unfortunately this report involved small numbers and one would be reluctant to make generalizations of the safety of ECT in this setting. A thorough anaesthetic assessment is essential patients with physical illness. before EGT is given to
Psycholo~icalinterventions have been widely and successfully used in liaison psychiatry for the treatmentof functional syndromessuch as chronic fatigue syndromeand irritable bowelsyndrome (Guthrie et al., 1993; Sharpe et al., 1996; Deale et al., 1997). However, theliterature is less clear for those with a defined physical illness. The majorityof psychotherapeutic treatments which have been evaluated in patients with physical illness have aimed to improve a d j u s ~ e nto t the illness and thereby improve quality of life. These studies have focused on outcomes such as behaviours, complianceand psychological adjustment,and might be considered as preventive trials in that most of the subjectsdid not suffer depressionat the outset. Guthrie (1996) has reviewed the literature on psychotherapy in chronic illness and comments that relatively brief, focusedinterventions have been shown to have long-lasting efficacyserious in illnesses if they are given early. Thus Thompsonand Meddis (1990) showed reduced anxietyand better adjustment when men were given a brief packageof counselling followingheart attacks. This study used coronary care nurses, which impliesthat the intervention could be widely available. Other studies in cancer (Burton et al., 1991; Greer et al.,1992; Forester et al., 1993)have shown similar benefits again with relatively brief focused psychotherapies; further studies in cancer are reviewed elsewhere (Fawzyet al., 1995).
A range of different psychotherapies, discussed in greater detail elsewhere in this book, may be useful patients in with physical illnessand depression. As we have emphasized elsewhere in this chapter, it isimportant not to dismiss the symptomsof depression in this context as simply 'understandable' (especially in the elderly) and therefore notworthy of treatment. Many of the preventive trials emphasize the importance of relatively simple interventions. Psychoeducation is especially important, given the bewilderment many patients feel for their illnesses. It isworthwhile explaining to patients how
s occurand,whereapriate,todescribestrate symptoms. These ight involve giving thema rational approachto selfas relaxationor medication or usisimplebehaviouraltechniques,such assessed in primary c
ch of the emphasis of problem-solvi~gis on edom, which are common conse~uences of ~hysicalillness. ~o~itive-behavioural therapy has been assessed in two ndomi~edcontrolled trials (RCTs), described below. As in other cases of sion the emphasis is on understanding mood terms of i n ~ o l ~ t a r y patterns. In this sett in^ the thera~istwoul a u t o ~ a t i cnegative thoughts related illness. For example, a heart attack many individ r heart tobe a battery in power andhave auto ts such as: ‘if T do too much I will have another heart found behavioural consequence ance of activities and ption of maladap~ve a sick role. ould aim to challenge d replace them with more lated to their illness. amic ~terventionshav this setting. They are f most use inpatient are p r o ~ i n ~ nand t, nt to the physical illness ful of studies of ~sychothera~y in the specific contextof deysical illness have been identified/ and these are reviewed
ent maintainedat 6 months. andomized 68 depressed men with HIV infection to upf a social support group or a no-treatbrief (8 weeks) and took placein small ngerous b e h ~ v i ~ u r s
Larcombe (19 and Wilson nitive-behavioural list controls for depressed patients with mult moderately to severely depressed on the Be The t r e a ~ e nconsisted t of 90 min, small for 6 weeks. The therapy aimed to increase social interac itive model of depression. Despite small numbers (only 20 ey were able demonstrate to impressive benefitsof treatment DI score, which appeared to be maintained at ~terpersonaltherapy for epression in human immunodeficiency vi (HIV) positive individ~alshas been assessed in an uncontrolled pilotS arkowitz et al. (1992). therapy tookplaceover 16 and used a mixture of psyc ucation and behavioural stra in-depth discussion of emrelatedto grief and interpersonal situations. Within the limits of the de is intervention appears to have been successful with resolution of depression in20 of the 23 patients. a clear ideaof the role of psychoWith so few studies it is difficult to have therapy in the treatment of ession in the setting of chronic illness. The results of some o research are none the less encou relatively brief, focused therapies could make a clinically meaningfuldiffe~ence to depression and quality of life in physically-ill patients, they might be the treatment of choice as they would avoid possible harmful side-ef or interactions and have the theoretical advantage of bringing abou term change. Without more complete assessment of these thera would be difficult to justify their widespread provision.
y parts of the psychological inte~entionsdescribed above is to social consequencesof physical illness and losses patients are faced with are social ones, which soc ameliorate.A. detailed social history and referral toa social tional therapist is important in any chronic medical illn involves substantial physical disability. These measures h ated specifically in this setting.
in patientswith physical illness is ne the less it is frequently not detecte reduci sequences of this are likely to be grave inofterms worsening prognosis for both illnesses. There is some evidence antidethat
pressants are useful in a wide varietyof physical illnesses. There is rather less evidence for psychological interventions, however, from the evidence available, we suspect that focused treatments which relya mixture on of supportive, behavioural, educational and cognitive elements are likely be helpful. to Future research should concentrate on the role of simple psychotherapeu~c interventions in the patients with physical illness and depression and should concentrate on long-term outcomes such as survival and quality of life. Table 19.3 Key points. Depression is common in all physical illnesses. Depression frequently goes undiagnosed among patients with physical illness, especially in hospital settings. Depression worsens qualityof life and possibly survival for some physical illnesses. Special emphasis should be placed on symptomssuch as guilt, low self-esteem and worthlessness in the diagnosis of depression in patients withphysical illness. Antidepressants are more effective than placebo, although there is no clear evidence that any one class of antidepressants is superior. Choice of antidepressants dependson specificcontraindications of the physical illness. Psychotherapies have not been extensively evaluated in this setting, but there is some evidence to suggest brief structured psychotherapies are effective.
Ades, P.A., Waldmann, M.L., McCann, W.J. & Caine, E.D., Lyness, J.M.& King, D.A. (1993) ReWeaver, S.O. (1992) Predictors of cardiac rehaconsidering depression in the elderly.A ~ e r ~ c n n ~ o ~ r nofa Geriatric l Psychiut~, 1 ,4-20. bilitation participation in older coronary patients. Archives of ~nternalMedicine, 152, 1033- Case, R.B., Moss, A.J., Case,N., McDermott, M. & Eberly, S. (1992) Living alone after myocardial 1035. infarction: impact on prognosis. ~ournalof the American PsychiatricAssociation (1994)DiagnosA~erican Medical Association, 267,515-519. tic and S~a~istical Manual of Mental ~isorders: Cavanaugh, S.V.A. (1991)Depression in the mediDSM W , TV edn. APA, Washington DC. cally ill. In:N a n ~ b o ofo ~Stu~ieson General NosAnderson, J., Aabro, E., Gulmann, N., Hjelmsted, pita2 P s y c k i a t ~ (eds.Judd, Burrows and Lipsitt), A. & Pederson, H.E. (1980) Anti-depressive pp. 281-303. B.V. Elsevier, Amsterdam. treatment in Parkinson‘s disease. Actn PsychiCosta, D., Mogos, I. & Toma, T. (1985) Efficacyand atrica Scandinavica/62,210-219. safety of mianserin in thetreatment of depresBorson, S., McDonald, G.J., Gayle,T., Deffebach, sion of women with cancer. Acta Psyckiatr~ca M., L a k s M a r a y a n , S. & VanTuinen, C. (1992) Scandinavica,72 (Suppl. 320). 85-92. Improvementin mood, physical symptom, and Cu-gs, J.L. (1992) Depression and Parkinson’s function with nortriptyline in patients with of PsyckiDisease: a review. America~ ~ournal chronic obstructive airways disease. Psychosomatics, 33,190-201. at% 1491 443-454. Bridges, K.W. & Goldberg, D.P. (1984)Psychiatric Cunningham, L.A. (1994)Depression in the medically ill choosing an antidepressant.~ o u r n oaf~ illnessin inpatients with neurolo~caldisorders: Clinicul Ps y c h ia t ~/ 55,90-97. patients’views on discussionof emotionalprobo u r ~ aG.W. ~ , & Roose, S.P. (1990)Perspectives on l e m with neurologists. ~ r i t i s h M e ~ i c a ~ ~Dalack, the relationshipbetween cardiovasculardisease 289,654458. l Psychiand affective disorder. ~ o u ~of a~Zinical Burton, M.V., Parker, R.W. & Wollner, J.M. (1991) The psychotherapeutic value of a ’chat’: a atry, 51 (Suppl. 7),4-9* verbal response modes study of a placebo Deale, A., Chalder, T., Marks, I. & Wessely, S. (1997) A randomised controlled trial of cognitive attention control with breast cancer patients. behavior therapy for chronic fatigue synPsychothera~yResearch, 1,39-61.
drome. American Journal of Psychiatry, 154, 40%-414. Derogatis, L.R. (1975)The SCL-90. J o h s Hopkins University, Baltimore. Endicott, J. (1984) Measurement of depression in patients with cancer. Cancer, 53 (Suppl.), 22432248. Evans, R.L. & Connis, R.T.(1995) Comparison of brief group therapies for depressed cancer patients receiving radiation treatment. Public Health R ~ o r t s11,306-311. , Fam, J.R. & Tucker, G.J. (1995) Mood disorders with general medical conditions. Current Opinion in Psychiatry,8,13-18. Fava, G.A. & Sonino, N. (1996) Depression associated with medical illness. Drug Therapy, 5, 175-189. Fawzy, F.I., Fawzy, N.W., Arndt, L.A. & Pasnall, R.O. (1995) Critical review of psychosocial interventions in cancer care. ~ r c h i ~ofe sGeneral Psychiatry, 52,100-113. Forester, B., Kornfeld, D.S., Fleiss, J.L. & Thompson, S. (1993)Group psychotherapy during radiotherapy: effects on emotional and physical distress. American Journal of Psychiatry, 150, 1700-1706. Frasure-Smith, N., Lesperance, F. & Talajic, M. (1993) Depression following myocardial infarction: impact on 6-month survival. Journal of the American MedicalAssociation,270,1819-1825. Godding, P.R., McAnulty, R.D., Wittrock, D.A., Britt, D.M. & Khansur, T. (1995) Predictors of depression among male cancer patients. Journal of Nervous Mental Disorders, ~$3~95-98. Goldberg, D. (1972)The Detection of PsychiatricIllness by Questionnaire.Oxford University Press, London. Greer, S. (1991) Psychological response to cancer and survival. PsychologicalMedicine, 21,43-49. Greer, S., Moorey, S., Baruch, J.D.R. et al. (1992) Adjuvant psychological therapy for patients with cancer: a prospective randomised trial. British Medical J o u ~ a l304,675480. , Guthrie, E. (1996) Emotional disorder in chronic illness: psychotherapeutic interventions. British Journal of Psychiatry, 168,265-273. Guthrie, E., Creed, F., Dawson, D. & Tomenson, B. (1993)A randomised controlled trial of psychotherapy in patients with refractory irritable bowel syndrome. British Journal of Psychiatry, 163, 315-321. Harrison, J. & Maguire, P. (1994)Predictors of psysh chiatricmorbidity in cancerpatients. ~ ~ t iJournal of Psychiatry,165 (5),593-598. Hotopf, M,, Hardy, R. & Lewis, G. (submitted) Discontinuation rates of SSRIs and tricyclic anti-
depressants: a meta-analysis and investigation of heterogeneity. British Journal of Psychiatry, xyo, 120-127. & Norrnand, C. (1996a) Are Hotopf,M.H., Lewis, G. SSRIs a cost effective alternative to tricyclics? British Journal of Ps y c h ia t ~, 168,404-409. Hotopf, M.H., Noah, N. & Wessely, S. (1996b) Chronic fatigue and psychiatric morbidity folJourlowing viral meningitis: a controlled study. nal of Neurology, Neurosurgery and Psychiatry, 6, 504-509, House, A. (1988) Mood disorders in the physically ill-problems of definition and measurement. Journal of PsychosomaticResearch, 32,345-353. Huyse, F.J., Zwaan, W.A. & Kupka, R. (1994) The applicabilityof antidepressant in the depressed medically ill: an open clinical trial with fluoxetine. J o u ~ aofl Psychoso~aticResearch, 38, 695702. Jeffcoate, W.J., Silverstone, J.T., Edwards, C.R.W. & Besser, G.M. (1979) Psychiatricmanifestations of Cushing’s Syndrome: response to lowering of Medicine, of plasma cortisol. Quarterly~ourna2 191,465-472. Katon, W. & Raskind, M. (1980)Treatment of depression in the medically ill elderly with methylphenidate.American Journal of Psychiat~, 1371 963-965. Katon, W. & Sullivan,M.D. (1990)Depression and chronic medical illness. Journal of Clinical Psychiatry, 51, 3-11. Kelly, J.A., Murphy, D.A., Bahr, G.R. et al. (1993) Outcome of cognitive-behaviouraland support group brief therapies for depressed, HIV infected persons. American Journal of Psychiat~, 150, 1679-1686. Ladwig, K.H., Roll, G.,Breithardt, G., Budde, T. & Borggrefe, M. (1994)Post-infarction depression and incomplete recovery 6 months after acute myocardial infarction. Lancet, 343, 20-23. & Frengly,J.D. (1986) Lakshmanan, M,, Mion, L.C. Effective low dose tricyclic antidepressant treatment for depressed geriatric rehabilitation patients: a double-blind study. Journal of the American Geriatric Association, 34,421-426. Larcombe, N.A. & Wilson, P.H. (1984) An evaluation of cognitive-behaviour therapy for depression in patients with multiple sclerosis. British Journal of Psychiatry, 145, 366-371. Levenson, J.L., Hamer, R.M. & Rossiter, L.F. (1992) A randomized controlled study of psychiatric consultations by screening general medical inpatients. American Journal of Psychiatry, 149, 631-637. J., Araya, R. & Dum, G. (1992) Lewis, G., Pelosi, A. Measuring psychiatric disorder in the commu-
nity: a standardised assessment for lay inter~ n ~ e r n a l ~ e d i~50,1083-1088. cine, viewers. Psychological~ e d i c i n e22,465-486. , Popkin, M.K., Callies, A.L. & Mackenzie, T.B. Lipsey, J.R., Robinson, R.G., Pearlson, G.D., Rao, (1985) The outcome of antidepressant use in the Price, T.R. (1984) Nortr~ptyline treatment medically ill. Archives of General Psychiatry, st-stroke depression: a double blind study. 1160-1163. Lancet, X,297-300. Rabkin, J.G., Rabkin, R., Harrison,W. & Wagner, Lishman, W.A. (1998) OrganicPsychiatry,3rd edn. G. (1994) Effect of imipramine on mood and Blackwell Science, Oxford. enumerative measuresof immune status indess, J.M., Caine,ED., Conwell, Y., King, D.A. pressed patients with HIV disease. A ~ e r i c a n Cox, C. (1993) Depressive symptoms, mediJournal of Psychiatry, X ~ X 516-523. , cal illness, and functional status in depressed Reding, MJ., Orto, L.A., Winter, S.W., Fortuna, .A ~ e r i c a nJournal of PsyI.M., Di Ponte,P. & McDowelI, EH. (1986) Antidepressant therapy after stroke. Archives of D.L., Hawton, K.E. & ~ e u r o ~43,o ~ 763-765. , Bancroft,J.H.J. (1974) Psychiatric morbidity and Rifkin, A., Reardon, G., Siris, S. et al. (1985) referral on two general medical wards.~ r i t ~ s h Trimipramine in physical illness with depression. Jou~nalof Cl~nicalP s y c h i a ~46,4-8. ~, ~ e d i c a ~ J o 1,268-270. ~rna~, alzberg, B. (1937) Mortality among patients with Robertson, M.M.& Trimble, M.R. (1985) The treatinvolution melancholia. ment of depression in patients with epilepsy: A ~ e r i c a ~ J o ofu Psyr~al a double blind trial. J o ~ r n oaf ~A ~ e c ~~ ~~ s~o e r ~ ~ s , ,G.L. &Perry,S.W. (1992) 9, 127-136. ~terpersonalpsychotherapyof depressed HIV Robin, L.N., Helzer, J.E., Croughan, J. & Ratcliff, ~ o s p i t aC l o ~ ~ uPsyn i ~ K.S. (1981) National Institute of Mental Health Diagnostic Interview Schedule; its &$tory, characteristics and validity. ,M.L. & Mirabi, M. (1981) Archives of General PsyPhysical symptoms of depression.~ r i t i s hJourchiatry, 38,381-389. Robinson, R.G., Morris, P.L.P. & Federoff,J.P. (1990) Depression and cerebrovascular disease. Journal of C~inicalPs y c h ia t ~, order in the general hospital. ~ r i t i s hJournal of 51 (Suppl. 7), 26-31. Robinson, R.G. & Price, T.R. (1982) Post-stroke Psychiatryf149, 172-190. depressive disorder: a follow-up study of 103 Max, M.B., Lynch, S.A., Muir, J., Shoaf, S.E., patients. Stroke, ~3,635-641. Smoller,B. &: Dubner,R. (1992) Effects of desipramine, amitriptyline, and fluoxetine on Rodin, G.M., Chmara, J. &: Ennis, J. (1981) Stopping life-sustaining medical treatment: psychipain in diabetic neuropathy. New EngZan~ Jouratric considerations in the termination of renal dialysis. CanadianJ o u r ~aof l Psych~atryf 26,540McDaniel, J.S.,Musselman, D.L., Porter, M.R., Reed, D.A. &:Nemeroff, C.B. (1995) Depression 544‘ in patients with cancer: diagnosis, biology andRoose, S.P., Glassman, A.H., Giardina, E.G.V., Johnson, L.L., Walsh, B.T. & Bigger, J.T.(1987) Cartreatment. Archives of General Ps~chiatry, diovascular effectsof ~ i p r aand ~ ebupropion 89-99. in depressed patients with congestive cardiac Morris, P.L.P., Robinson, R.G.& Samuels, J. (1993) Depression, introversion and mortality follow- failure. Journal of Clinical P s y c ~ o p h a r ~ a c o l o ~ , ing stroke.Australianand New~ e a l a n d J o ~ rofn a l 7, 24-7-251 Roose, S.P., Glassman, A.H., Attia, E.& Woodring, Psychia~r~f 27, 443-449. S. (1994)Comparative efficacy of selective ~ u r r a yG.B., , Shea, V. & Conn, D.K. (1986) Elecserotonin reuptake inhibitors and tricyclics in troconvulsive therapy for poststroke depresthe treatmentof melancholia. A ~ ~ i c a n J oof u ~ a l sion. J~urnalof Clinical Psychiatry,47, 258-260. L.M., Gath, G.H.,Lloyd-Tho~as, P s y c ~ ~15a1,~173y”-739. ~, inson, D. (1995) Randomised con- Schifano, F., Garbin, A., Renesto,V. et al. (1990) A double-blind comparison of mianserin and trolled trial comparing problem solving treatmaprotiline in depressed medically ill elderly ment with amitriptyline and placebo for major people. Acta Psychiat~caScandinavica,81,289-294. depression in primary care. ~ r i t i s h ~ e dJourical Schiffer, R.B.& W ~ e m a nN.M. , (1990) Antidepresnal, 310,441-445. sant pharmacotherapyof depression associated P4rez-Stable, E. J.,Miranda,J., Munoz, R.F. & Ying, with multiple sclerosis. A ~ e r i c a n J o uofrPsy~a~ Y.W. (1990) Depressionin medical outpatients: ~nderrecognitionand misdiagnosis.Arch~vesof chiatry, 147,1493-1497. *
Schubert, D.S.P., Burns, R.,Paras, W. & Sioson, E. Thompson, D.R. & Meddis, R. (1990) A prospective evaluationof in-patient counselling for first (1992) Increase of medical hospital length of time myocardial infarctionin men. ~ o ~ r ofn u ~ stay by depression in stroke and amputation Psychosomutic Research, 34, 237-248. patients: a pilot study. Psychotherupy and Van Hemert, A.M., Hengeveld, M.W., Bolk, J.H., ~ s ~ c h o s o ~ u59, t ~61-66. cs Rooijmans, H.G.M.& Vandenbroucke, J.P. (1993) Schwartz, J.A., Speed, N.M., Brunberg, J.A., Psychiatric disorder in relation tomedical illBrewer, T.L., Brown, M. & Greden, J.F. (1993) ness among patientsof a general medical outDepression in stroke rehabilitation. Biologicul patient clinic,Psychoiogical Medicine,23,167-173. 331 694-699. Psychiut~, Series, H.G.(1991) Drug treatmentsof depression Veith, R.C.,Raskind, MA., Caldwell, J.H.,Barnes, in medically illpatients.Journal of Psychosomatic R.F., Gumbretch, G. & Ritchie, J.L. (1982) CarResearch, 36,1-16. diovascular effects of tricyclic antidepressants in depressed patients with chronic heart disease. Sharpe, M., Hawton, K.,House, A. et al. (1990) New Eng2 and Journal of Medicine, 306,954959. Mood disorders in long-term survivors of stroke: associations with brainlesion locationand vol- Wessely, S., Chalder, T., Hirsch, S., Pawlikowska, T., Wallace, P. & Wright, D.J.M. (1995) Postume. Psycholo~culMedicine, 2,815-828. infectious fatigue: prospective cohortstudy in Sharpe, M., Hawton, K.,Simkin, S. et al. (1996) Cognitive therapy for chronic fatigue syndrome:primary care. Lancet, 345, 1333-1338. Newton, R.W., Lind, a randomized controlled trial. British M e d ~ c u ~ Wilkinson, G., Borsey, D.Q., disorder Journul, 3~2~22-26. C. & Ballinger, C.B. (1987) Psychiatric Silverstone, P.H. (1990)Depression increases mor- in patients with insulin dependent diabetes tality and morbidity in acute life-threatening mellitus attending a general hospitalclinic: (i) J o u ~ aofl Psychoso~atic Research, two stage screening and (ii) detectionby physimedical illness. cians. PsychologicalMedicine, 17,515-517. 341 651-657Silverstone, P.H. (1996) Prevalence of psychiatric Wing, J.K., Cooper, J.E. & Sartorius, N.(1974) The of Nervof psychia~ricsymp~ e a s u r ~ eand n t c~uss~cation disorders inmedical inpatients.J o u ~ u Z toms, Cambridge: Cambridge ous and ent tal Diseuse, ~84~43-51. University Press. Tan, R.S.H., Barlow, R.J., Abel, C.et al. (1994)The Zigmond, A.S.& Snaith, R.P. (1983) The Hospital effect of low dose lofepramine in depressed eld- Anxiety and Depression Scale. Acta Ps~chiutr~cu erly patients in general medical wards. British Scundinavica, 69,361-370. Journal of Clinical P h u r ~ u c o l o37, ~ , 321-324.
The aim of this final chapter is to make some practical suggestions for the assessment and management of patients with resistant depression. Preceding chapters have summarizedwhat is knownabout the treatmentof depression in the differentsettings and age groups in which it presents. Depressed patients who do not respond to the standard treatment as described in these chapters are frequently referred for a second opinion with a diagnosis of resistant depression.
The term resistant depression is used when a patient has made an inadequate clinical responseto an adequate course of antidepressant treatment. The definition of what comprises an adequate course of treatment with, for example, electroconvulsive therapy (ECT), antidepressant medication or cognitive therapy is more difficultand will be discussed in this chapter. It will be seen that the assessment of the adequacy of prior treatment is one of the most important aspects of the assessmentof patients with resistant depression.The assessment of the adequacy of the response to each treatment will also be discussed. Although common usage refers to resistant depression as if it represented a distinct and exclusively biologicalvariant of depression, there isno scientific basis for this view. The genetic studies which have been reviewed in Chapter 3 have demonstrated an environmental contribution to all forms of depression, and the environmental modelof depression which has been described in Chapter 4 explains most of the variance in onset and much of the variance in rates of recovery in most formsof non-psychotic unipolar depression. This chapter will describe the assessment of the biological and psychosocial contributions to any case of resistant depression from which assessments follow the need for physical and psychosocial interventions. The order to be followedin this chapter will that be followed in the assessment and treatment of a patient with resistant depression. The first assessment is of the diagnosisand time courseof the depressive illness and of any
associated (comorbid) mental or physical illness. The second assessment of is the respective contributionsof biological and psychosocial factors to the onset and chronicityof the illnessin question: discussionof psychological and psychosocial inte~entionin patients with resistant depression follows logically from discussionof the psychosocial causes. The third assessmentis of the adequacyof prior physical treatments and the clinical response to them. The fourth section will describe the use of physical treatments in patients with resistant depression: most of this sectionwill concern the physical treatment of resistant unipolar depression, but the special needs of patients with treatment resistant bipolar, winter, brief recurrent and psychotic depressions will alsobe discussed.
Although it is usually a simple matter to confirm a diagnosis of major depression in a patient referred with resistant depression, it is much more difficult to evaluate the time course of the illness which may extend back over several decades. The historiesof such illness provide a wealth of information about the effects upon the illness of psychosocial stress on the one hand and physical and psychosocial treatments on the other. Help is needed from independant informants andpast medical recordsif an accurate history is to be obtained, Often it is helpful to plot the courseof the illness over time as illustratedin the chart in this chapter (Fig.20.1)
Although the diagnosis of depression is rarely altered in patients referred with a diagnosis of resistant depression, frequently a second (comorbid) diagnosis is made.
The distinction between major depression and dysthymia is important because in the caseof depression higher expectations of the benefit of treatment can be raised than with dysthymia. The distinction between dysthymia and depression is not madeby most patients,but with the aid of operational criteit can be ria suchas DSM-IV (American Psychiatric Association (APA), 1994) made with reasonable relia~ility(Holzer et al., 1996). Both illnesses respond to antidepressant medication (Thase et al., 1996) andso it is necessary to evaluate the effectsof past treatment on the two disorders separately.
TAD SSRl CBT
ANlA
Removal of adrenal tumour Severe
Moderate 1975 Mild 1970
I
YEAR
1965
Mild Moderate Divorce
Remarried
VEk I
Chronic difficulties
Severe
emigration
Cushing’s Syndrome
Fig. 20.1 A typical life chartof the courseof a chronic depressive illness which shows the
effects of treatment with tricyclic antidepressants (TADs), selective serotonin re-uptake inhibitors (SSNs),cognitive behaviour therapy(CBT) and removal of an adrenaltumour, Also shown are the timing of a numberof severe life eventsand chronic difficultiesand the courseof this patient’s Cushing’s Syndrome.
Confusion between personality disorder and depression should not arise if the episodes of depression are carefully charted andif the diagnosis of personality disorderis based on enduring personality characteristics as assessed in between episodes of depression. There is some evidence from controlled studies that fhe coexistence of personality disorder with depression results in poorer response to antidepressant medication(Sato ef al., 1993)~to interper-
sonal psychotherapy (IPT) (Shea et al., 1990,1992) and to antidepressant medication with IPT (Frank &:Kupfer, 1990). The presenceof personality disorder may not effect the response of depression to cognitive behaviour therapy (CBT) (Shea et al., 1990) and depressed patients with dysfunctional attitudes have a better response to pharmacotherapy combined with CBT than depressed patients withoutdysf~ctional attitudes (Milleret al., 1990; Whismanet al., 1991). A particular CBT package has been developed for patients with borderline personality disorder (Linehan, 1993) and the use of cognitive techniques in the managementof patients with personality disorder has been described by Thase and Rush (1995). The importance of measuring objectively depressive symptoms in patients with comorbid personality disorder is emphasized by the finding that although a second diagnosis of personality disordkr predicted a poorer response to pharmacotherapy and IPT when social outcome was measured, very similar reductions in objective measures of depressive symptoms were found in depressed patients with and without personality disorder (Shea et al., 1990).
When a patient with resistant depression also has features of an obsessive compulsive disorder, then the time course of the two syndromes needs tobe distinguished. The possibilities include: 3. depression secondary to obsessive compulsive disorder; 2 obsessive compulsive symptoms developing de novo in a depressed patient; 3 obsessivecompulsive traits exacerbated by depression(andsometimes ameliorated by hypomania). Once the two disorders have been separated it is then possible to target physical and psychological treatments towards each. It will be recalled from that selective serotonin re-uptake inhibitors (SSRIs) have a parChapter 10 ticular roleto play in the t r e a ~ e nof t patients with depressive and obsessive compulsive symptoms (Goodmanet al., 1992).
As was the case with comorbid obsessive compulsive disorder, it is the same with comorbid anxiety disorder and panic disorder; and it is necessary to chart the time course of the different disorders. Cognitive therapy is particularly effective against panic attacks andboth the SSRIs and the monoamine oxidase inhibitors (MAOIs) can be used to treat most formsof anxiety states. But before planning such treatments for comorbid disorders it is essential to
433
have a good baseline assessmentof the severity and the time course of the different disorders which requiresimultaneous t r e a ~ e n t .
ubstance abuse is associated with resistant depression ( ~ o s e n b et a ~al., 1995) and a systematic historyof drug and alcohol intake will reveal undetected casesof substance abuse in some patients with resistant depression.Selfmedication with alcohol and/ or benzodiazepines is not uncommon in resistant depression and both tolerance and dependence are seen.An important ssessment of such depressions is an attempt to withdraw the ances, if necessary with the patient as an in-patient, while trying to teach alternative coping strategies.Mrhen depression develops in the presence of substance abuse then the treatment of the substance abuse the is first part of the assessment: substance related mood disorders in DSM-IV are those which resolve spontaneously within I month of abstinence. Antide~ressant medication has some therapeutic effect in recently abstinent alcoholics ason Br Kocsis, 1991) in non-abstinent alcoholics (McGrath et al., 1996) and in opiateaddicts maintained on methadone (Nunes et al., 1991).
epression is associated with medical illnesses such as carcinoma, myxoed and diabetes, therefore screening for these disorders is part of the evaluation of any patient with resistant depression.As with other comorbid disorders, the need is to distinguish between the time course of episodes of affective disorder and the time course of episodes of medical illness(Fig. 20.1). Any debilitating illness or one which imposes physical handicap is a risk factor for chronic depression (Prince et al., 1992).
epression is a well-known side-effect of hypotensive agents which act by i ~ b i t i n noradrenergic g neurotrans~ssion(Goodwin & Bimey, 1971; Paykel et al., 1982)but is also seen in patients treated with calcium channel blockers (Pascualy Br Veith, 1989), steroid hormones (Lewis & Smith, 1986) and H2 blockers (Billingset al., 1981; Billings& Stein, 1986). Whenone of these drugs is prescribed for a patient with resistant depression then an alternativeshould be sought.
43
Previously undetected cognitive decline is detected frequently when elderly patients with resistant depression are investigated. Cognitive decline by itself is not a cause of ~eatment resistance (Reynoldset al., 1987; Reifleret al., 1989; ~toudemireet al., 1991). When cognitive decline is suspected in a case of resistant depression, ~ e m o r ytesting is necessary to measure the baseline cognitive state which may deteriorate as a result of subsequent EGT. differential diagnosis of pseudodementia is well described by Lishman (Lishman, 1987).
The assessmentof the causes of any depressive illness involves the separate but related assessment of possible genetic, biological, medical and psychosocial factors. The evidence for the genetic aetiology of depression has been reviewed earlier and varies considerably with depressive subtype. In the case of bipolar illness as much as 0% of the variance in lifetime risk is due to genetic cGuffin Sr Katz, 1989), andin multiply affected families the genetic f lifetime riskis even greater. In the case of recurrent unipolardepression as seen in secondary and tertiary care, the genetic influence may ~ 1996) andin milder S explain 50% of lifetime risk ( ~ c G u f&rf Katz, cases of depression, as typically seenin the communit~the genetic contribu50% (Kendler et al., iggz). Thus deprestion to lifetime risk is much less than sive subtype (bipolar and unipolar), course of illness (sporadic or recurrent) and family history are important evidence in the assessment of genetic aetiology. As well as genetic evidence, other clinical evidence in favour of a ~iological (as opposed to exclusively psychosocial) aetiology of a given depression includes regular periodicity of the course of the illness, presence of the features of melancholia and severity of depression. Each of these indicate that neurobiological changes in sleep and neuroendocrine function are is possible. likely (Chapter 3) and that response to physical treatment Important ~ e ~ i ccauses al of depression include myxoedema (Cleare et al., 1996), Cushing’s Syndrome (Kelly et al., 1983)’ hypercalcaemia, carcinoma and any debilitating or painful chronic disorder (Prince et al., 1996). The careful screeningof patients with resistant depression regularly identifies patientsin which such medical conditions have been overlooked. Whether or not the onset of any individual episode of depression is influc i a l is best determined by assessment of the enced by ~ s ~ c ~ o s o factors psychosocial adversity and vulnerability as described in Chapter 4 of this
ITV
CHlL
~arital dj~iculties
Fostered then a d o p t ~ dfrom birth
Suicide of one son Emotional neglect
migration of another son
L
Chronic subclinical depression Low self-esteem
T IN
Lack of supp~rtive networks ~ u ~ t i pcarers le in childhood
Lack of con~dants
Fig. 20.2 A schematic summary of the psychosocial formulation of a depressed patient. The headings printed in bold type are assessed for all patients. The information in italics refers to the individual patient who is the same individual as in Figure20.1.
volume. The author Is depressed patients are routinely assessed as in Fig.20.2, using the LEDS (Brown Br Harris, 1978) for research purposes, with routine clinical enquiry for clinical assessment. There is evidence that psychosocial adversity and~ul~erability explain most of the variance in onset of depression and much of the variance in rates of recovery (Brown Br Brown et al., 1994.b). However, for recurrences of bipolar and psychotic/ melancholic episodes there is no such evidence (Brown et al., 1994a). With the exception of recurrences (not inceptions) of bipolar and psychotic depression, it can be assumed that when psychosocial adversity (see Chapter 4), precedes an episode of depression then a causal relationship is likely. Furthermore, if such adversity continues while a patient remains chronically depressed, then it is likely that psychosocial factors are contributing to chronicity of illness.
Most of the psychosocial triggers for the onset of depression also serve to maintain a depressive illness (Chapter 4). These factors include persistent chronic difficulties (such as long-term marital discord) and additional life events involving loss. Childhood adversity and currentoflack social support & Moran, 1994; Brown et al., 199413). Conversely also predict chronicity (Brown ’fresh start’ events, like a new relationshipor new job, which cancel out the et effect of a depressogenic eventor difficulty, predict good outcome (Brown al., 1992). There have been few detailed interview-based studies of the effects of these same psychosocial influences on response to medication(as measured in a placebo-controlled study),but in one such study social adversity (life events and chronicdi~icultie§) had a greater effect on outcome than the et al., 1994). Whether administration of the antidepressant desipramine (Lam ongoing social adversity and lack of social support interacts specifically in response to antidepressant medication (Lamet al., 1994) or whether it has a et al., 1994a) is a somewhat academic quesdirect effect on outcome (Brown tion: either way ongoing social adversity and lack of social support resultin resistant depression. Since there are interventions which can influence the psychosocial forces bearing upon depression, it is necessary to identify social adversity when present and to formulate ways in which its effects can be minimized.
Social support is a critical variable in determining whether social adversity results in onsetof depression: it also influences the chronicityof depression (Brugha et al., 1987). The elementof social support which has this protective effect is the presenceof a confiding relationship, i.e. a close friendship within which it is possible to disclose personal difficulties s and h o r t c o ~ n g(Brown s et al., 1986). For this reason it is helpful to see depressed patients with their spouses on one occasion to evaluate the communication that takes place between them. Sometimes the non-depressed spouse demonstrates a high level of expressed emotion (EE) and makes many critical remarks about the depressed patient who says little but c o ~ u ~ c a tpassively es (Hooley, 1986). High levels of EE in the spouse predict depressive relapse in the patient (Vaughan & Leff, 1976; Hooley et al., 1986): marital therapy can help such couples toc o ~ u n i c a ton e a more satisfactory manner. The spouses of chronically depressed patients are themselvesby definition exposed to a chronic difficulty and hence are vulnerable to depression. They suffer a restriction in social and leisure activities, a fall in family income and a strain on the marital r e l a ~ o n s ~‘negative’ p: behaviour is particularly hard for them to bear (Fadden et al., 1987). Family members often say that they need further info~mation be provi about the natureof the illness (Faddenet al., 1987) and this can
for example, in a single psychoeducational group meeting (Andersonet al., 1986). Divorce and separation are seen in as many as 50% of the spouses of bipolar depressives and so this is another reason for offering marital assessment and/or therapy (Brodietk Leff, 1971).A behavioural approach to marital therapy has been shown to increase marital satisfaction evenin the presence of significant depression (O’Leary &c each, 1990). When a patient with resistant depression has little in the wayof social support and in particular has no close confidant, then interventionsto change this state of affairs are warranted. Patients with chronic depression and marital difficulties can be helped significantlyby a combination of emotional and practical help given, for example,by a social worker (Corney, 1987). The effectivenessof a befriending scheme providedby trained lay volunteers has been described by Harris in Chapter 4. Finally, there are a number of excellent self-help and relative support groups which can provide information and contacts with UK,the Manic Depressionello ow ship and others in similar situations. In the the Seasonal Affective Disorder Association are particularly recommended.
Psychological interventions such cogniti~e-behavioural as therapy (GBT) can interrupt the mechanisms whereby social adversity influences the course of depressive illness. Thus, low self-esteem (~egativeEvaluation of the Self) which is a consequence of childhood adversityis also a vulnerability factor which increases the likelihood that social adversitybewill followed by onset and chronicity of depression (Brown et al., 1994a): conversely, high selfesteem (Positive Evaluation of the Self) predicts recovery from an establi§hed depression (Brownet al., 1994b). Cognitive therapy addresses these negative cognitions and also the depressive schemata from which they arise (Teasdale tk Barnard, 1993). The useof co~itive-behaviouraltherapy with pharmacotherapy for depressed in-patients with resistant depression has been describe by Scott (1988). Effective coping strategies have a protective effect against social adversity (see Chapter 4) and can be taught as partof a CBT package which includes problem-solving. They can also be taught to patients with bipolar illness who learn to recognize early signs of hypomania. The roleof cognitive therapy inthis situation is being evaluated at present (Chapter5).
nce rates can be as low as 50%indepresse
43
et al., 1965) and even lower in elderly patients (Salzman, 1995) the first investigation to make in anypatient with resistant depression is measurement of plasma drug concentrations, together with a pill count and an enquiry about complianceand side-effects. M e n investigation reveals total non-compliance it is necessary to find out the reasons. It may be that the patient has not received an adequate explanation of the wanted and unwanted effects of the drug, or there may have been troublesome side-effects prior to utic response; on occasions non-complianceanisexpresitions concerning their future and their treatinadequate, it isimportant to ensure good compliance for a trial period of 6 weeksand to measure any symptom reduction using an observer rating scale. Frequent measurementof plasma drug concentrations is needed, and with in-patients medication can be givenas syrup. Time spent on educating a patient in simple coping strategies (Cochran, 1984) can bring compliance rates up to 85% (Frank et al., 1995).
In the absence of clear dose-response relationships for mostantidepressant 11)the recommended doses of tricyclic drugs are arbitrary. rugs (see Chapter daily is the recommended maximum ( ~ ~ ~ t ~ Fors ~ reas in the USA 2.50 mg daily would be consideredan adequate doseof imipramine, amitrip~lineand chlomipramine, whereaszoo mg daily is considered adequate for desipramineand trimipramine: higher doses aregiven when acce atients(seeAmsterdam 8r Hor~g-Rohan, for further details). nship between dose and side-effects of S and other recently introduced antidepressants has been ~vestigatedmore carefully and for thesedr S similar rangesof dose and plasma drug concenn the USA, UK and the rest of Europe. In assessantide~ressantit is sensibleto ask firstwhether ose (and plasma drug level) has been followed tricyclic antidepressants)/ to ask whether the maximally tolerated doseof a tricyclichas been usedwith concurrent monitoring of plasma drug levels and electrocardiograph (ECG).
~ a t ~ o ~
There is now much evidence that the time needed to evaluate the efficacy of an antidepre~sant drugis 6 weeks, most of the variance in outcome being uring the first4-6 weeks of t r e a ~ e n (AHCPR t Depression Guideline report suggests that much longer treatment periods (of 10-16 weeks) may be needed in patients with personality difficulties and social 439
adversity (Frank8r:Kupfer, 1990); and in the case of fluoxetine there is some evidence that 8 weeks is needed for maximal antidepressant response (Schweizer et al., 1990). A recent meta-analysisof published clinical trials has concluded thatif there hadbeen no symptom reduction after 4 weeks of treatment then future treatment with the same drug would be of help, not whereas if a small reduction in symptoms was seen after4 weeks of treatment then 2 weeks of t r e a ~ e nwith t the same drug further improvement over the next was likely( ~ u i t k i et n al., 1996). In general, an adequate duration of t r e a ~ e n t can be considered tobe one which has been continued at optimal dose for 6 weeks.
The 'effect size' of any ~eatmentcan be defined as a changeby one standard deviation in the baseline measure (see Chapter 11). The effect size of treatment with tricyclic antidepressant and placebo have been estimated to be 1.5 and 0.8, respectively (Quality Assurance Project, 1983) which means that com pared with placebo treatmentan antidepressant drug canbe expected to reduce a depression rating by less than one standard deviation of the baseline measure. Fava and Davidson(1996) have reviewed a more recent cohortof clinical trials published between 1993 and 1995. They defined treatment resistance as a fallby less than 50% in the Hamilton Rating Scale after at least 6 weeks of treatment withan antidepressant given in the dose ranges recomto this definimended in the USA (see above): treatment resistance according tion was found in 29-30% of all studies sampled (Fava &: Davidson, 1996). From this it follows that patients with resistant depression can at least be expected to have received only partial antide~res~a~t bene~it from prior antidepressant treatment: they may be too epressed and too negative in their thinking to be able to recall which of the many antidepressant drugs produced a small reductionin depressive symptoms. When different treatment imes are given to patients with resistant depression it is essential to monitor treatment response using an objective measure of depression.
ince guidelines vary between countries thereno is absolute criterion for an dequate EGT stimulus. In the case of bilateral ECT a stimul~swhich is x threshold is certainly adequate (Sackeim et al., 1990): a right unilateral x thresholdstimulusisprobablyalsoadequate.practice of titrating T stimulus against duration of seizure (Fink, ) cannot guarantee that an adequate stimulus has been given (Robin&:De Tissera, 1982). Thisis articularly so in the case of unilateral EGT when a barely suprat~eshold
pulse can produce a seizure of ’adequate’ duration but of no therapeutic value (Sackeim et al., 1987). The titration ofECT stimulus against duration of seizure as recommended by the Royal College of Psychiatrists is, however, better than the use of arbitrary fixed low-dose stimulus which makes no allowance for variation in ECT threshold. The use of a fixed high-dose stimulus (Abrams et al., 1991) is also an adequate therapeutic strate though one which may cause unnecessary memory loss. There is even less empirical evidence to help the clinicianwhe evaluate h r previous coursesof EGT have been given for a sufficient length of t ferent authorities have recommendedECT thatbe discontinuedif no rnent has been seen after eight ( oyal Collegeof Psychiatrists, 1995) (APA, 1990)~12 (Abrams, 1992)and even 15 (Devanand et al., 1991) adequate seizures: the author’s practice follows the most cautious of these r e c o ~ e n dations. Although thereis doubt about the definition of an optimal courseof E there is sufficient consensus to be able to identify awhen patient with resistant depression has received inadequate ECT treatment. A treatment in. which none of the recommended methods for controlling the stimulus was use should be considered inadequate,as should a course which wasabandone on account of non-response beforeat least eightadequate seizures were duced. In view of the superior efficacy of bilateral ECT over unilateral (see Chapter 12),an unsuccessful courseof unilateral ECT should also be considered to be inadequatein a patient with resistant depression. Under all these circumstances a second course of ECT should be considered U placement of electrodes, stimulus dosing and recording of se measuring any symptomatic improvement with an observer after at least eight adequate seizures before concluding th ineffective.
In the lightof current evidence, the following treatment options can be considered in most patients with resistant depression.
Uncontrolled studies have shown that out-patients’ resistant to t antidepressant(TAD) t r e a t ~ e n have t a 30-70% chance of respondi or it is effective t change to an SSRI (Thase Ss Rush, 1995). Whether not the converse strategy of changing froman SSRI to a TAD in cases of t resistance isnot known.
se rates of over 50% have been reported in a number of controlled in which patients who had failed to respond to a tricyclic antidessant were changed to an MAO1 (Thase et al., 1992; McGrath et al., 1993). ical depression are particularly likely to benefit (~cGrath et wever, the usefulness of this inte~entionis limited by the need to ate medication-~ee’washout’ period which should be 3 weeks eks forSSIpls (other than fluoxetine) and 6 weeks for fluoxetine. not possible to keep treatment resistant depressed dm patients S way without causing a severe exacerbation of depression.
onse rates of around 50% have been reportedin a number of ies in which lithium treatment for at least4 weeks has been to a tricyclic or heterocyclic drug which had been given in adequate t with inade uate response (Thase & Rush, 1995).~ i t h i u m a u ~ m e ~ t a 01’s has been reportedin a smaller numberof studies, features of the 5-HT syndrome have been seen with lithium-SS tions (Fontaineet al., 1991).
the antidepressant effectof TADs is weaker entation, although some ~atientsdo benefit 6).T3can cause a moderate ave a place in the outof other t r e a ~ e n tincluding s EGT are worth considering in the the author’s experience when ve been given adequate trial on an outufficient efficacy, then in-patient treatment orous physical and~ s y c h o l o ~ i -
ized indication for ECT and response rates of in patients who have not responded to an adequate courseof drug (Prudic et al., 1990, 1996). Althou~hEGT is the most layed in less depressed patients to allow
tre In the absenceof an established therapeutic window for most anti drugs, it is reasonable to conductan empirical trial of the effectiv maximally tolerated doseif there hasbeen inadequate response mended dose, provided that close supervision of unwanted dr possible. Thisis particularly appropriate for TCAs provid of EGG and plasma concentrations are possible(see Ch drug effects).
oni it or^ The co~binationof tryptophan with an MAOI needs careful on accountof the 5-HT syndrome (Sternbach, 1991) and the possibility Eosinophilic Myalgic Syndrome(EMS) (see Chapter IO), but it is a treat that has been shown tobe effective in severely depressed & Platma resistant to MAOI treatment alone (Pare, 1963; Glassman
For all of the treatments discussed so far there is either evidenc randomized controlled trialsof efficacy in patients with or, inthe caseof high doseanti~epressanttreatment, a cle backed upby extensive experience. The mainstay of the physical treat resistant depression is the correct use of these treatmentsof proven effi (as discussed above) and most patients with resistant depression can by these treat~ents,particularly when they are CO psychosocial inter~entions.A number of other biolog rently being evaluatedin specialist centres and, tho ommended for general use, several of these have be in this book. The combination of pindolol with anS ated or enhanced antidepressant response toSSan & Bergeron, 1995) as a consequenceof blockade oft autoreceptors which normally inhibit 5-HTcell fir increased. Another strategyof theoretical interesti synthesis and/or the blockade of central corticos et al., 1985; O’Dwyeret al., 1995).
r 3, the stereotactic e followed by res
essed treatment resistant patients (Lovett et al., 1989; Poyntonet al., 1995). ontraindications include personality disorder and drug and alcohol dependence, but in appropriately selected cases response rates of up to 50% have been reportedin patients who have failed to respond to the main treatments for resistant depression. Epilepsyis seen in 1% of cases operated upon but other permanent unwanted effectsof surgery are extremely rare.This treatUK requires the approvalof the Mental Health Act Comment, which in the missioners, needs further evaluation and only is available at a few specialist centres. In the caseof all treatments for resistant depression it is necessary to monithe treatment (with plasma drug concentrations) and the response observer rating scale). This chapter will conclude awith discussion so the n p a t h ~ a y between s the above treatments, but before doing ent of treatment resistant psychotic, bipolar, winter brief andrecurssion willbe considered.
is the most effective treatment for psychotic depression and response rates of 50% (Sackeim et al., 1990) and 88% (§piker et al., 1985) have been reorted when ECT has been given to patients with psychotic depression who ave not responded to combined treatment with neuroleptic medication and of a e than ECT (Parkeret al., 1992) the combination with a neuroleptic drug achieves partial improvement in 70--8o% of ).M e n psychotic depression is resistant to treatt it is wise to use the maximally tolerated dose of TAD with a moderate ose of neuroleptic and to measure plasma tricyclic drug concentrationsin of some neuroleptics to inhibit the metabolism of S. The combination of the §§RI fluoxetine with the neuroleptic perstigated in psychotic depression but resulted in S in half of the patients studied (Rothschildet al., f individual patients with treatment resistant psyequently responded to treatment with clozapine assa et al.,1993) areof interest and merit further chotic depressives relapse overa follow-up petenance treatment with TA is recommended. withdrawn cautiously once patients have fully yskinesia: research on e is needed since these
ajor life events
Fig. 20.3 A typical lifechart of a patient with a severe bipolar affective disorder, which shows the effects of different treatments and life events. For discussion, see text.
The physical treatments available for treatment resistant bipolar affective disorder have been reviewed earlier and will notbe discussed further here. In this sectionof this chapter the general principles which have been developed for treatment resistant unipolar affective disorder will be applied to the treatment of resistant bipolar affective disorder. x The most important aspect of assessment is the charting of the illness over time with detailed attention to the numberof admissions to hospital and the duration of illness on each trial of mood stabilizer (Fig. 20.3). The constructionof such chartsis time consuming and involves checking from records of plasma drug concentrations to determine whether prescribed mood stabilizers were in fact taken prior to relapses. Predictions can be made on the basis of the literature reviewed in Chapter 13, but the individual casedecisions about the useof mood stabilizers are made on empirical grounds. In the example citedin Fig. 20.3, charting of the illness revealed that neither lithium nor carbamazepine reduced the severity of illness, that both tricyclic antidepressants and ECT triggered hypomania, that rapid cycling onlydeveloped while tricyclic antidepressants were prescribed and that life events had no apparent effecton the courseof the illness. 2 In general, the use of antidepressants is to be minimized in bipolar depression since hypomania or mania have been reported to develop in 31-70% of the series studied (Goodwin &: Jamison, 1990).Antidepressants
445
especially in patien shuler et al., 1995 an g. 20.3). In such patients lithium should be tried asan antidepressant U ximally tolerated plasma lithium concentration (i.e. 1 . 2 n ~ L -(Sachs, ~) 1996). Carbam epine can also be use for the treatment of bipolar depression (Postet al., 1 Lithium discon~nuationis another hazar m =50% of cases (Faeddaed al., 1993; ~ o o d w se with ratesof up to rium is a time of ogenic effects,lithiu rted (see Chapte should be w i ~ d r a ~ Fortunately, relapseis ra of relapse during the puerperium, very close supe~ision is needed followin childbirth. The mood stabilizers, which are the mainstay of treatm which there is good evidence from controlled studies of l lactic efficacy. As has been described in Chapter 13, only lithium and carbama%epine and (possibly) valproate meet these criteria. It is therefore essential to ensure that adequate courses of each have been given an response to each is charted. M e n bipolar illness is truly resistant to courses of t r e a ~ e n tboth , with lithium and carbama%epine, selection of future treatments is then made on theoretical rather than empirical grounds. C o ~ o n l adopted y strategies in this s i ~ a ~ are o nthe combin with ~arbama%epineand the addition of valproate and S anticonvulsants. "he rationale for combining lithium and carbama%epine is own but through different pharmacological mechathat eachis effective on its nisms. The combination may therefore be expected tobe more effective than the effecti~enessof either drug given alone. The rationale for the use of anticonvulsants other than carbama%apine and valproate is that as carbama%ep~e and valproate are effective mood stabilizers, very careful charting of the effect of such experimental treatmentsis needed; but when a patient has frequent episodes then it is possible to determine on empirical grounds whether an experimental treatment such as valproate isof help in an in~ividualcase. Other preventative treatments for w ~ i c hthere is less empirical evidence are reviewed in Chapter 13, including the use of other anticonvulsants. Patients with good outcome often develop coping strategies as avoidsuch ing sleep deprivation and the extremes of overwork. Some learn to recognize signs of incipent hypomania, such as insomnia, and have agreed with their sychiatrist medication changes tobe made under such circumstances. The in such patients is being evaluated at effect of cognitive, behavioural therapy present (see Chapter 5).
Social adversityhas little, if any, influence on the course of recurrent bipolar affective disorder and there is no reason to thinkthethat course of such an illness can be changed by social management. However, depression is common and understandable in thespouses of such patients and rates of marital breakdown as high as 50% have been reported (Brodie & Leff, 1971). It is for this reasonthat time should be spent giving informationand support to the families of patients with resistant bipolar affective disorder. Mem~ership of self-help groups such as the Manic Depression Fellowship in the UK is recommended.
Although somepatients with seasonal affectivedisorder (SAD) do not meet criteria for resistant depression, someand do,it is convenient to discuss SAD in the chapter devoted to resistant depression. SAD is, as its name suggests, an affective disorder which meets operational criteria for seasonality. The original criteriaof seasonality were that major depression should be present inat least two consecutive previous years in which the depression developed during a u t u m or winterand remitted by the following spring or summer (Rosenthal et- al., 1984). The subsequent DSM-111-R (American Psychiatric Association, 1987) criteria were more specific and required that two-thirds of all onset occurred during the same 60-daywindow and that two-thirds of all offsets occurred during a second 60-day window. The DSM-IV criteria are less specific but remain controversial, and many SAD researchers still use Rosenthal's original criteria.SAD is characterized by the 'atypical' symptoms of increased sleeping, carbohydrate craving,and weight gain. Many, though not all, cases of SAD follow a bipolar Iand a few a bipolar I1 course (Oren& Rosenthal, 1992). The treatmentof SAD is the same as for any other form of affective disorder with the exception that light treatment is used in the treatment of those SAD patients whose depressions are present only in winter. The best evaluated treatment regimes are those which involve 2500 lux of bright artificial daylight given in the morning from 7:oo to 1o:oo and in the evening from8:oo to 11:oo. Treatment response is extremely rapid and most patients with SAD note a50%reduction insymptoms after 4 days of treatment. Relapse occurs equally quickly after stopping light treatment; therefore if the treatment is effective it should be continued indefinitelythroughout the winter months. More recently, to shortenthe period of treatment, strongerlight (10000 lux) has been given for just half an hour early in the morning (Terman et al., 1990) and two un~ublishedstudies have found this to be more effective than a 'placebo' treatment which was provided by an ion generator. Despite this, doubts remain about the possibility of a placebo response to light treatment for a conditionin which patients expect bright light to help them.As in all
447
patients with resistant depression it is essential to use objective rating scales of the severity of depression-adapted for SAD-before and after the administration of light treatment and periodically thereafter (Termanet al., 1992). When light treatment is effectiveit needs to be continued throughout the winter period. There is some evidence thatlight treatment given before the onsetof depression can prevent relapses of SAD. In theUK,an excellent self-help group is available: the Seasonal Affective Disorder Association (SADA)
~
Brief recurrent depression is a variant of unipolar depression in which the full range and severityof depressive symptoms are present for episodes lasting no more than 2-14 days and recurring at least once a month for at least a year (Angst, 1994). Brief recurrent depression responds poorly to all antidepressant drugs (Montgomery, 1994).
The decisions which need tobe made concerning the physical treatmentof most patients with resistant depression be can summarized in four diagrams. Since each treatment decision takes 6 weeks to evaluate it is necessary tobe economical in the selection of treatment options. The following selectionis based on the critical evaluation of each of these options as described in Chapter 11. Inevitably the selectionis a matterof judgement and alternative decision pathways have been describedby others (e.g. AmsterdamSr: HorningRohan, 1996). The first decision is whether the patient is best treated with a TAD, an SSRI or an MAOI. The factors which will influence this decision will include some evidence that TADs may be more effective thanSSRIs and MAOIs in more severely depressed patients (Perry, 1996), some evidence that MAOIs may be more effective than other antidepressants in patients with atypical profile of side-effects uitkin et al., 1989), and the ~i€ferent and drug interactions of the three classes of drugs (Chapter 9). Most important of all is the empirical evidence derived from the patient’s history and, if necessary, confirmedby therapeutic trial of the differen~al effect of the three main classesof antidepressant upon the illness in question. The choice of the optimal classof antidepressants for each patient is therefore madepartly on empirical grounds andpartly with reference to the literature. Despite the suggestion raised in Chapters 10 and 11 for severalSSNs and for venlafaxineit is not yet established that there are important within, class differences for the efficacy of individual drugs in resistant depression. The
44
choice of the individual memberof the class, therefore, canbe made on the basis of the side-effect profiles as tabulated in Chapter IO. Once an individual drug has been selected it is essential to ensurethat an adequate plasma drug concentration is obtained, that an adequate courseof treatment as described above is given and that antidepressant response is measured using an observer-rated depression scale. The next decision how tois augment the probably modest antidepressant response to the first drug. A different decision tree applies to each of th6 three classes of antidepressant drug, although in each case the first option is lithium augmentation.
of the antiFigure 20.4 lists the options tobe considered in the augmentation depressant effectof a TAD. Lithium augmentationis considered first in view of the strength of the evidence from controlledtrials that this is an effective intervention in treatment resistant depression (Chapter 10) and the acceptability by most patientsof this treatment combination. When lithium augmentation is contraindicate^, for example by serious kidney disease, or when it has been tried adequately but without success, then the additionof T3can be considered. The evidence from controlled trials is positive (Aronson et al., 1996) though not as strong as the case is with lithium augmentation (Joffeet al., 1993). Care is needed since theT3 may exacerbate anxiety and weight loss, and it may also cause cardiac arrhythmias. Since lithium and T3 are the only agents which have been shown in con%rolled studies to augment the antidepressant effect of TADs in resistant depression, the choiceof treatment strategyif they have both failed is arbitrary. The useof high doseTAD treatment at this point is logical. Caution is needed
Fig. 20.4
Available options for trying to enhance the antidepressant action of a tricyclic antidepressant drug.
449
with the monitoring of all the usual side-effects of TADs and the monitoring of the ECG and plasma TAD concentrations is essential. The final option to consider in the enhancement of the antidepressant effect of a TAD is the addition of an MAOI. Although the evidence for the effectiveness of TAD-MAOIs combinations in resistant depression is not strong (Chapter 11) the combination has been used for many years. The precautions that shouldbe taken include strict adherence to a low tyramine diet, the avoid ance of tranylcypramine and sympathomimetic agents, the avoidance of imipramine and chlomipramine (both of which cause features of the 5-HT syndrome in combination with MAOIs) and monitoring of pulse and blood pressure. A further reason for adding MAOIs is that if the combinationis not effective then theTAD can be withdrawn and several options for enhancing the antidepressant effect of MAOIs (see below) can be explored. The same of an SSIU to a TAD, although againno logic would also support the addition evidence of efficacy from controlled trials is yet available for SSRI-TAD 11). combinations in patients with resistant depression (Chapter a ~ ~ eof~ theea ~ t ti ressant ef~ectofan
The first decision to consider in the attempted enhancement of an antidepressant action of an SSRI (Fig. 120.5) is whether or not to add lithium. Lithium augmentation is the only procedure which has been shown to enhance the antidepressant effectof SSRIs in con~olledstudies. Careis needed with the of 5-HT (Chapter combination. since both treatments increase the availability 3) and featuresof the 5-HT syndrome have been reported withSS~-lithium" combination (Young & Cowen, 1994). The addition of pindolol to an SSIU is worth considering. Pindolol isan antagonist at betal and 5-HT1, receptors and through its action at 5-HTl, somatodendritic receptors has been shown to enhance the effect of SSMs on
Fig. 20.5 Available options fortrying to enhance the antidepressant effect of a selective serotonin re-uptake inhibitor.
5-HT neurotransmission by inhibiting negative feedback control of 5-HT cell firing byintras~aptic 5-HT (Arboreliuset al., 1995). Pindolol might therefore be expected both to accelerate and to enhance the antidepressant effect of SSRIs as has recently been reported in two preliminary studies (Artigas et al., lier &: Rergeron, 1995). Particular care is neededwith pindolol-§§M combination in view of the limited experience to date and the likely ynamic ~teractionsbetween the two. The fourth decision to make in the enhancement of the antidepressant effect of an SSM is whether or not to add a TAD. However, the evidence that this combination isof value in the treatment of resistant depression comes only from uncontrolled studies (Weilburget al., 1989,1991; Sethet al., 1992). S
The safestand best established way to enhance the antidepressant effect of an MAOI isto add lithium (Fig. 20.6) (Johnson, 1991). omb bin at ions of tr~ptophanwith MAOIs have been used for many years in the treatment of resistant depression and with encouraging results in severely depressedpatients (Pare, 1963and Glassman &:Platman, 1969). The combination producesa powerful enhancementof 5-HT neurotransmissions in animals, and in man featuresof the 5-HT syndrome are seen which limit the dose of tryptophan to 2 g. ore recently tryptophan has been withdrawn a conon account of the EMS. However, theEMS is thought to be caused by taminant in the preparation of tryptophan (Slutsher et al., 1990). The tryptophan-MAO1 combination should be considered seriously when and if tryptophan becomes readily available for clinical use. The combined use of MAOIwith TAD has already been discussed (above).
Fig. ~ 0 . 6Available options for trying to enhance the antidepressant action of a ~ o n o a oxidase ~ ~ e inhibitor.
Perhaps the most important decision in the physical treatment of resistant depression concerns the timing of the ECT. use No of treatment has been shown to be more effective or more quickly effective in the treatment of resistant depression. Consequently, ECT is used either when isthere a threat to the life of the patient (from suicide, starvation or dehydration), or when patient and doctor agree that thelikely delays implicit in future drug treatments are not warranted. In the UK,though not always in the USA, it is normal practice for an antidepressant to be continued throughout the course of treatment. This is one reason ,whyit is important to decide as soon as possible which class of antidepressant drugis most effectivein an individ~alcase.
The treatment combinations which have been discussed so far include the use of no more thantwo antidepressant drugs at any one time. Drug combinations involving three or more antidepressant drugs have been described (Hale et al., 1987). The intention of such combinations is mostly to enhance 5-HT neurotransmission. Since features of the 5-HT syndrome are seen with several of the two drug Combinations (e.g. lithium.+,SS~I and tryptophan.+,MAOI) the theoretical case for a third or fourth drug is hard to justify. It is also more difficult to evaluate the effectiveness of the different componentsof such treatment combinations. To date, there is no evidence from randomized controlledtrials that drug combinations which involve three or two. more psychotropic agents are more effective than those which employ
Most patients with resistant depression gain some benefit from the optimal use of the physical treatments for which there is good evidence from controll studies of efficacy in resistant depression. The clinical challenge in these patients is to convert an improvementin physical symptomsinto a rewarding change in lifestyle using the behavioural and psychosocial strategies described elsewhere in this chapter. Meanwhile, research into new treatments for resistant depression continues.
Abrams, R. (1992) ~ l e c t r u c u ~ v ~ ~ s ~ v e 2nd T h e y a ~electroconvulsive ~, therapy. Archives of General edn. Oxford University Press, Oxford. ~sy~hiatyy, 48,746-748. Abrams, R., Swartz, C.M. & Vedak, C. (1991) Anti- Altshuler, L.L., Post, R.M., Leverich, G.S. et al. (1995) Antidepressant-induced mania and cycle depressant effect of high-dose right unilateral
C
~20: ~ ~e s i s ~~ ~ n ~~
~ee ~ ~
sion: a comparative studyof patient characteracceleration: a controversy revisited.American istics. America Journal of Psychiatry, 127, 1086Journal of Psychiat~, 152,1130-1138. 1090. American Psychiatric Association (1987) Di~gnostic and Statistical Manualof Mental Disorders,re- Brown, G.W. & Harris, T.O. (1978) Social Originsof vised 3rd edn. American Psychiatric Association,Depression: A Study of Psychiatric Disorder in London. ~ o ~ eTavistock, n . Washington, DC. American Psychiatric Association (1990) Clinical Brown,G.W. & Moran, P. (1994) Clinical and psychosocial origins of chronic depressive epiGuidelines for the Practice of ECT. American sodes. 1: A community survey.British Journalof Psychiatric Association, Washington, DC. Psychiatry, 165,447-456. American Psychiatric Association (1994) Diagnostic and Statistical Manual of M e ~ t aDisorders, l 4th Brown, G.W., Andrews, B., Harris, T.O., Adler,Z. & Bridge, L. (1986) Social support, self-esteem edn. American Psychiatric Association, Washand depression.Psychological Medicine/16,813ington, DC. 831. Amsterdam, J.D.& Hornig-Rohan,M.(1996) Treatment algorithms in treatment resistant depres- Brown, G.W., Lemyre, L. & Bifulco, A. (1992) Social factors and recovery from anxiety and sion. Psychiatric Clinicsof NorthAmericff/zg, 371depressive disorders: a test of the specificity. 386. British Journalof Psychiatry/ 161,44-54. Anderson, C.M., Griffin,S., Rossi, A., Pagonis,I. et all (1986) A comparative study of the impact Brown, G.W., Harris, T.O. & Hepworthr C.(q94a) of education vs. process groups for families of Life events and 'endogenous' depression: a puzzle re-examined. Archives of General Psychiat~, patients with affective disorders. Family Process, 25,185-205. 51, 525-534. Angst, J. (1994) The history and concept of recur- Brown, G.W., Harris, T.O., Hepworth, C. & rent brief depression.ropean an Archives of PsyRobinson, R.(1994b) Clinical and psychosocial chiatry and Clinical Neuroscience/244,171-173. Britorigins of depression. 2: A patient enquiry. AHCPR (1993) Diagnosis and treatment of depres- ish ~ournalof Psychiatry, 165,457-465. sion in late life: NIH Consensus Development Cleare, A.J., MacGregor, A.,& Chambers, S.M.et al. (1995) Thyroxinereplacementincrease Panel on depression in late life. JournaZ of the central 5-hydroxytryptamine activity and reAmerican MedicaZ Association, 268, 1018-1024. duces depressive symptoms in hypothyroidism. Arborelius,ANomikos, G., Grillner,P., Hertel, P., Hook, B.B., Hacksell, U. & Svensson, T.H. (1995) Neuroendocrinology, 64, 65-69. 5-HTI, receptor antagonists increase the activ- Cochran, S.D.(1984) Preventing medical non-comity of serotonergic cells in the dorsal raphe nu- pliance in the outpatient treatmentof affective cleus in rats treated acutely or chxonically with disorders. Journal of Consulting and Clinical citalopram. Naunyn Schmiede~erg's Archives of PSYChOlO~,52,5440. Corney, R.(1987) Marital problems and treatment P h a r ~ a c o l o352,157-165. ~, outcome in depressed women: a clinical trial of Aronson, R., Offman, HJ., Joffe, R.?: & Naylor, social work intervention. British J o ~ r n aofl C.D. (1996) Triiodothyro~neaugmentation in the treatmentof refractory depression.Archives Ps~chiatry,151,652-659. Dassa, D., Kaladjian, A., Azorin, J.M. et al. (1993) of General Psychiat~,53, 842-848. Artigas, F., Perez,V. & Alvarez, E.(1994) Pindolol Clozapine in the treatment of psychotic refracinduces a rapid improvement of depressed patory depression.British Journalof Psychiatry,163, tients treated with serotonin reuptake inhibitors.822-824. Archives of General Psychiatry/51,248-251. Depression Guideline Panel (1993) ClinicaZ Practice Guideline: Depression in Primary Care. Billings, R., Tang, S.W. & Rafkoff V.U. (1981) DeVol. 2: pression associated with ranitidine. American ~ r e a ~ m eofn tMajorDepression.US Department of Health and Human Services, Washington DC. Journal of Psychiatry, 743,915-916. Blier, P. & Bergeron, R. (1995) Effectiveness of Devanand,D.P., Sackeim, H.A.& Prudic, J. (1991) pindolol with selected antidepressant drugs in Electroconvulsive therapy in the treatment-rethe treatment of major depression. Journal of sistant patient. Psychiatric Clinics of North Clinical Psychopharmacolo~/15/ 217-222. America, 15,902-923. British National Formulary (1996) British National Fadden, G., Bebbington, P. & Kuipers, L. (1987) For~ulary,32 (Sept.). (ed C. F. George) British Caring andits burdens: a studyof the spouses Medical Association and Royal Pharmaceutical of depressed patients. brit is^ Journal of PsychiSociety of Great Britain, London. atry, 151, 660-667. Brodie,H.K.H. & Leff, M.J. (1971) Bipolar depresFaedda, G.L., Tondo, L., Baldessarinin, R.J.,
453
Suppes, T. & Tohen, M.(1993) Outcome after rapid vs. gradual discontinuation of lithium treatment in bipolar disorder. Archives of General Psychiatry,50,448-455. Fava, M. & Davidson, K.G. (1996)Definition and epidemiologyof trea~ent-resistantdepression. Psychia~ricClinics of NorthAnzeric~,19,179-200. Fink, M. (1992) Electroconvulsive therapy. In: andb boo^ of Afective Disorders (ed. E. S. Paykel), pp. 359-368. Churchill Livingstone, Edinburgh. Fontaine, R., Ontiveros, A. & Elie, R. et al. (1991) Lithiumcarbonateaugmentation of desipramine and fluoxetine in refractory depres~, sion. Eiological P s y c h i a ~29,946-948. 11personality Frank, E. & Kupfer, D.J. (1990) Axis disorders and personality features in treatmentresistant and refractory depression.In:Treatment Strategiesfor ~ ~ a cD~ression t o ~ (eds S. P.Roose & A. H. Glassman),pp. 207-221. American Psychiatric Press, Washington DC. Alliance Frank, E., Kupfer, D. & Siegel, L.R. (1995) not compliance:a philosophy of outpatient care. ~ o u ~of aClinical l Psychiat~,54 (Suppl. I), 11-16. Ghadirian, A.M., Engelsmann, F. & Dhar, V. et al. (1985) The psychotropic effects of ~ ~ i b i t oofr s steroidbiosynthesis in depressed patients refractory to treatment. Bi~logicalPsychiatry, 37,368375. Glassman, A.H. & Platman, S.R. (1969) Potentiation of monoamine oxidase inhibitor by tryptophan. ~ournalof Psychiatric~esearch7,8388.
Goodman, W.K., McDougle, C.J. & Price, L.H. (1992) Pharmacotherapy of obsessive compul, sive disorder. ~ournalof Clinical P s y c k i a t ~53 (SUPP14) 29-37. Goodwin, G.M. (1994) Recurrenceof mania after lithium withdrawal: implications for the use of lithium in the treatment of bipolar affectivedis144, 149-152. order. British ~ournalof Psychi~~ry, Goodwin, F.K. & Bunney, W.E. (1971)Depression following reserpine: a re-evaluation. Seminars in Psychiatry, 3,435-448. Goodwin, E.K. & Jamison, K.R. (1990)Manic Depressive Illness. Oxford University Press, New York. Hale, A.S., Procter, A.W. & Bridges, P.K. (1987) Clomipramine, tryptophan and lithium in combination for resistant endogenous depression: 7 case studies. Eritisk ~ournalof Psychiatry, 151, 213-217. Holzer, C.E., Nguyen, M.T. & Hirschfield, R.M.A. (1996)Reliabilityof diagnosis in mood disorder. Psychiatric Clinicsof North America,xg,73-84. Hooley, J.M. (1986) Expressedemotion and depression: interactions between patients and high vs.
454
low EE spouses. ~ o ~ r nofaAl b n o r ~ aPsychology l 951 237-246. Joffe,R.T., Singer,W., Levitt, A.J.et al. (1993)Aplacebo-controlled comparison of lithium and triidothyrone augmentation of tricyclic antidepressants in unipolar refractory depression. Archives of General P s y c h i a ~50,387-393. ~, Johnson, F.N. (1991) Lithium augmentation in Con~emporary therapy for depression. Re~)iews P ~ a r ~ a c o t h e r a2,332. ~y, Kelly, W.F., Checkley, S.A. & Bender, B.A. et al. (1983) Cushing's Syndrome and depression: a prospective study of 26 patients. Eritisk ~ o ~ r n a l of Psychiatry, 142,16-19. al. (1992) Kendler, K.S., Neale, M.C., Kessler, et R.G. A population- base^ twin study of major depression in women: the impact of varying definitions Psychiat~, 49,272of illness.Archives of General 281. Lackheim, H.A.,Prudic, J. & Devanand, D.P. et al. (1993)Effects of stimulus intensity and electrode placement on the efficacy and cognitive effects of electroconvulsive therapy. New England ~ournalof ~ e d i c i n e328,839-846. , Lam DH,Green B and Power MJ et al(1994) The impact of social cognitive variables on the initial level of depression and recovery. ~ o u rof~ ~ l Afective Disorders, 32,7543. Lewis, D.A.& Smith, R.E. (1983)Steroid induced a ~Afective Dispsychiatric syndromes. ~ o u r n of orders, 5, 3q-332. Linehan,M.M. (1993)C o ~ i t i v e - ~ e ~ a v i o r a l T r e a t ~ ~ t of E o r d e r ~Personal~ty ~~e Disorder. Guilford Press, New York. Lishman, W.A. (1987) Organic Psychiat~:the Psyc k o ~ o g i c a ~ ~ o n ~ u of e nCerebral ces Disorder, 2nd edn. Blackwell ScientificPublications, Oxford. Lovett, L.M., C r i m i n s , R. & Shaw, D.M. (1989). Outcome in unipolar affective disorder after sk ofPsych~stereotactictractotomy. B r i ~ ~~ournaZ atry, 155, 547-550. Mason, B.J. & Kocsis, J.H. (1991) Desipramine treatment of alcoholism. P s y c ~ o p ~ a r m a c o l ~ g ~ ~ u ~ ~ 27,155-161. e t i ~, McGrath, P.J.,Stewart,J.W., Nunes, E.V. et al. (1993) A double-blind crossover trial of imipramine and phenelzine for outpatients with treatmentof Psyrefractory depression. American ~olfrnal chiatry, 150,118-123. McGrath, P.J., Nunes, E.V., Stewart,J.W.et al. (1996) Imipramine treatment of alcoholics with primary depression: a placebo controlled clinical trial. Archives of General Psychiat~,53,232-240. McGuffin, P. & Katz, R. (1989) Thegenetics of depression and manic depressive illness. Br~tish ~ , 294-304. ~ournalof P s y c h i a ~155,
~ ~ a ~20:t ~ee sri s t a n t ~ e ~ r e s s i o n McGuffin, P. & Katz, R. (1996) Nature, nurture and (1995). A prospectiveclinicalstudyof stereotactic subcaudate tractotomy.Psychologidepression: a twin study. PsychologicalMedicine, cal Medicine, 25,763-770. 21, 329-335. Miller, I.W., Norman, W.H. & Keitner, G.I. (1990) Prince, M., Harwood,H., Blizard, R., Thomas,A. Treatment responseof high cognitive dysfunc& Mann, A. (1997) Impairment, disability and handicap as risk factors for depression in old tion depressed inpatients. Com~arativePsychiage. Psychoio~calMedicine, 27,311-322. atry 31,62-71. Prudic, J., Sackeim, H.A. & Devanand, D.P. (1990) Montgomery, D.B., Roberts, A., Green, M., Bullock, T., Baldwin, D. & Montgomery, S.A. (1994) Lack Medication resistance and clinical response to electroconvulsive therapy. Psychia~ryResearch, of efficacy of fluoxetine in recurrent brief depression and suicidal attempts. ~uropean Archives of 31,287-296. Clinical Psychiatryand ClinicalNeuroscience,244, Prudic, J., Haskett, R.F., Mulsant, B. et al. (1996) Resistance to antidepressant medications and 211-215. short-term clinical response to ECT. American Nunes, E., Quitkin, EM., Brady, R. et al. (1991) Imipramine treatment of methadone mainte~ o u ~of aPsychiatry, l 153,985992. nance patients with affective disorderillicit and Quality Assurance Project (1983) A treatment outline for depressive disorders. Australian and New drug use. American Journal of Psychiatry, 148, 667-669. Zealand Journal of Psychiatry,17,129-146. Quitkin, EM., McGrath, P.J. Stewart, J.W. et al. & Marks, M.N. et O ’ h y e r , A.M., Lightman, S.A. (1989). Phenelzine and imipramine in mood al. (1995) Treatment of major depression with reactive depressives.Archives of General Psychimetyrapone and hydrocortisone. Journal of Afective Disorders, 33,123-128. atry, 46,787-793. O’Leary,K.D. & Beach, R.H. (1990) Marital Quitkin, F.M., McGrath, P.J.,& Stewart, J.W. et al. therapy: a viable treatment for depression and (1996) Chronological milestones to guide drug marital discord.American J o u ~ aof change. Archives of General Psychiatry, 53, 787l Psychiatry, 147, 229-235. 792Reifler, B.V., Teri, L.,Raskind, M. et al. (1989) DouOren, D.A. & Rosenthal, N.E. (1992) Seasonal ble-blind trialof imipramine in Alzheimer’s disaffective disorder.In:~ a n d ~ofoAfective o ~ Disease patients with and without depression. orders (ed. E. S. Paykel), pp. 551-568. Churchill Livingstone, Edinburgh. Amffican~ o u ~of aPsychiatry, l 146,45-49. et al. (1987) Pare, C.M.B. (1963) Potentiation of monoamine Reynolds,C.F., Perel, J.M., Kupfer, D.J. oxidase inhibitors by tryptophan. Open trial response to antidepressant treatment Lancet, ii, 527in elderly patients with mixed depression and 528. cognitive impairment. Psychiatry Research, 21, Parker, G., Roy, K., Hadzi-Pavovic, D. e f al. (1992) Psychotic (delusional) depression: a meta111-122. analysis of physical treatments.Journal of AfecRobin, A. & De Tissera, S. (1982) A double blind tive Disorders, 24,17-24. controlled comparisonof the therapeutic effects Pascualy, M. & Veith, RC. (1989). Depression as of low and high energy electroconvulsive an adverse drug reaction. In: Aging and Clinical therapy. British ~ o u ~of aPsych~try, l 141,357-366. Practice (eds R. G. Robinson& P.V. Rabin), pp. Rosenbaum, J.F., Fava, M., Nierenberg, A.A.et al. 132-151. Igaku Shoin, New York. (1995) Treatment-resistant mood disorders. Paykel, E.S., Flemingero, R.F. & Watson, J.P. (1982). In: ~reatmentof Psychiatric Disorders (ed. G. 0. Psychiatric side effects of antihypertensives Gabbard) ,2nd edn, Vol.1, pp. 1275-1283. otherthanreserpine. Journal of Clinical American Psychiatric Press, Washington, DC. P s y c h o p h a ~ a c o l o2,14-39. ~, Rosenthal, N.E., Sack, D.A. & Gillin, J.C.et al. (1984) Perry, P.J. (1996) Pharmacotherapy for major Seasonal affective disorder: a description of the depression with melancholic features: relative syndrome and preliminary findings with light efficacy of tricyclic vs. selective serotonin retherapy. Archives of GeneralPsychiatry,41,72-80. uptake inhibitor antidepressants. Journal of Rothschild, A.J. (1996) Management of psychotic, Afective Disorders, 39,1--6. treatment-resistant depression.PsychiatricClinPost, R.M., Uhde,T.W., Ballenger, J.C. et al. (1983) ics of North America, 19 (2),237-252. Carbamazepineandits IO-, 11-epoxide Rothschild, A.J., Samson, J.A., Bessette, M.P. & metabolite in plasma and CSF relationship to Carter-Campbell, J.T. (1993) Efficacy of the comantidepressant response. Archives of General bination of fluoxetine and perphenazine in the Psychiatry,40,673476. treatment of psychotic depression. Journal of Poynton, A.M., Kartsoukis, L.D. & Bridges, P.K. Clinical Psychiatry,54,338-342.
455
Royal College of Psychiatrists (1995) The ECT Spiker, D.G., Stein, J. & Rich, C.L. (1985) Delusional ~ a ~ d ~ Royal o o k . College of Psychiatrists. Londepression and electroconvulsivetherapy: One , don. year later. Convulsive ~ ~ e r u pI,y 167-172. Sternbach, H. (1991) The Serotonin Syndrome. Sachs, G.S. (1996)Treatment-resistantbipolar depression. Psychiatric Clinicsof North America,19, American ~ o u ~ nofaPsychiat~, l 148,705-713. 215-236. Stoudemire, A., Hill, C.D., Morris,R. et al. (1991) Sackeim, H.A., Decina,P., Canzler, M. et al. (1987) Cognitive outcome following tricyclic and elecEffects of electrode placement on the efficacy of troconvulsive treatment of major depression in titrated low-dose ECT. American ~ournulof Psythe elderly. American ~ournulof Ps~chiatry,148, chiatry, 144,1449-1455. 1336-1340Sackeim,D.G., Prudic, J. & Devanand, D.P. (1990) Teasdale, J.D. & Barnard, P.J. (1993)Aflect, CogniTreatment of medication-resistant depression tion and Change:Re~odelling Depressive ~hought, In:A~ericunPsywith electroconvulsive therapy. Lawrence Erlbaum, Hove. chiatric Press Reviewof Psychia~ry, Vol. 9 (eds A. Terman, J.S., Terman, M., Schlager, D. et al. (1990) Tasman, S. M. Goldfiner & C. Kaufman), pp. Efficacy ofbrief, intense light exposure for treat91-115. American Psychiatric Press, Washington, ment of winter depression. P s y c h o p ~ a r ~ a c o ~ o ~ BulZetin, 26,3-11. Salzman, C. (1995)Medication compliance in the Terman, M,, Williams, J.B.W. & Terman, J.S. (1992) elderly. ~ournaZof Clinical Psychiatry,56 (Suppl. Light therapy for winter depression. In:InnovaI), 18-23. tions in Clinical Practice: A Source Book, Vol. 10 Sato, T., Sakado, K. & Sato, S. (1993) Isthere any (eds P. Keller & S. R. Heyman), pp. 179-221. Prospecific personality disorder or personality disfessional Resources Executive, Florida. order cluster that measures short-term treatment Thase, M.E. (1996)The roleof Axis I1 comorbidity outcome of major depression? Acta Psych~atrica in the management of patients with treatment Scandinavica, 88,342-349. resistant depression. Psyc~iatricClinics of North America, 19,287-310. Schweizer, E., Rickels, K.,Amsterdam, J.D. et al. (1990) What constitutes an adequate antideThase, M.E., Mallinger, A.G., McKnight, D. e f al. (1992)Treatment of imipra~e-resistantrecurpressant trial for fluoxetine? ~ournulof CZinical Psyc~ia~ry, 51,8-11. rent depression. IV A double-blind crossover study of tranycypromine for anergic bipolar Scott, J. (1988)Cognitive therapy with depressed depression. American ~ournalof Psychiatry, 149, inpatients. In: Develop~entsin Cognitive Psycho195-198. therapy (eds P.Tower & W. Dryden), pp. 177189. Sage Publications,London. Thase, M.E.& Rush, A.J. (1995)Treatment-resistant depression. In: Psyc~opharmacology: The Scott, J., Cole, A. & Eccleston, D. (1991) Dealing ~ o u r t hGeneration of Progress (eds F. E. Bloom & with persisting abnormalities of mood. InternaD. J. Kupfer), pp. 1081-1097. Raven Press,New tional Review of Psychia~ry,3, q-33. York. Seth, R., Jennings, A.L., Bindman,J., Phillips, J. & Bergman, K. (1992) Combinationtreatment with Vaughn, C.E. & Leff, J.P. (1976) The influence of family and social factors on the course of psynoradrenalineand serotoninreuptake inhibitors chiatric illness. British ~ournalof Psychiatr~,129, in resistant depression. British ~ournalof Psychi125-137. atry, 161,562-565. Shea, M.T., Pilkonis, P.A., Beckham,E. ef al. (1990) Wehr, T.A.& Goodwin, F.K. (1987)Can mtidepressants cause mania and worsen the course of Personality disorders and treatment outcome in u~ affective illness? A ~ e r i c a n~ o ~ rofnPsychjat~, the NIMH treatment of depression collaborative 144,1403-1411. research program. Ame~icun~ournalof PsychiWeilburg, J.B., Rosenbaum, J.F., Biderman, J., atry, 147, 711-718. Sachs, G.S., Pollack, M.H. & Kelly, K. (1989) Shea, M.T., Widiger, T.A. & Klein, M.H. (1992) Fluoxetine added to non-MOI antidepressants Comorbidity of personality disorders and deconverts non-responders to responders: a prepression: Implications for treatment. ~ournulof liminary report. ~ournalof Clinical Psychiutry,50, Consulting and ClinicalPsycholo~J,60,857-868. Slutsher, L., Hoesly, F.C., Miller, L., Williams, I?, 447-449, Watson, J.C. & Hemming, D.W. (1990) Weilburg, J.B., Rosenbaum, J.F. & Meltzer-Brody, S. et al. (1991) Tricyclic augmentation of Eosinophilia-myalgiasyndrome associated with fluoxetine. Annals of Clinical Psychiatry,3,209exposure to tryptophan from a single manufacn a ~ A~ericun Medical ~ssoc~ation, 213. turer. ~ o ~ rofthe Whisman, M.A., Miller, LW., Norman, W.H. & 264,213-217.
x.
Keitner, G.I. (1991) Cognitive therapy with Wood, M.J. & Rubirtstein, M. (1990) An atypical depressed inpatients: specific effects on dysresponder to clozapine (letter). American Journal of Psychiatry,147,369. functional cognitions. ~ournalof C ~ n s u l ~ and ~ng Young, A.H. & Cowen, P.J.(1994)Antidepressant Clinical Psychology,59, 282-288. a ~(eds drugs. In: Side-efjcects of Drugs A ~ n u 17 Wilcox, D.R., Gilan, R.&Hare, E.H. (1965)Do psychiatric outpatients take their drugs? ~ r i ~ i s J.~ K. Aronson & C. J. Van Boxtell), pp. 14-24. Elsevier, Amsterdam. Medical ~ournal/ 2,790-792.
457
This Page Intentionally Left Blank
Please note: page numbersin tables; thosein italics refer to figures. acetyl choline 53,285 acne284 Addison’s syndrome adenosine 3’,5’-monophosphate (cyclic AMP) 281 adenylate cyclase 56 ’adjuvant psychological therapy’ 38 adolescents age and depression 20 EEG abnormalities 352 management of depression see children, treatment of depression puberty20,358 sex ratio and frequency of depression 18-19,20,348 see also children adrenal gland hypertrophy 47 steroid suppression 245-6 a,-adrenoceptor antagonists 246 adrenoceptors59,246 Adult Attachment Interview 83 aetiology of depression 436-8 biological 4 7 4 3 , 4 3 6 cognitive103-19,120 environmelltal 44-7,70-4,75-9 genetic 42-4,55,436 medical 46,49,410-12,436 psychosocial72-5,437-8 affective disorders, classifications 1-5 affective psychosis, incidence 8 age and abnormalities in brain structure51 and frequency of affective disorders 19 in the elderly 396 lifetime prevalence 13 and recoveryin the elderly 396 of risk 11 and sex ratio 19-20 and women’s marital status 15,16 Agency for Health Care Policy and Research (US.)144 agmatine 59 agranulocytosis293
albumin, serum 391 alcohol abuse in recurrent affective disorder 278 dependence85,339 and antidepressants 434 aldosterone285 alpha-methyl-para-tyrosine (AMPT) 62 alprazolam 33,36 Alzheimer’s disease 388,393,417 and linkage analysis 43 ambivalence, anxious/ambivalent attachment styles 131-2, 133,135,136 amiloride283 amine precursors 206 see also tryptophan amineptine203 am~oglutethamide246 amitryptiline238 adverse effects 218,220,392 for children 353 combined with perphenazine 238 compared with problem-solvin~192 psychotherapy145,314,315 dosages 440 in primary care 33,312,314 withdrawal201 amoxapine, adverse effects 218 amphetamine245,278 AMPT (alpha-methyl-para-tyrosil~e)62 amputees 414 amygdala49 anaclitic depresssion/ traits see dependency Anaclitic and Introjective Dysfunctional subscales 106,1017 anaemia, aplastic 293 angiotensin285 anhedonia384 antibiotics, and lithium toxicity 289 antibodies, antithyroid 47 anticholinergic effects of atypical antidepressants 229 of SSIU treatment 222,223 of TCA treatment 218-19 antidepressants adverse effects 36,196-7,212-29 in physically-ill patients 418-21 ’atypical’ see ’atypical’ antidepressants
antidepressants (Cont.) augmentation for resistant depression 23744,443, 450-3 see also lithium, augmentation for children and adolescents 202,352-5, 360 ’cocktails’453 combined with psychotherapies, cognitive“ behavioural therapy 175 and comorbidity 201-2 compared with psychotherapies cognitive-behavioural therapy 166-7, 168,174,177, 184-!j1 442-3 interpersonal therapy 148-50 for maintenance treatment 151 comparison between 2 0 5 4 compliance 197’ cost-effectiveness 205,207 delay in action 198-9 discontinuation 201 in placebo-controlled studies 194 dosages 196-7 adequacy in resistant depression 440 for physically-ill patients 421 used in primary care 313-14 and drug abuse 2 0 3 4 duration of treatment 200-1,440-1 effect on 5-HT54-9,62-3 corticosteroid receptors 62-3 noradrenaline 59-62 efficacy 188-207 in primary care 312-13 for the elderly 202-3 fatal overdoses 204 historical background 189 for hospitalized depressions 38 interactions 190 long-term therapy 1934 MAOIs see monoamine oxidase inhibitors and patients experiencing severe events and difficulties 85 and personality factors 203 pharmokinetics 190 for the physically-ill see physical illness plasma concentrations 190,205,440 in primary care 32,33-5,196,312-14,320 prophylactic, for postnatal depression 375 reassessed in resistant depression4 3 9 4 1 response to 197-200 reviews and key references 206 in specialist care 35-6 SSRIs see selective serotonin re-uptake inhibitors and suicide 204-5,220,223 switching between classof 235-7,442-3 TCAs see tricyclic antidepressants treatments for single-drug resistant depression234-48 when psychosocial management fails90 worsening depressive symptoms 204-5 antipsychotic drugs 234
460
“atypical‘298 in drug combinations 238 with lithium 283,290,297 for recurrent affective disorder297-8 see also neuroleptic drugs anxiety 85,202 anxiety/depression prevalence 4,28-9 anxious/ambivalent attachment styles 131-3,135,136 comorbidity with depression 434 in the elderly 385,402 environmental triggers 43 arginine47 assessment procedures,in children 346-8 assumptions, dysfunctional 96-7,113-19 attachment theory 125-38 attachment styles 82-5,130-5 attachment system in adults127-8 attachment system in children 103-4, 126-7,129 and clinical practice 130,135-8,153 and depression 128-30 ’atypical’ antidepressants adverse effects 204,226-9 see also bupropion; mianserin; nefazodone; trazodone; venlafaxine ’atypical’ antipsychotics, in bipolar disorder298 ’atypical’ depression, treatment with MAOIs 192-3,449 automatic thoughts 94 Automatic Thoughts Questionnaire 99 autonomous individuals 105,110,112-13, 120,133 see also self-criticism avoidant attachment styles 131-2,134, 135,136 Beck Depression Inventory (BDI) 108,110, 114,115, 116 and cognitive-behavioural therapy 178 used for the elderly 388 beds, hospital and community, management by CMHTS 332-3 befriending intervention 79-80’86,439 behavioural therapy compared with experiential therapy156 efficacy from published data144 for elderly patients 403 behaviours, antidepressant 102 benzodiazepines202 bereavement counselling 402 bias, in psychosocial data collection71-2 biological rhythms, and depression50 bipolar disorder (manic depressive disorder) and age 276 and alcohol or drug abuse 278 antidepressants for 192,446-7 antipsychotic drugs for 297-8 charting over time 446 classification275 and driving vehicles 284 episode frequency 276-7
frequency 8,9,10,11,13 and gender 18,276 heritability 4 2 , 4 3 4 and life events 277-8 mania and classificationof bipolar disorder 275 diagnostic criteria for 276 and lithium therapy 279 onset276 postpartum 377 triggered by drugs 278 triggered by life events 277-8 and valproate treatment 295 and marriage 278,439 mood stabilisers carbarnazepine 2 9 2 4 in children 355 lamotrigine 296-7 lithium278-9,280-92 and rapid cycling phases 277 valproate295-6 number and duration of episodes 276 outcome277 in the postnatal period 376-7 rapid cycling 277,299,447 sex ratio 339 and social class 14-15 and subcortical lesions 51 and tardive dyskinesia 282,297 thyroid hormone treatment 299 treatment resistant 446-8 see alsomania; recurrent affective disorder birth see childbirth blood disorders, and use of antidepressants 422 bones, effect of lithium treatment on 284 BP-I, -11 and -111 disorders 275 brain changes in cortical activityin depression 50-2 changes in limbic forebrain function 47-50 changes in structure 50-1 brain disease, causesof organic depressions387 breastfeeding and antidepressants 371-2 MAOIs 216 moclobemide217 SSRIS 2256,371-2 TCAS221,371-2 brief recurrent depression 449 Brief Symptom Inventory108 brofaromine212,216 key reference 206 bulimia43 bupropion205 adverse effects 226,229 combined with lithium 239 Burns, DavidD. Feeling good 170 Thef e e l i ~ g g o harz~book o~ 170 buspirone 33,36,54,57
calcium, serum levels 387 calcium channel blockers 435-6 Camberwell, surveys in5,12 cancer patients depression among 37-8,387,410 antidepressant treatment 419 psychotherapy423,424 cannabis278 carbamazepine244 combined with lithium 245,294,447 compared with, lithium 292 interaction withSSRIs 225 interaction with TCAs 220 in recurrent affective disorder 2 9 2 4 , 4 4 7 and pregnancy 286 cardiovascular disease and depression 410 treatment of depression 418, 420,422 heart attacks depression at the time of 414 psychological interventions following 423,424 cardiovascular effects of lithium therapy 284 of TCA therapy 218,219,392 of tradozone therapy 228 care in the communitysee community, depressions in the levels of and presentationof depression 27-39 see also primary care; secondary care ‘multifaceted’ (MC) versus ’usual’ 33 see also treatment Care Programme Approach (CPA) 330,331 care-givers of elderly patients, and depression 401, 4034 primary and attachment theory 126-7,128-9, 131, 132-37 134 see alsomothers; parenting case conceptualization 118-19 case management,in the community 331-2 case register studies 9,10 CATEGO 1 , 5 , 6 PSE-ID-CATEGO system 5 categories, of affective disorder1-5 causes, of depressionsee aetiology of depression CBT see cognitive behavioural therapy cerebral blood flow, regional (rCBF)51’52 charting, bipolar disorder 446 child care,as a determinant of depression 15,16 childbirth, depression following 46,365,447 bipolar depression 376-7 and the family 367-70 frequency 364-5 and postparturn psychoses 365 prevention374-5,377 and suicide 365
children adversity in 80,81,88 and attachment problems 1034,126-8, 132-3 and cognitive theory 9 5 4 and depressionin the elderly 397 sexualabuse 45 and vulnerable personality types 104-10 age and depression 20 assessment of depressive disorders in346-8,358 attachment bond with primary caregiver 103,126-7,129 and comorbidity 348 and definitionof depression 345-6 and diagnosisof depression 346 early ’core beliefs’ 95-6 ‘exploratory behaviour’ 127 family and genetic factors350-1 increased incidenceof depression 349 infants cot deaths 368 depressed mother’s impacton 367-8, 377 impact on mother 367 link between childhood and adult depression350 mother-loss/loss of primary caregiver 80-1,126,128 neurobiological studies 351-2 outcome studies in 349-50 response to stress 45 and risk factors for depressive disorders 348-9 sex ratioof depressive disorders 18-19, 20,348 sleep abnormalities in 352 and suicidal behaviours 350 treatment for depression antidepressants 202,352-5,360 ECT 2 6 5 4 , 3 5 5 and the family/parent component357-8 generic counselling 357 psychological therapies 355-7 relaxation training 356 selection of treatment modality359-60 social skills training 357 treatment setting 359 see also adolescents chlordiazepoxide 30,36 cholinergic rebound 201 chronic fatigue syndrome 414,423 chronicity, of depression 200 CID1 see Composite International Diagnostic Interview cimetidine, and moclobemide metabolism217 circadian system 50 CIS-R 6-7 citalopram244 6
and clozapine plasma levels 225
overdose224 class, socialsee socioeconomic status class, social 34-15 classifications, of affective disorder 1-5 client-centred therapy 142 clinical ~uidelines319-20 Clinical Interview Schedule 307 Clinical practiceg ~ j ~ e l i n e s : ~ e ~in~ e s s i o n ~ r i ~care a i312 ~ ~ clinical trials 191-2 clinics, sessional 335 clomipramine190,244 for children 354 clonidine 59,60,298 clozapine in bipolar disorder 298 effects of SSRIs on 225 cluster B personality disorder 84 CMHTs see community health care teams cocaine278 cog-wheel rigidity 283 cognitive analytic therapy 88 cognitive behavioural therapy (CBT) 165,166 availability 181-2 for children and adolescents3 5 5 4 , 3 5 9 cost-effectiveness 182-3 drop-out rates 183 efficacy160,165-7,184-5 combined with drugs 175,439 compared with pharmacotherapy166-7, 168,174,177,184-5,402-3 compared with psychodynamic therapy 145-6,403,404 cornpared with p s y c h o d ~ a m i c interpersonal therapy 1 5 3 4 , 1 5 5 in in-patient settings 36,171-2 magnitude of treatment effects178 in out-patient settings 170-1 for patients with cancer 38 in primary care 168-70 and relapse/recurrence prevention 180-1,184 and time-course of treatment effects 178-81 for the elderly 4 0 2 4 future developments 1 8 3 4 for personality disorder 433 for physically-ill patients 424,425 in primary care 34,315 suitability of patients for 3 73-7 treatment duration 183 see also cognitive therapy Cognitive Bias ~ u e s t i o ~ a i r100 e cognitive theory 94-7,119-20 causal aspects 103-19,120 clinical implications 118-19 see alsocognitive therapy cognitive negativity and thinking errors 99-100 cognitive triad 94,97-8,129
descriptive aspects 97-103 links between thought, mood and behaviour 100-2 cognitive therapy (CT) 94,118-19,165-6 cognitive negativity and thinking errors 99 and the cognitive triad97-8 compared with experiential therapy156 compared with interpersonl therapy 148-50 and dysfunctional assumptions 119 efficacy from published data144 and link between negative cognitions and mood101,103 in primary care 315 the therapist’s case conceptualization 118-19 vertical exploration technique 119 see alsocognitive behavioural therapy; cognitive theory cognitive triad 9 4 , 9 7 4 , 1 2 9 cohort studies 9-11,12-13 College Student Life Events Schedule 116 community, depressions in the 28-30,38-9 classification 4-5 contact points for patients329-30 ethnic minorities 340-1 identification and assessment 329-30 management 30,39 by community teams see community health care teams and patient compliance 337-8 surveys 5,9,11,12 women 3 3 9 4 0 community health care teams (CMHTs) 3 2 5 4 bed management 332-3 case management 331-2 consultation services 334-5 crisis services 334 day care services 334,339 home care 332,340 identifying people with mental health disorders 329-30 interventions in primary care336-8 liason with other agencies and primary care 335-6 needs assessment system330-1 and rehabilitation programs 334 risk management 338-9 special neeeds groups 3 3 9 4 0 strategies 326-9 see also community, depressions in the community psychiatric nurses (CPNs) see nurses comorbidity188,201-2 in children 348 in resistant depression 4 3 2 4 compliance see patients, compliance Composite International Diagnostic Interview (CIDI) 6 primary care version (CIDI-PC)30-1 computed tomography 51 computer systems
and diagnosis 1-2 and treatment guidelines 311,320 conceptualization, cognitive 118-19 congenital abnormalities 286 congruent hypothesis 111 see also events, congruent consultation sevices 334-5 conversational strategies for treatment137 ’conversational’ therapy model 88 coping, after depressogenic events78-9 Coping with ~ e ~ r e s s 177 io~ ‘core beliefs’ 9 5 4 , 1 0 5 Cornell Scale for Depression in Dementia 388 corticosteroid hormones 44-6 corticosteroid receptors, effects of antidepressants on 62-3 corticotropin-releasing hormone 47 cortisol 44,45,47,351 inhibition of synthesis 45,444 and patients experiencing severe events 85 cot deaths 368 counselling bereavement 402 in primary care 314,315,317 courses, in cognitiv~behavioural therapy 181-2 CPA (Care Programme Approach) 330,331 CPNs (Community psychiatric nurses) see nurses creativity, and lithium treatment 283 Creutzfeld-Jakob-like syndrome 286 crisis response services 334,338-9 Cross-National Collaborative Group 5 4 CT see cognitive therapy culture341 Cushing’s syndrome 45,46,49,410 and the hippocampus 49 cyproheptadine214,223 cytochrome p450 system 190 inhibition by SSRIs 224-5,244 data psychosocial data collection 70-2 see alsoquestionnaires day care services 334,339 for elderly patients 401 DCRs (Diagnostic Criteria for Research)1,3 ’Defeat Depression’ campaign 311,398 definitions, of disorders 1 delusional depressionsee psychotic depression dementia compared with pseudodementia 385-6 in physically-ill patients 417 treatment of elderly patients with 393 demography, and depression prevalence29, 31 ’denial’, of stressful events 78/79 dependency111,117 and type of depression 104,105,130-3 see also sociotropic individuals depression 188 major, criteria for 191-2
463
Depression Attitude Questionnaire 312 Depression Guideline Panel 33,35-6 depression styles/ types dependent 130-3 self-critical 130,131-2,1334 depressive episode (DE), prevalence3 4 Depressive Experience Questionnaire (DEQ)106,107-9,115,116 Revised116 depressive psychosis, incidence 8,lO 'depressive spectrum disorder' 339 desipramine adverse effects 218,220 antidepressant response to 62 combined with fluoxetine treatment 243 cornpared with CBTin the elderly 403 dosages440 not for children 354 detection, of depression see diagnosis, of depression dexamethosone suppression test (DST) 2644,351-2 dexamphetamine245 diabetes, and antidepressant treatment 203,284 diagnosis, of depression 32 in children and adolescents345-6 computer programs aiding1-2 diagnostic classifications 1-2 in e t h i c minorities 341 and patient outcomes 32 in primary care 30,31,307-l0 see also symptoms Diagnosis and Statistical Manual of Mental Disorders see DSM Diagnostic Criteria for Research (DCRs)1 , 3 Diagnostic Interview Schedule (DIS) 3,5, 7,11 dialysis patients 415 diazepam36 differential activation hypothesis 117-18 difficulties 7 1 , 7 3 4 coping after 78-9 vu~erabilityto 75-8 see also events dihydroepiandosterone46 diltiazem298 DIS see Diagnostic Interview Schedule 'dominant goal' personality type 105,132 see also self-criticism 'dominant other' personality type 105,132 see a2so dependency dopamine53 and bupropion 226 dopamine-blocking drugssee antipsychotic drugs and sex steroid withdrawal 376 dothiepin204 adverse effects 218,220 key reference 206 'double depression' 433 'downplaying', of stressful events 78-9 doxepin
464
adverse effects 218,221 compared with CBT in the elderly 403 in pregnancy and breastfeeding 221,372 drugs antidepressant see antidepressants antipsychotic see antipsychotic drugs causes of organic depressions 387 drug abuse in recurrent affective disorder278 predisposing to lithium toxicity 289 psychostimulant245 steroid-suppressing 2 4 5 4 DSM criteria for depressive episode (DE)3 , 4 DSM-III-R 1 , 3 , 4 DSM-IV 1,4,346,383 diagnosis413 diagnostic criteria formania 276 DST (dexamethosone suppression test)264-5,351-2 dynamic psychotherapysee psychodynamic therapy dysfunctional assumptionssee assumptions, dysfunctional Dysfunctional Attitude Scale (DAS)114,115, 116,119 and cognitive-behavioural therapy 177 dyskinesia, tardive 282,297,445 dysthymia 4,346,433 dystonia223 Ebstein's anomaly 286 ECA studiessee epidemiological catchment area studies ECT see electroconvulsive therapy education, and riskof depression 397 EEGs see electroencephalograms effect size 191,441 elderly patients, depression in delusional (psychotic) depression 391 diagnosis and associated ill health383-4 difficulties in 383-6 laboratory investigations 387-8 and screening questionnaires 388-9 maintaining factors 400-2,415 management 389-90 antidepressants 202-3,390,391,392-5 ECT 265,272,394 psychological 402-5 resistance to treatment 3 9 4 5 social m ~ a g e ~ e 395-402 nt precipitating factors 3 9 8 4 0 0 predisposing factors 396-8 presentation 383,386 electrocardiograms (ECGs), and tricyclic treatment 354,355 electroconvulsive therapy (ECT) adequate ECT stimuli441-2 and brain damage 252 for children and adolescents2 6 5 4 , 3 5 5 discontinuation442 efficacy 38,234,24743,252-72
compared with drug treatment 240, 2524,271 and dosage 259,260-1,262,266-8,272 duration of clinical effect 262-3 early trials 252 in the elderly 265,272 and frequency of treatment 269-70,272 likelihood of recovery 271 practical issues 271-2 and pretreatment medication 247-8, 268-9 real versus simulated ECT25642,271 unilateral versus bilateral 26743,270, 272,442 in young people 265-6 for elderly patients 265,272,394 fixed high dose 442 geographical usage 252 for hospitalized depressions 38 and 5-HT 55,56 indications for 247, 271,272 non-response to 270,442 for physically-ill patients 421-3 for psychotic depression 445 and regional cerebral blood flow 51 for resistant depression 24743,443, 445,453 selection of patients for 263-5 and subsequent relapse 248 electroencephalograms (EEGs) abnormalities in adolescents 352 inthe elderly 388 emotional disorder cognitive theoryof 94-7 see alsocognitive theory emotional functioning, and self-esteem 134-5 employment status, unemployment16-17 ’empty chair’ technique 156,157 EMS (Eosinophilia Myalgic Syndrome) 241, 444 endocrine disease, and use of antidepressants 422 endocrine system, and causes of organic depressions387 endogenous depression 74-5 and cognitive-behavioural therapy 1 7 3 4 effects of antidepressants on 191,199,200 entrapment 86 depressogenic 73,74 environment, causes of depression see aetiology of depression, environmental Eosinophilia Myalgic Syndrome (EMS) 241, 444 epidemiological catchment area (ECA) studies 3/11 and social class 14-15 epidemiology 5-21, 29/31 in children and adolescents 348-9 of depression associated with physical illness 410-13 see alsoage; frequency, of depressive disorders; gender
epileptic patients and atypical antidepressants 229 and MAOIs 215-16 and moclobemide 217 and SSRIs 225 and TCAs 221 and use of antidepressants419,422 epileptics, and antidepressant treatment 203 ethnic minorities 340-1 depression prevalence 29 see alsorace events congruent, and vulnerable personality types 110-13,114-16, 119 depressogenic 72-3 coping after 78-9 and early life experiences 133 origins 84-5 vulnerability before 75-8 ’fresh start’ 79,82,86,438 life events 52,70-1,133 and bipolar disorder 277-8 bringing hope 79 and cortisol secretion 45 and depression in the elderly 398-9,401 loss 73,398-9,416 see alsodifficulties; entrapment; humiliation; loss experiences see difficulties; events experiential therapy 142-3,155-8,160 defining characteristics 142 systematic reviews 1 4 3 4 eye disease, and treatment of depression 219,422
facilitators/facilitation 311-12,319 families assessed in childhood depression358 depression in 135,350-1 effects of physically-ill membersin 417 family therapy 357-8 for elderly patients 399,404-5 members understanding depression438-9 of those with resistant bipolar disorder 448 see alsogenetics; marital relationships; spouses Family Attitudes Questionnaire 109 Father/Mother Trait Scale 108 fathers and mothers’ postnatal depressions368-9 see alsoparenting fatigue31 chronic fatigue syndrome 414,423 Feeling good (David D. Burns) 170 Feeling goodh a ~ d b o oThe ~, (David D. Burns) 170 fenfluramine 48/57 d-fenfluramine57 ’filters’, to treatment 27 fluid retention 284-5 fluoxetine adverse effects 222,224-5 and breastfeeding 225-6
465
fluoxetine (Cont.) combined with perphenazine238,445 pindolol247 TCAs 243,244 delay in action 199 dosage196,199 duration of treatment 441 efficacy 205,244 for adolescents 353 interactions 224-5 key reference 206 lithium augmentation of 36 long-term trials 194-5 for physically ill depressives 393-4 for postnatal depression 374 and suicide 204-5 withdrawal224 fluvoxamine57 adverse effects 222,224-5 combined with moclobemide 243 interactions 224-5 key reference 206 flying, and bipolar disorder 278 focusing 142 food interactions,MAO1 214-15 frequency, of depressive disorders and high risk groups 13-21 incidence 8-9,329 with age 19 postnatal365 measurement 5-9,10,11,329 over time 9-13 prevalence changes over time9-13 in children and adolescents 348 of depression and mental disorders28-9,31 of depressive episode (DE)3-4 in the elderly 396 international locations 6 , 7 lifetime 12-13 measuring 5-7 in the physically-ill 410-12 of postnatal depression 364,365 'fresh starts' 79,82,86,438 friends, of depressed people 130 frusemide283 functional neuroimaging 51-2 GABA (gamma-amino-butyric acid) gamma-amino-butyric acid (GABA) 295 gastro-intestinal disease, and use of antidepressants 422 gastro-intestinal effects, of lithium treatment285 gastro-intestinal haemorrhage, depression at the time of 414 GDS (Geriatric Depression Scale) 388,389 gender differences in response to distress 339 and elderly patients 396,400
466
and frequency of depression 15-16/19, 28-9,30-l admitted to hospital 8 in children 18-19,348 sex ratio 18,19,20,339,348 and social factors 18,339 and vulnerability 110 see also men; women General Health Questionnaire(GHQ) 307,310 GHQ-28 5 general practitionerssee primary care generic counselling,for childhood depression 357 genetics and aetiology of depression 42-4,55,350-l see also families genito-urinary diseases, and treatment of depression422 gepirone54,57 Geriatric Depression Scale (GDS)388,389 Gestalt therapy 142,156 GH (growth hormone) 47,352 GHQ see General Health Questionnaire glaucoma, and treatment of depression 219,422 Global Chronicity/Severity Index 176 glucocorticoids, secretion of2 4 5 4 glucose metabolism, and lithium treatment284 goals, demanding/unattainable 136 grief see mourning growth hormone (CH) 47,352 guidelines, clinical 319-20 H, blockers 435-6 HADS (Hospital Anxiety and Depression Scale)416 hair, effects of lithium treatment on 284 for HAM-D see Hamilton Rating Scale Depression Hamilton Rating Scalefor Depression (HRSD/HAM-D) 192 and cognitive-behavioural therapy 178 and sedative antidepressants199 hands, tremor 283 Health of the Nation Key Area Handbook 325 health visitors 373 heart diseasesee cardiovascular disease hepatitis219 heritability see genetics heterocyclics effectiveness for hospitalized depressions 38 for patients receiving specialized care36 hippocampus49 corticosteroid receptorsin 49/62 histrionic personalities 84,85 HIV-positive patients antidepressant trials with420 psychotherapy 424,425 home-based care 332,340
hopelessness 98 hormones and causesof depression 467,435-6 corticosteroid 44-6 neuroendocrine changesin depression 47-9 for treatment of postnatal depression 49, 372-3 Hospital Anxiety and Depression Scale (HADS)416 hospitals day334,401 see also secondary care HPA axis, activation 47 HRSD see Hamilton Rating Scale for Depression 5-HT see 5-hydroxytryptamine 5-HTP see 5-hydroxyt~ptophan 5-HTT (5-hydroxytryptamine transporter)55 humiliation, depressogenic 73 5-hydroxyindoleacetic acid 58 5-hydroxytryptamine (5-HT) 52 depletion53 effect of antidepressal~t drugs on 54-9,246, 281-2 5-HT behavioural syndrome and 5-HT syndrome54-5,444,453 neurotoxicity 239,242 precursors 241 and stress 46 5-hydroxytrypta~inetransporter (5-HTT) 55 5-hydroxytr~tophan(5-HTP) 241,242 toxicity243 hypercalcaemia46 hyperparathyroidism284,387 hypertension, and treatment of in the elderly 387 and use of antidepressants42 hyponatraemia219,223 hypotension due to antidepressant treatment229 postural392 treatment causing depression 435 hypothalamic nuclei 49 hypothalamus and biologial clocks 50 hormones47 hypothroidism 47,277,288,293,387,422 ICD-,, 1,3,390 criteria for depressive episode (DE)3 , 4 primary care version(ICD,,-PC) 32 and screening instruments 5 idazoxan246 illness, physicalsee physical illness imidazoline receptors 59 i ~ i n i d o d i b e ~derivatives yl 189 imipramine and cancer patients 37 combined with pindolol247
cornpared with ECT 253-6,268,271 compared with other antidepressants 205,288 compared with psychotherapies cognitive-behavioural therapy 166-7, 403 interpersonal therapy 148,349 development 189 dosage440 and onsetof action 198 efficacy in children 352,353 for elderly patients with dementia 393 for hospitalized depressions 38 and5-HT 58 prophylaxis studies using 194 for treatmentin the community 30 in-patients children and adolescents as359 c o ~ i t i v ~ b e h a v i o u rtherapy al for 171-2, 178-9 medical antidepressant trials with elderly419 detection of depression in412-13 prevalence of depression 411,412 post-natal depressive mothers 370 see also secondary care inactivity97 incidence of depression see frequency, of depressive disorders ’incidents’, life’s 71 infections, causes of organic depressions387 inositol phosphates 56,281 insecurity, experience stemmingfrom childhood128 insight/relational therapy 403 insomnia 31 and bipolar disorder 278 insulin metabolism, and lithium treatment284 ‘internal working theory’ 104,134,138 International Classificationof Diseases see ICD-10 hterparental Influence Scale 108 interpersonal interactions 125 interpersonal therapy (IPT) 130,142-3, 147-52,160 for adolescents (IPT/A) 356-7 compared with imipramine treatment194 defining characteristics 342 efficacy and effectiveness 36,144, 148-52 for HIV-positive patients 425 systematic reviews 143-4 trials of maintenance 151 interventions, for people with depression 328,331 Interview for Recent Life Events 71 introjective depression/traits 132,134 see also self-criticism iproniazid 189,213 ipsapirone S4 IPT see interpersonal therapy irritable bowel syndrome 423
467
isocarboxazid adverse effects 213 drug interactions 215 ketaconazole246 ketanserin55 kidneys, and lithium treatment282-3 ’kindling’ effect 75,277-8,292 LA see lithium, augmentation laboratory investigations, for causes of depression in the elderly387-8 lactation see breastfeeding lamotrigine, for recurrent affective disorder 296-7 lecithin298 LEDs (Life Events and Difficulties Schedule) 71/73 leucocytosis285 life eventssee events Life Events and Difficulties Schedule (LEDS)71,112 Life Experience Scale 116 Life Experiences Survey 116 lifespan modelof depression 80-5 used for therapy 87-8 light treatment, for SAD patients448-9 limbic system function 47-50 and sleep deprivation 49-50 linkage analysis, genetic 43-4 lithium augmentation (LA) 56,234 for children and adolescents 354 for elderly patients 394 for hospitalized depressions 38 and 5-HT neurotransmission 54/58, 281-2 of MAOIs 239,443,452 in mental health care 36 for resistant depression 23841,245 of SSRIs 36,239,443 of TCAs 36,239,443,450 combined with antipsychotics 283,290,297 carbamazepine245,294,447 valproate296 during pregnancy 447 and 5-HT 55,56,58,281-2 mode of action 55/56! 280-2 for physically-ill patients 421,422 and rapid cycling bipolar disorder 277 for recurrent affective disorder 278-92,447 adherence to treatment291-2 alternative prophylaxis 292 antidepressants and lithium 288 contra~dications285-6 drug combinations 245,283,290,294, 296,297,447 mode of action280-2 monitoring of lithium levels287-8 natural outcome 291 patient selection 287
postpartum377 for prophylaxis 292,377 response prediction 280 side-effects282-5,291 toxicity286-7,288-90,294 withdrawal of lithium 279,290-1 withdrawal279,290-1,447 liver failure, anduse of antidepressants 422 lofepramine adverse effects 218 for elderly patients 203,392 key reference 206 lithium augmentation of 36 overdose220 versus placebo 33 loss, depressogenic 73,398-9,416 Lundby Study 12 magnetic resonance imaging(MH)51 maintenance treatment 151 major depressive disorder (MDD) categorization 3 , 5 prevalence 6 treatments compared 192,194 management of depression see treatment mania and antidepressant medication 192 carbamazepine292 lithium 279,280,291 lithium withdrawal mania 279, 291 diagnostic criteria for 276 prevalence from community surveys 11 triggered by drugs 278 see also bipolar disorder manic depressive disorder see bipolar disorder MAOIs see monoamine oxidase inhibitors maprotilene 189 dosage 196 MARIS (monoamine reuptake inhibitors) 288,299 marital assessment/therapy 438-9 marital relationships 80,130,136,137-8,438 and postnatal depression 368-9 supportive 137 see aEso families; marital status marital status and bipolar disorder 18 in different cultures 16 and elderly patients 396,400 marital termination 15 and postnatal depression 369 and risk of depression 15-16,396 and women’s employment 17 see also marital relationships marriage see marital relationships; marital status MDD see major depressive disorder melancholic-psychotic depressions 75 rnela tonin 61-2 memory, and lithium treatment 283 men depression prevalence 29,30-1
see also gender menopause19,20 mental disorders, prevalence 28 mental health services and depressive illnesses 35-8 structural components 326 see also secondary care metabolism altered by lithium treatment 284 and causes of organic depressions387 in the elderly 387,391 methadone434 methylphenidate245,418 methysergide298 metyrapone246 mianserin37,205 adverse effects 226-7,229 mice, fearfulness 44 Midtown Manhattan Study12 midwives375 migraine, and use of antidepressants 422 milnacipran227 'minimization', of stressful events 79 mirtazapine, adverse effects 228,229 moclobernide 189 adverse effects 213,216-17 in the elderly 392 combined with pindolol247 SSRIs 243 for elderly patients with dementia 393 in epileptics 217 interactions217 key references 206 overdose217 in pregnancy and breastfeeding 217 withdrawal217 monoamine oxidase inhibitors (MAOIs) 53, 189,234 adverse effects 197,204,212-16 for atypical depression 192-3, 449 changing antidepressants to 443 combined with hydroxytryytophan242 lithium 239,443,452 psychostimulants245 SSRIs 225,243 242-3,451 TCAS tryptophan 58,242,444,452 delay in action 198 dosage196,197 drug interactions 215 efficacy for children and adolescents353-4 compared withECT 253,254-S for hospitalized depressions 38 for patients receiving specialized care36 and predictionof response 199 for elderly patients 394-5 for epileptics 215-16 food interactions 214-15 and 5-HT neurotransmission 55,56,57, 58,246
'
i
not for physically-ill patients421,422 overdose214 in pregnancy and breastfeeding 216 reversible see reversible inhibitors of monoamine oxidase withdrawal201,214 and change toSSRIs 225 see also isocarboxazid; phenelzine; tranylcypromine monoamine reuptake inhibitors (MARIs) 288,299 Montgomery-Asberg Depression Rating Scale 192 mood and dysfunctional attitudes 117 relationship with cognition and behaviour 100-2 mother and baby units371 mothers and childcare 16 loss by children 80-1,126,128 mother-child interactions of depressed136 overprotectiveness 107 perfectionism 109 single 15 see also care-givers, primary; childbirth; parenting motivation, low 97 mourning, and interpersonal therapy147-8 Multidimensional Perfectionism Scale 108 'multifaceted care' (MC) 33 multiple sclerosis 419 and cognitive behaviour therapy 425 myasthenia gravis 285 myxoedema46,49 NA see noradrenaline nails, effectsof lithiurn treatment on 284 narcissism153 National Comorbidity Survey, United States (NCS) 6/7/11 and social class 14 National Instituteof Mental Health see NIMH National Survey of Psychiatric Morbidity (British) 4 , 6 , 7 and sex ratio changes with age 19 and single mothers 15 and social class 14 nationality and prevalence of depressive disorders 5-9 see also ethnic minorities; race natural killer cell activity( N U ) 37 nausea and atypical antidepressants 227 and SSRI treatment 222 needs, of those with mental health disorders328-9,330-1 nefazodone adverse effects 229 interactions 229 key reference 206 withdrawal229
469
Negative Evaluation of Self (NES) 78 children and adolescents as 359 and cognitive behaviour therapy170-1, see also self-esteem, low negativity, in thinking 99-100 178-9 detection of depression in 413 NES (Negative Evaluationof Self) 78 prevalence of depression 411 neurochemistry, of depression 52-3 see also secondary care neuroendocrine changes in depression outcome, of depressions encountered in 47-9,52 primary care 32 neuroimaging, functional 51-2 overprotectiveness, parents’ 107,109 neuroleptic drugs oxaprotiline 61-2 for treatment-resistant psychotic depression445 see also antipsychotic drugs; clozapine; panic disorder 43,434 perphenazine; risperidone parachlorophenylalanine 58 neuroleptic malignant syndrome 290 parathyroid, effects of lithium treatment neuropeptides53 on 284 neurotic depressions 75 Parent-Child Relations Questionnaire 109 neuroticism 44 Parental BehaviourForm 108 nifedipine298 Parental Bonding Instrument 108 NIMH Parental Bonding Inventory 109 Collaborative Program on the PsychoParental Inconsistencyof Love Scale 108 parenting biology of Depression 200 inconsistent 132-3 Consensus Development Panel 33 TDCRP see Treatment of Depression negative132 Collaborative Research Program and the attachment system 126-7,131, nimodipine298 132-3,134 nomifensine205 see also children, adversity in non-neurotic depressive disorder, positive 134 incidence 8 and vulnerable personality types 107,108 non-steroidal anti~inflammatorydrugs, and see also fathers; mothers parkinsonism lithium toxicity 289 noradrenaline (NA) 52-3 and antidepressant treatment 203,223,283 see also Parkinson’s disease depletion53 Parkinson’s disease 410,419,422 effects of antidepressant drugs on59-62 see also parkinsonism and stress 46 nortriptyline paroxetine adverse effects 222,223 adverse effects 218,219 combined with, pindolol 247 combined with dosage 199 fluoxetine243 sertraline243 key references 206 compared with IPT for elderly patients 404 long-term trials with placebo295-6 withdrawal224 efficacy in children 352-3 partners, of depressed people see spouses for physically-ill depressives 393,418 patients nuclei, hypothalamic 49 nurses compliance197,337-8 community psychiatric (CPNs) 315,317, in physically-ill patients 414-15 335-6 in primary care 311,314,320 for elderly patients 398,401-2 in resistant depression 4 3 9 4 0 general practice 314,318,319 ’refractory’ 199 pemoline245 obsessive compulsive disorder 49 perfectionism 109 with resistant depression 433-4 PERI71 oedema285 period effects, on prevalence11 oestrogen 46,49 periodicity, of depression 50 and postnatal depression 372-3,376,377 perphenazine238 receptors49 in drug combinations 238 offenders, mentally disordered 339 personality disorders Office of Population and Census National and cognitive-behavioural therapy 176 Morbidity survey 329 with depression 433 ondansetron 55 histrionic/dral~atic84,85 OPCRIT1 Personality Organization 131,133 opiate addicts 434 personality types organic depressions, causes 387 and antidepressant therapy 203 out-patients and cognitive-behavioural therapy 176 ,
vulnerable and adversityin childhood 104-10 and congruent events 110-13,114-16, 119 see also depression styles/types PES (Positive Evaluation of Self) 78 PET (positron emission tomography) 51,52 phaeochromocytoma, and use of antidepressants422 pharmacological treatments see drugs pharmacotherapiessee drugs phenelzine 34 adverse effects 213,214,215 combined with pindolol 247 compared with ECT 254-5 dosage196 for elderly patients 39&5 phenytoin interaction withSSNs 225 interaction with TCAs 220 phobias43 phosphatidyl inositol 56,281 physical illness, depression in 410,425-6, 434-5 assessment 3834,415-17 in elderly patients 3 8 3 4 , 3 9 3 4 , 4 0 1 epidemiology 410-13 from treatmentfor medical illness 435-6 hospitalized patients with depressions 36-8 and presentation of mental disorders 31 in primary care 30-2,307 and prognosis of physical illness 414-15 treatment of depression 417-18, 419-20,422 antidepressants37,203,417,418-21 antidepressants for the elderly 3934 ECT417,421-3 psychological interventions 423-5 social measures 425 physostigmine298 PI see psychodynamic-interpersonal(PI) model of depression pindolol54 in drug combinations 246-7 with SSRIs 58-9,234,247,444,451-2 placebos39,190-1 in community cases 30 in hospitalized depressions 38 in primary care 3 3 4 , 3 5 Plan Formulation method 147 plasma concentrations,of antidepressant drugs190,205 porphyria, and use of antidepressants 422 Positive Evaluation of Self (PES)78,439 potassium, serum levels 387 predictors of depression see risk, of depression pregnancy premarital 80-1 treatment during lithium447 MAOIS 216
moclobemide217 SSRIs 225 TCAs221 Present-State Examination (PSE)4,5 PSE-ID-CATEGO system 5 prevalence, of depression see frequency, of depressive disorders primary care classification of depressive disorders in 4-5 management of depression 32-5,307-8, 320-1 antidepressants 32,334196,312-14, 320 'counselling' 314,315,317 detection of depression 30,31,32,307-l0 and facilitators 311-12,319 liason withCommunity health care teams 335-6 multidisciplinary teamwork 312, 317-19,321 and patientattitudes 311,314,320 and practice nurses 314 primary-secondary care interface319-20 psychotherapies34,35,168-70, 314-15,321 social treatments 314,31617,321 strategies 310-12 mental illness caseload 329 patients' needs in 328-9 patients presenting somatic symptoms to30-2,307 sessional clinics 335 training for care teams 337 see also community health care teams problem-solving 313,315 in elderly patients 403 for physically-ill patients 424 Process Experiential therapy142 progesterone and postnatal depression 372,373,376 withdrawal46 prolactin 48,282 propanolol283 prophylaxis see recurrence of depression, prevention; relapse, prevention protriptyline, adverse effects 218 provoking agents for depression 72-5,437-8 see also aetiology PSE see Present-State Examination pseudo-unipolar disorder 275 pseudodementia 3854,436 psoriasis284 psychiatrists collaboration withGPs 308,321 compared toGPs on treatment strategies 313-14 psychiatric referral for perinatal women370 see also secondary care psychoanalysis 144 psychoanalytic theories of depression125-6 psychodynamic therapy 142-5,159-60
471
psychiatrists (Cont.) clinical trials 145-6 efficacy from published data 144 for elderly patients 404 compared with CBT 403,404 mechanisms of change in146-7 and physically-ill patients 424 see also psychod~amic-interpersonal model of depression psychod~amic-interpersonal (PI) model of depression and attachment theory 125-38 therapy130,142,152-5 psychoses incidence 8,lO postpartum365 psychosocial modelof depression and data collection 70-2 lifespan model of depression80-5 and management of depression79-80, 85-90 provoking agents for depression72-5, 437-8 see also difficulties; events vulnerability to depressogenic events75-9 psychostimulants234,245 psychosurgery248,444-5 psychotherapy 158-9 counterproductive and useful150 for the depressed physically-ill 37-8, 423-5,426 for post-natal depression 3 7 3 4 in primary care 34,35,168-70,314-15,321 in specialist care 36 systematic reviews 1 4 3 4 psychotic (delusional) depression in the elderly 391 incidence 10 treatment 192,199 treatment resistant 445 puberty20,358 see also adolescents pulmonary disease, and antidepressant trials420 pulmonary embolus, depression at the timeof 414
brief recurrent depression 449 in children and adolescents349 prevention using antidepressants 1 9 3 4 using cognitive-behavioural therapy 180-1 versus relapse 201 see also recurrent affective disorder; relapse recurrent affective disorder 275-8 drug treatments 3 9 3 4 , 2 9 9 antimanic drugs 298 antipsychotic drugs 297-8 carbamazepine 2 9 2 4 lamotrigine 296-7 lithium 278-92 thyroid hormone 299 valproate 295-6 recurrence versus relapse 180,201 see also bipolar disorder; recurrence of depression regional cerebral bloodflow (rCBF) 51,52 rehabilitation patients 334 antidepressant trials with 420 relapse 111,117 in children and adolescents349,360 compared with recurrence 180,201 see also recurrence and congruent events 113,119 and long-term antidepressant treatment 193 prevention using antidepressants 194 using cognitive-behavioural therapy180-1,184 rates35 risk of 35,200-1,400 relationships personal128,129-30 see also marital relationships relaxation training, for children and adolescents356 remission200 in bipolar disorder 192 renal dialysis patients 415 Research Diagnostic Criteria 364 residence, urban versus rural 396 resistant depression, definitionof 431 questionnaires resistant depression management aiding diagnostic classifications 2 aetiology reassessed 436-8 for the elderly 388-9 antidepressants234-6 psychosocial checklists 71,72 adequacy of previous treatments self-report346 assessed 4 3 9 4 2 augmentation 23744,443,450-3 race 'cocktails' 453 and depression in the elderly 396 decision pathways for 449-52 experimental treatments 444 see also ethnic minorities; nationality high dose TAD treatment 444 rapid cycling illness 277,299,447 ~AOI-tryptophancombination 444 rCBF (regional cerebral blood flow)51,52 switching235-7,442-3 reappraisal, of unpleasant situations 74 assessment of diagnosis and time course receptors, and mechanisms of drug action 53 of depression 431-2 recovery35 bipolar affective disorder 446-8 and changein cognitive measures 117 brief recurrent depression 449 recurrence of depression
Index and comorbidity 432-6 ECT 2474,443,453 experimental treatments 444 pharmacological (with limited efficacy) 244-7 psychological439 psychosocial 438-9 psychosurgery248,444-5 psychotic depression 445 seasonal affective disorder 448-9 respiratory effects,of lithium treatment 285 reversible inhibitorsof m o n o a ~ i n eoxidase (RIMAs)206,212 see alsobrofaromine; moclobemide Revised Clinical Interview Schedule 310 Revised Depressive Experience Questionnaire (DEQ) 116 RIMAs see reversible inhibitorsof monoamine oxidase risk of depression in children and adolescents 348-9 in elderly patients 396-8 high risk groups 13-21 in the physically-ill 412 of suicide 338-9 risperidone, and bipolar disorder 298 Roles and Goals Questionnaire 116 romantic relationshipssee relationships, personal
psychotherapy36 security experience stemming from childhood128 felt 127-8 sedation, caused by antidepressants 218, 222-3,229 selective serotonin re-uptake inhibitors (SSRIs) 189 adverse effects 36,55,197,204,221-6 in elderly patients 392 for cancer patients 37 changing toa MAO1 443 for children and adolescents353,354, 355,360 choice206 combined with antipsychotic drugs 238 lithium 239,443,451 MAOIs225,243 moclobemide 21 7 pindolol 58-9,234,247,444,451-2 TCAs 2 4 3 4 , 4 5 2 comparative efficacy 191,2054,288 delay in action 198,199 dosages196,197,440 dropout rates 201 efficacy for children and adolescents 353 for hospitalized depressions 38 for patients receiving specialized care 36 and prediction of response 200 SAD (seasonal affective disorder), light for the elderly 203,391,392 treatment 448-9 in epileptics 225 SCAN (Schedules for Clinical Assessment in and 5-HT function 55-9,61,246 interactions224-5,391 Neuropsychaiatry) 5,12 long-term trials 194-5 Schedules for Clinical Assessment in and mania 288 Neuropsychaiatry (SCAN) 5,12 and NA neurones 61 'schemata' 96 schizophrenia9 overdose224 pharmacokinetics 190 and cognitive-behavioural therapy 173 and in-patient bed use 332 for the physically-ill 418,419,420,421,422 seasonal affective disorder (SAD), light in pregnancy and breastfeeding 225-6, 371-2 treatment 448-9 in primary care 313 secondary care reviews and key references 206 and cognitive-behaviour therapy 170-2, 178-9 and suicide 204,223 withdrawal224 detection of depression in the physicallyill 412-13 and changeto MAOTs 225 see also citalopram; fluoxetine; fluvoxamine; hospitals paroxetine; sertraline bed management 332-3 day care 334 selective serotonin/noradrenaline re-uptake depressions admitted to 38 inhibitors (SNRIs) first admission rates to 8 and 5-HT neurones 61 and lengthof stay for physical and NA neurones 61 illnesses414 reviews and key references 206 mother and baby units371 see also venlafaxine physically ill patients with self-complexity 112 depressions 36-8 greater 112 see also in-patients; out-patients self-criticism (self-blame) 78,109,113 interface with primary care319-20 and attachment theory 130,131-2,1334 patients ineligible for 327 and interpersonal therapy 149 and responseto treatment with self-esteem drugs35-6 high439
473
self-esteem (Corzt.) low 76/77! 78,439 and attachment theory 134-5 and cognitive theory 96-8,106 see also cognitive triad self-harm, deliberate 385 serotonin see 5-hydroxytryptamine (5-HT) serotonin reuptake inhibitor and antagonists reviews and key references see also nefazodone serotonin syndrome 215,217,225 serotonin transporter gene 44 sertraline244 adverse effects 222,224 combined with nortriptyline243 pindolol247 key reference 206 sex see gender sexual abuse,of children 45 sexual dysfunction, causedby antidepressant drugs219,223,228,285 Sheffield Psychotherapy Project, second (SPP,) 130,152-5,160 skin and carbamazepine treatment 293 and lithium treatment 284 sleep abnormalities in children 352 deprivation and bipolar disorder 278 and regional cerebral blood flow51 disturbance 49-50 SNRIs see selective serotonin/noradrenaline re-uptake inhibitors social adversity 438 and bipolar disorder 448 and postpartum depressions 366 see also events social characteristics, and depression prevalence 29/31 social classsee socioeconomic status social measures/treatments in old-age depression 395-402 in physically-ill patients 425 in primary care 314,316-17,321 Social Readjustment Rating Scale 71 social skills training, for childhood depression357 social supportsee support, social social workers, and elderly patients 398, 401-2 socioeconomic status, and depression14-15, 396,414 sociotropic individuals 105,110,111,112-13, 120,132 see also dependency Sociotropy-Autonomy Scale(SAS) 106-7, 111,114,115,116 somatic symptomssee physical illness somatization 307,411 specialist care treatment in 35-6
see also secondary care Spectrum of Care 325 SPET (single photon emission tomography) 51 spireprone 55 spouses of depressed people 130,133,438-9 see also families; marital relationships SPP, see Sheffield Psychotherapy Project, second SSRIs see selective serotonin re-uptake inhibitors SST (stereotactic subcaudate tractotomy) 52, 444-5 stereotactic subcaudate tractotomy (SST) 52, 248,444-5 steroid hormones 435-6 steroid-suppressing drugs 245-6 Stevens-Johnson syndrome 293 Stirling County Study 12 Strengthening the Primary CareTeam 308 stress environmental hormonal responseto 44-6/48 see also difficulties; events, depressogenic StrictControl, Conformity and Achievement Control subscale 108 stroke 411,414,420 substance abuse, with resistant depression434 succinyl choline 285 suicide 98,200 amongst Asian women341 and antidepressants 204-5,220,223 in bipolar disorder cases 277,278 in children 350 postnatal365 risk factors 338-9,417 support from the therapeutic service 88-90 social78,179-80,340,438 and depression in the elderly 397,400 poor76-7/87 'suppression', of stressful events 78-9 susceptibility to depression see risk, of depression Symptom Checklist-90 109 symptoms and diagnostic classes 1,2 physical see physical illness see also diagnosis syndromes 1
T,, augmentation with 443,450 Taiwan, community surveysin 12 tardive dyskinesia 282,297,445 TCAs see tricyclic antidepressants TDAs see tricyclic antidepressants (TCAs) TDCRP see Treatment of Depression Collaborative Research Program teeth, effectof lithium treatmenton 284 temperaments see personality types
therapists, trainingin cognitiv~behavioural for patients receiving specialized care36 therapy 181-2 and prediction of response199 thiazide diuretics, and lithium toxicity 289 for the elderly 203,391,392 thinking, cognitive errors and in epileptics 221 negativity 99-100 historical introduction 189 thoughts, automaticsee automatic thoughts and 5-HT neurotransmission 55,56,57,58 thyroid gland interactions220 changes in depression 47 overdose220 effects of lithium on 282 for the physically-ill function tests 387 in pregnancy and bre thyroid hormones in primary care 313 receptors49 reviews andkey references treatment with 49,299 sudden deathsin chi1 thyroid-stimulating hormone (TSH) 47, withdrawal201,220 281,282 see also amitryptiline; clomipramine; thyrotoxicosis288 desipramine; dothiepin; daxepin; thyroxine 49,299 imipramine; nortriptyline; and TCA augmel~tation 241 trimipramine triiodothyronine (T,) 2 time, of risk 11 time course,of depression 432 in TCA-resistant dep trimipramine tinnitus384 training, therapists’, in cognitive-behavioural adverse effects 218 therapy 181-2 dosages440 tranylcypromine TRP see tryptophan tryptophan (TRP) 48,57,298 overdose214 in drug combinations 241-2 withdrawal214 with lithium 281,282 trazodone205,243 with MAOIs 58,234,242,444,452 adverse effects 228,229 with SSRIs 242 for elderly patients 203,392 with TCAs 234 Treatment of Depression Collaborative key reference 206 Research Program (TDCRP) 148-50, and NA uptake inhibitors 62 152,166-7,194 and 5-HT synthesis 58,281 tremor TSH see thyroid-stimulating hormone and lithium treatment 283 tyramine 214-15 and tricyclic treatment 219 tricyclic antidepressants (TCAs) 53 ultra-rapid cycling, of bipolar disorder 277 adverse effects 196,201,203,204,218-21 ultradian cycling,of bipolar disorder 277 in children 354 unemployment 16-17 in the elderly 392 unipolar disorder changing to a MAO1 443 genetic mechanism 44 and childhood depression 202,352-3, heritabili~ 424 354-5 pseudo-unipolar disorder 275 combined with urban versus rural residence 396 adrenoceptor antagonists 246 antipsychotic drugs 238 valproate lithium 239,443 in recurrent affective disorder 295-6 and pregnancy 286,296 value-achievers 106 T, 443,450 vasculitis224 delay in action 198,199 vasopressin47 dosages196-7,440 venlafaxine244 for children 354-5 adverse effects 227,229 for the elderly 391 dosages198 high dose 444,450-1 drug interactions 215 used in primary care 313 key references 206 dropout rates 201 withdrawal227 effect size 441 verapamil, and mania 298 efficacy vertical exploration technique 119 comparative 205-6 viamin B,, deficiency 387 compared with ECT 253-6 viloxazine, adverse effects 226,229 compared with psychotherapy314 viral infections 414 for hospitalized depressions 38 vulnerability, to depression 45
vulnerability, to depression(Cont.) and attachment styles 130,131,136 before depressogenic events75-8 and cognitive theory 96,117 vulnerable personality types 104-13 interpersonal 131 self-evaluative 131 weight gain dueto antidepressant treatment in elderly patients 392 lithium284 SSRIs223-4 TCAs 219 loss/gain accompanying depression31 WHO, Study of Mental Disorders in General Medical Settings 31 women 339-40 Asian341 befriending interventions 79-80 with bipolar disorder 18
476
childhood adversity mirroredin later life 80-2 depression prevalence 28,30-l and employment 17 and loss events 73 marital status and child care as determinants of depression1516 menopause 19 postnatal depressionsee childbirth, depression following psychosocial vulnerability factors for depression 75-7 and social relationships 130 see also childbirth; gender work aspects of 16 unemployment 16-17 yohimbine 59 zimeldine 189 Zung Depression Rating Scale 108,116,388