sports medicine Volume 40 Issue 2 February 2010

Sports Med 2010; 40 (2): 95-111 0112-1642/10/0002-0095/$49.95/0 CURRENT OPINION ª 2010 Adis Data Information BV. All r...

Author: adisonline.com

74 downloads 756 Views 1MB 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

DOWNLOAD PDF

Sports Med 2010; 40 (2): 95-111 0112-1642/10/0002-0095/$49.95/0

CURRENT OPINION

ª 2010 Adis Data Information BV. All rights reserved.

Recommendations for Improved Data Processing from Expired Gas Analysis Indirect Calorimetry Robert A. Robergs,1,2 Dan Dwyer3 and Todd Astorino4 1 2 3 4

Exercise and Sports Sciences, University of Western Sydney, Sydney, New South Wales, Australia Exercise Physiology Laboratories, University of New Mexico, Albuquerque, New Mexico, USA Exercise Science, University of Newcastle, Newcastle, New South Wales, Australia Department of Kinesiology, California State University, San Marcos, California, USA

Abstract

There is currently no universally recommended and accepted method of data processing within the science of indirect calorimetry for either mixing chamber or breath-by-breath systems of expired gas analysis. Exercise physiologists were .first surveyed to determine methods used to process oxygen consumption (VO2) data, and current attitudes to data processing within the science of indirect calorimetry. Breath-by-breath datasets obtained from indirect calorimetry during incremental exercise were then used to demonstrate the consequences of commonly used time, breath and digital filter post-acquisition data processing strategies. Assessment of the variability in breath-by-breath data was determined using multiple regression based on the independent variables ventilation (VE), and the expired gas fractions for Based on the oxygen and carbon dioxide, FEO2 and FECO2, respectively. . results of explanation of variance of the breath-by-breath VO2 data, methods of processing to remove variability were proposed for time-averaged, breathaveraged and digital filter applications. Among exercise. physiologists, the strategy used to remove the variability in sequential VO2 measurements varied widely, and consisted of time averages (30 sec [38%], 60 sec [18%], 20 sec [11%], 15 sec [8%]), a moving average of five to 11 breaths (10%), and the middle five of seven breaths (7%). Most. respondents indicated that . they used multiple criteria to establish maximum VO2 (VO2max) including: the attainment of age-predicted maximum heart rate (HRmax) [53%], respiratory exchange ratio (RER) >1.10 (49%) or RER >1.15 (27%) and a rating of perceived exertion (RPE) of >17, 18 or 19 (20%). The reasons stated for these strategies included their own beliefs (32%), what they were taught (26%), what they read in research articles (22%), tradition (13%) and the influence of their colleagues (7%). The combination of VE, FEO2 and FECO2 removed . breath-by-breath variability in incremental and steady-state 96–98% of VO 2 . . exercise VO2 data sets, respectively. Correction of residual error in VO2 datasets to 10% of the raw variability results from application of a 30-second time average, 15-breath running average, or a 0.04 Hz low cut-off digital filter. Thus, we recommend that once these data processing strategies are

Robergs et al.

96

used, the peak or maximal value becomes the highest processed datapoint. Exercise physiologists need to agree on, and continually refine through empirical research, a consistent process for analysing data from indirect calorimetry.

Exercise testing that incorporates the use of indirect calorimetry to calculate whole body rates . of oxygen consumption (VO2), carbon dioxide production (VCO2), respiratory exchange ratio (RER) and energy expenditure has a long history dating back to the late 19th century.[1] Hill[1] and Hill and Lupton[2] further consolidated the scientific base of the application of indirect calorimetry to study body metabolism and physiology during exercise in the 1920s. Hill introduced the scientific community to the presence of a plateau . in VO2 during discontinuous incremental exercise testing, . the measure of the maximal rate of . VO2 (VO2max), and provided at that time an interpretation, based on logic and not empirical evidence, of the central cardiovascular and . pulmonary determinants of VO2max. Numerous reviews and both supportive[3-7] and critical commentaries[8-11] have been written on the historic work of Hill and his contributions to the academic field and developing professions of exercise physiology and the exercise and sports sciences. Within the first 60 years after Hill’s pioneering work, researchers identified the need to establish . criteria to classify VO2 responses to incremental exercise as being truly maximal.[12-17] The most notable and influential of these studies was published in 1955 by Taylor et al.[17] In this study, men (n = 115) were required to complete multiday discontinuous treadmill testing characterized by both walking and running. To confirm attain. ment. of VO2max, these scientists used a change in VO2 <150 mL/min or 2.1 mL/kg/min with a final 2.5% increase in grade at a speed equal to 7.0 miles per hour (11.3 km/h). This proto. col stage increment has a VO2 demand of 4.23 mL/kg/min,[18] meaning that this criterion accepts that any increment <50% of the metabolic demand of this protocol stage was accepted as a . VO2 plateau. Results showed that . 108 out of 115 men met the aforementioned VO2max criterion. ª 2010 Adis Data Information BV. All rights reserved.

However, this criterion was not immediately adopted by all exercise physiologists, as only 4 years later Wyndham et al.[19] argued against it, stating that ‘‘no simple criterion to define the level of maximum O2 intake will suffice, such as that proposed by Taylor.[17]’’ In fact, Wyndham and colleagues, as . early as 1959, favoured the use of curve fitting VO.2 versus work rate to confirm the attainment of VO2max. Despite efforts by numerous exercise physiologists to more validly and objectively process data from indirect calorimetry, especially in. re. lation to the measure of VO2max and/or a VO2 plateau,[20-37] there currently exists no universally . recommended procedures for processing VO2 data acquired from breath-by-breath indirect calorimetry, or from time-averaged systems. This is all the more alarming given that current technological advances in indirect calorimetry allow practitioners and researchers to acquire several hundred datapoints per incremental test to maximal exertion. As many data processing techniques used today have a history dating back to the 1950s, prior to the electronic revolution in the biological sciences, how applicable are these ‘older’ or more ‘traditional’ data processing methods to today’s breath-by-breath conditions? It is our firm belief that a lack of consistency in data processing breath-by-breath data from indirect calorimetry can only decrease the validity at which data is processed, which in turn can only lead to errors in data presentation and interpretation. Consequently, given the increasing use of breath-by-breath indirect calorimetry in education, research and professional practice, we find the lack of any objective criteria to follow when processing data to be problematic, and constraining to further research and interpretation of whole body oxygen consumption. In this commentary, we review prior research and reviews on data processing in indirect Sports Med 2010; 40 (2)

. VO2 Data Processing

calorimetry, and present empirical evidence for the positive and negative attributes of different data processing strategies. We then provide recommendations for data processing based on specific types of systems used in indirect calorimetry, as well as raise important suggested solutions for then identifying maximal values, such as . VO2max. Once the field . of exercise physiology agrees on methods of VO2 data processing, and consistently uses these methods in research and peer-reviewed publication, then added topics . such as how to define and detect a VO2 plateau, . whether there are slope changes in the VO2 response to incremental exercise, what . are the most valid protocols for measuring VO2max or what is the data suited to . . difference between labels of VO2max versus VO2 peak can be readdressed and hopefully resolved. 1. Clarification of Terminology Before we commence with the more detailed content of this commentary, it is important to pre-define specific terms as well as explain difficulties with such definitions. Past research of the . VO2 acquired during .maximal. exercise testing presents terms such as VO2max, VO2 peak and the . VO2 plateau. As we will argue, there are no universally accepted criteria for defining any of these measures within exercise physiology, sport and exercise. science. Consequently, we will use the . terms VO2max and VO2 peak simply as labels for terms currently used in research to identify the . incremental exerhighest VO2 recorded during . cise. We will use the term VO2 plateau to identify the concept, first proposed by Hill, for a . leveling in the VO2-time response despite continued increases in exercise intensity. Defining and recommending procedures used to measure such terms is not the purpose of this commentary, but clearly we hope that such direction can be proposed with greater empirical support based on the data, explanations and recommendations we provide in subsequent sections. We also refer to breath-by-breath and timeaveraged data processing, with unavoidable inference to breath-by-breath and mixing chamber systems of expired gas analysis indirect caloriª 2010 Adis Data Information BV. All rights reserved.

97

metry. It is important to recognize that this commentary is not about different methods of data acquisition, but rather is focused on methods of post-acquisition data processing. Nevertheless, it is true that most (but not all) commercial systems used in research that are based on an expired gas mixing chamber configuration employ time averaging. Some mixing chamber systems, such as the Parvomedics (Sandy, UT, USA) have a breath-by-breath option, and therefore can provide breath average and time average options for data processing and can even be used to apply digital filtering. Conversely, use of breath-by-breath systems, regardless of the type of hardware and computer software used to acquire and process data, can be processed by: time averaged, breath averaged, or more sophisticated digital filtering. Once again, it is important to understand that this commentary does not concern itself with differences in the validity of different methods of data acquisition. Rather, the scope of this commentary is confined to different methods of data processing. We direct the reader to a relatively recent review by Macfarlane[38] for an assessment of the validity of different methods and instrumentation used in expired gas analysis indirect calorimetry. 2. How do Researchers Currently Process . Oxygen Consumption (VO2) Data? We wanted to present an in-depth assessment of how researchers currently address the issues of data processing from indirect calorimetry, and how such data. processing. is used to detect the . VO2 plateau, VO2max and VO2 peak, and present this data to support our rationale for this review. We identified two potential methods for this content: (i) a summary of published research, and (ii) a survey circulated via the Internet to as many exercise physiologists as possible. We chose the latter due to the absence of many details of data processing and important definitions concerning . . VO2max and the VO2 plateau in most journal [24] publications. . For example, Whyte et al. mentioned the VO2 plateau as a criterion for . detecting . VO2max, but did not define what the VO2 plateau was or how data was processed to reveal it. Sports Med 2010; 40 (2)

Robergs et al.

98

. Similarly, Day et al.[25] studied the VO2 plateau . and VO2max but did not define either entity, or explain how they were determined. In addition, our review of 298 manuscripts concerning . VO2max testing in healthy, adult populations revealed that 60% of authors failed to . denote criteria used to verify attainment of VO2max. Clearly, exercise physiology research is replete with vague coverage of important data proces. . . sing features for VO2max, VO2 peak and the VO2 plateau. An invitation to complete an Internet-based questionnaire was sent to all members of an international sport science email discussion list (www.sportsci.org). Participants were restricted to those who actually conduct exercise tests and . analyse VO2 data based on self report. Respondents were asked to answer questions about . the method of processing VO2 data that they used most often. A total of 75 questionnaires were completed by physicians (3%), and individuals with a Masters degree (25%) or PhD (64%) in exercise physiology (81%), and who worked in a university or corporate setting and performed exercise testing with indirect calorimetry (79%). Nearly half of all respondents (48%) collected expired gases using the breath-by-breath technique, while only one-quarter (25%) used a mixing chamber. The remaining proportion of respondents (27%) used either technique depending upon the purpose of the test. The strategy . used to remove the variability in sequential VO2 measurements varied widely, and consisted of time averages (30 sec [38%], 60 sec [18%], 20 sec [11%], 15 sec [8%]), a moving average of five to eleven breaths (10%), and the middle five of seven breaths (7%). The remaining 8% of respondents used even shorter durations or rolling averages of breath-by-breath data. Almost all respondents checked for a plateau . made in VO2 (93%), but not all . . respondents (76%) a distinction between VO2 peak and VO.2max. Of those that did check for a plateau in VO2, the most commonly used criteria for a plateau was an . increase in VO2 of <150 mL/min (34%), followed by its variant – an increase of <2 mL/kg/min (27%) and a subjective visual inspection (18%). The remaining 19% of respondents used a wide ª 2010 Adis Data Information BV. All rights reserved.

variety of criteria that included: lower limits of . increase in VO2 with . increasing workload (100 and 50 mL/min), a VO2 slope that is less than half that expected . or not different from zero, and an increase in VO2 that is less than the expected increase. Most respondents also indicated that they . used additional criteria to establish VO2max including the attainment of age predicted maximum heart rate (HRmax) [53%], RER >1.10 (49%) or RER >1.15 (27%) and a rate of perceived exertion (RPE) of >17, 18 or 19 (20%). Finally, throughout the questionnaire, the respondents were asked why they were using these methods. There were a variety of reasons given, including: their own beliefs (32%); what they were taught (26%); what they read in research articles (22%); tradition (13%); and the influence of their colleagues (7%). This was not an exhaustive survey of the practices of all exercise physiologists, but the results indicate that a broad . variety of methods are being used to process VO2 data. The .lack of a consistent approach to processing VO2 data exists despite the fact that this was a relatively homogenous group of scientists with a similar educational background. An especially concerning aspect of the results was that about half of the respondents used their processing method because of subjective (beliefs, tradition, colleagues, 52%) rather than objective influences (education, research, 48%). 3. Problems .and Concerns with Processing VO2 Data 3.1 . Current Methods Used for Processing VO2 Data

The diverse methods adopted by exercise physiologists for data processing can only . decrease . the validity of the measurement of VO2max or VO2 peak, and all variables subsequently dependent on this measure. This fact reflects poorly on exercise physiologists, sport and exercise scientists and general scientists who perform exercise testing to research their physiological specialty. In addition, the majority of scientists continue to use between 0.5 and 1 minute time Sports Med 2010; 40 (2)

. VO2 Data Processing

averages as their method of data processing. This procedure, in turn, lends itself to adoption of the . [17] criterion to verify a VO plateau Taylor et al. 2 . and VO2max. .While we do not .wish to focus on defining the VO2 plateau and VO2max, it is also clear that the type of data processing used has an overwhelming influence on subsequent definitions and data interpretation. It is for this reason that we have focused on data processing in this commentary. 3.2 Validity and Opposing Theories

If exercise physiologists do not use appropriate methods of data processing, then .data for . VO2max and the presence/absence of a VO2 plateau may be invalid. The result of any invalid method or accepted interpretation in physiology, or any field, is the reinforcement of incorrect data and interpretations. Such errors are further increased in magnitude and scope when the data are used to derive . additional theories. Thus, for the measure of VO2max, improper measurement and data processing can lead to errors in modelling such functions as central or peripheral cardiovascular function, blood oxygen delivery, peripheral oxygen diffusion, the influence of hypoxia on central and peripheral vascular function as well as muscle metabolism, the oxygen cost of ventilation, estimates of anaerobic capacity, exercise prescription, training adaptations and relative expressions of metabolic thresholds. Invalid methods and data processing also open the door to alternate explanations of muscle and whole body metabolism during exercise. For example, Noakes[8-10] . has vehemently challenged the concept of a VO2max, along with classic interpretations of the measure that date back to the writings of Hill.[2,3,6,7] Would the arguments of Noakes have been given the same latitude in the peer review and publication process if valid methods. and data processing . to quantify and interpret VO2max and the VO2 plateau had been resolved 50 years ago? Alternatively, if these issues were resolved 50 years ago, would the theories of Noakes be more accepted and appreciated than they are today? These questions reveal that it is impossible to validly argue for or ª 2010 Adis Data Information BV. All rights reserved.

99

against a measurement if there is evidence that the methods used in the measurement are invalid. The net result of this is a stifled advancement of research and knowledge of the measure(s) at question, and the applied topics within human physiology influenced by the measure(s).

3.3 Why New Approaches Are Needed

In the last 50 years we have progressed scientifically from a research and data processing environment based on manual calculations and chemical-based analyses of gas compositions, producing 10–20 datapoints per test, to the rapidly evolving age of electronics and microprocessors. There was an initial progression to semi- and then fully automated metabolic systems, to systems developed for retail sale, and then to a dependence on such commercial systems due to softwarerelated ease-of-use benefits.[20,21,26-29] However, with the more widespread acceptance of breathby-breath methods of indirect calorimetry during the 1980s, making the process of indirect calorimetry more expensive and complex, the dependence on commercial systems increased because of demand for data acquisition software and data processing methods (delay factors, aligning breaths to expired gas fractions, integrating inspired and expired gas fractions in real time, etc.). Breath-by-breath methods of indirect calorimetry can . produce >500 datapoints/test for a typical VO2max protocol. However, today many researchers are now sufficiently trained in electronics and computer programming to write their own software based on a variety of software platforms, develop their own systems, and use a myriad of software applications to easily process large data files generated from indirect calorimetry. Today, more than ever before, scientists need direction for how to process data from indirect calorimetry. Unfortunately, research and explanation of how to process this increasing volume of data has not kept up with the technological progress experienced in the field of indirect calorimetry and the training of many of today’s physiological scientists. Sports Med 2010; 40 (2)

Robergs et al.

100

. 4. Breath-by-Breath VO2 Data: Deciphering Signal from Noise

3.5 3.0 2.5 2.0 1.5 1.0

Heart rate (beats/min)

125 110 95 80

O2 pulse (mL/beat • 1000)

70 55 40 25 10 5

35 30 25 20 15 10

0.0655 FECO2

4 3 2 1

0.0605 0.0555 0.0505

0

0.17

35

0.16

25

FEO2

Breathing frequency (breaths/min)

Tidal volume (L)

VE STPD (L/min)

VO2 (L/min)

To acquire data from every breath involves the collection of approximately 12 datapoints/min at rest, to more than 60 datapoints/min during exercise at volitional fatigue. Any user of a breathby-breath system of indirect calorimetry will attest to the high degree of variability in the data, and recognize the obvious dilemma that follows: what is the most valid way to decrease this variability, and to what extent should this variability be decreased? An important first step prior to answering these questions is to present and explain the contributing factors causing variability in the typical breath-by-breath response to incremental exercise testing. A clear understanding of the

origin and magnitude of this variability is essential for recognizing the need to reduce the variability via one of several processing options. Figure 1 presents steady-state breath-by. breath data for VO2, ventilation, tidal volume, breathing frequency, heart rate (HR), oxygen pulse and expired gas fractions for oxygen (FEO2) and carbon dioxide (FECO2). Data from figure 1 reveal that there is considerable variability in all signals. Close visual inspection also reveals a similar pattern of variability between . variables. With reVO2 and ventilation-derived . spect to the VO2 data, the constant metabolic demand of this exercise condition resulted in . breath-by-breath VO2 data that averaged (mean – SD) 2.17 – 0.3 L/min, with a range of 1.4–3.3 L/min. To assume that breath-by-breath . VO2 data reflect body metabolism resolved to

15

0.15 0.14

5 3

5

7 9 11 Time (min)

13

3

5

7 9 Time (min)

11

13

Fig. 1. Stack plots of pertinent variables measured . and computed from expired gas exchange indirect calorimetry during steadystate exercise (calculated from oxygen consumption [VO2] and heart rate). FEO2 = oxygen pulse, and expired gas fractions for oxygen; FECO2 = oxygen pulse, and expired gas fractions for carbon dioxide; STPD = Standard Temperature and Pressure, Dry; VE = ventilation.

ª 2010 Adis Data Information BV. All rights reserved.

Sports Med 2010; 40 (2)

. VO2 Data Processing

101

b

Intercept VE FEO2

B

t (188)

p-Value

4.2202

26.25

0.000

0.908

0.0495

95.52

0.000

-0.249

-26.2398

-26.15

0.000

FEO2 = expired gas fractions for oxygen; VE = ventilation.

each breath is to assume that body metabolism deviates by as much as 86% during steady-state exercise. Although data at this frequency of col. lection for muscle VO2 has never been collected due to methodological constraints, we feel that it is logical to conclude that such a magnitude of muscle metabolic variability does not occur. For example, application of the Fick equation reveals that if such variability was real, then either muscle blood flow or oxygen extraction, or a combination of the two, would need to collectively change by this magnitude and cause similar oscillatory perturbations in alveolar ventilation and external respiration. We know of no evidence to support this interpretation. This assumed non-metabolic origin of the varia. bility in breath-by-breath VO2 means that a large . portion of the VO2 variability is caused by other variables/conditions. What variables account for this variability, and how much of the variability can be accounted for, are questions that have been addressed and adequately answered by past research. [29] measured breath-by-breath . Potter et al. VO2 in children and showed that tidal volume accounted . for a large proportion of variability for each of VO2 (43%) and VCO2 (49%), with additional variance explanation provided by breath. ing frequencies (22%) for both VO2. and VCO2. Myers et al.[30,31] reported 51% of VO2 variance due to tidal volume, with smaller contributions from breathing frequencies (25%). Data from LaMarra et al.[22] corroborated these findings,. as these scientists stated that breath-by-breath VO2 variability was due to natural breathing imperfections (irregularities). Clearly, past research shows that the variability in ventilation on a breathby-breath basis causes the majority of variability, close to 70%, in any gas volume computation in indirect calorimetry. As this past research has by no means been thorough, or included all possible ª 2010 Adis Data Information BV. All rights reserved.

a 4.0

Pred VO2 = (0.0495(VE)) − (26.2398(FEO2)) + 4.2202 R2 = 0.984 SEE = 0.07 L/min

3.5 Predicted VO2 (L/min)

Variables

variables that may alter metabolic computations (such as expired gas fractions), it is also likely that ventilatory parameters may account for an even larger proportion of explained variance. We completed a multiple regression analysis of . the data from figure 1, where VO2 was the dependent variable, and ventilation, breathing frequency, FECO2 and FEO2 were independent variables. We assumed that steady-state data reflect random signal generation, where each datapoint has independence from one .another in reflection of the actual whole body VO2. Based on a step-wise data entry model, multiple regression analysis produced the results presented in table I. Figure 2a presents the raw and predicted . VO2 data from figure 1, and. the resulting line of best fit. Breath-by-breath VO2 variability was

3.0 2.5 2.0 1.5 1.0 0.5 0 0

0.5

1.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

b 0.3 0.2 Residuals (L/min)

Table I. Results from the multiple regression analysis of breath-bybreath oxygen consumption variability

0.1 0 −0.1 −0.2 −0.3 0

1.5 2.0 2.5 3.0 Measured VO2 (L/min)

3.5

4.0

. Fig. 2. (a) Prediction of oxygen consumption (VO2) based on the independent variables ventilation (VE) and expired gas fractions for oxygen (FEO 2). The two independent variables accounted for 98% of the . raw VO2 variability. (b) Residuals analysis shows the near-even error distribution across the range of measurement, with a mean – SD residuals error = 0.000 – 0.069 L/min. SEE = standard error of estimate.

Sports Med 2010; 40 (2)

Robergs et al.

102

98% explained by a two-factor model comprising ventilation and FEO2. Figure 2b presents the residual distribution from the prediction. Such variability is magnified in the . breath-by-breath computation of ventilation, VO2 and VCO2 due to the convention of expressing this data relative to minute time intervals, thereby multiplying the breath-by-breath volume error by a factor (60/breath interval time [s]) that can range from 12 (rest) to 75 (peak exercise). . We applied the same statistical logic to a VO2 dataset from the linear segment obtained during a ramp incremental exercise protocol (figure 3a). Linear regression was applied to this data segment to derive the line of best fit and the computation of residuals (figure 3b). Multiple regression gave similar results to the steady-state

4.1 Options for Attenuating Variability . in Breath-By-Breath VO2 Data

a 4.5

There are three general approaches that one can use to . attenuate the variability in breathby-breath VO2: time averages, breath averages and digital filtering.

4.0

VO2 (L/min)

3.5 3.0 2.5 2.0

4.1.1 Time and Breath Averages

1.5 1.0 0.5 0 4

6

8

10

12

14

16

18

20

6

8

10 12 14 Time (min)

16

18

20

b 1.0 Residuals (L/min)

data, where a three-variable model of ventilation, accounted for FECO2 and FEO2 (p < 0.0001) . 99.46% of the variability in VO2, with a standard error of estimate (SEE) of 0.09 L/min. Given the clear evidence from our data, as well all of as from other researchers,[22,29-31] almost . the breath-by-breath variability for VO2 during both steady state and incremental exercise is caused by irregularities in the rate and depth of ventilation, which by definition and physiology also influences HR, FEO2 and FECO2, especially during low to moderate exercise intensities. The obvious question is how to remove this variability to better reflect what we . assume to be the smoother muscle VO2 response to metabolic demand?

0.5 0 −0.5 −1.0 4

. Fig. 3. (a) Data for oxygen consumption (VO2) from the linear segment of an incremental exercise test, with the line of best fit based on multiple regression prediction from ventilation and expired gas fractions for carbon dioxide and oxygen. (b) The remaining residuals from the multiple regression prediction (mean – SD residuals = -0.0005 – 0.1991).

ª 2010 Adis Data Information BV. All rights reserved.

Based on our preliminary survey, time averages remain . the most common approach to processing VO2 data. Such averaging is typically a fixed time period such as 15, 30 or 60 seconds. All breaths within the time window are used for the calculation. Breath averages are computed for a fixed number of breaths and these data are plotted versus time. Figures 4 and 5 present data for one individual during cycle ergometry exercise to volitional fatigue. The data were processed by time averages (figure 4a–e) and breath averages (figure 5a–e), with time-averaged data aligned to the end of the time interval based on what we view to be the current conventional, not necessarily correct (alignment should occur to the centre of the time interval), presentation. As revealed by the data in figures 4 and 5, time and breath averaging approaches are crude smoothing tools and have weaknesses in how they represent the measures of interest. For example, because the time interval between Sports Med 2010; 40 (2)

. VO2 Data Processing

103

a = bb

the protocol increased from 2 to 10 and 12 to 38, respectively. Conversely, when using breath averages, the time period in an averaging window shrinks during transitions from rest to maximal exercise. The data in figure 5 reveal that for the 5- and 21-breath averages, the range of time included in each average decreased from 0.51 to 0.1

b = 15 sec 4

4.5 4.0

3

3.0

VO2 (L/min)

2.5 2.0 1.5

2

1

1.0

a = bb

0.5 0 5

10

15

20

0

c = 30 sec

5

10

15

d = 60 sec

4

4

3 VO2 (L/min)

3

2

1

4.0

20

3.0 2.5 2.0 1.5 1.0

2 1

2

0

0 0

5

10

15

0

20

c = 11 breaths

5

10

15

20

5 10 15 Time (min)

20

d = 21 breaths

4

4

3

3

0 5 10 15 Time (min)

20

0

5 10 15 Time (min)

20 VO2 (L/min)

0

e 40

2

1

2

1

60 sec

30 Breaths

3

0.5

1

0

4

3.5 VO2 (L/min)

0

5

VO2 (L/min)

0

VO2 (L/min)

b = 5 breaths

4.5

VO2 (L/min)

VO2 (L/min)

3.5

0

0 0

20

5 10 15 Time (min)

30 sec

20

0

e 1.75

10 15 sec

0

5

10 Time (min)

15

20

. Fig. 4. Oxygen consumption (VO2) data processed as (a) breathby-breath (bb), and time averages of (b) 15 seconds, (c) 30 seconds and (d) 60 seconds. (e) The different breath numbers in the data time averages.

. adjacent breaths decreases as VO2 increases, the number of breaths/unit time increases with increases in exercise intensity. For the 15-second to 1-minute average comparison, the number of breaths/time interval from the start to the end of ª 2010 Adis Data Information BV. All rights reserved.

1.50 Time interval (min)

0

1.25

21 breaths

1.00 0.75

11 breaths

0.50

5 breaths

0.25

bb

0 0

5

10 Time (min)

15

20

. Fig. 5. Oxygen consumption (VO2) data processed as (a) breathby-breath (bb), and averages of (b) five breaths, (c) 11 breaths and (d) 21 breaths. (e) The different time durations for each breathaveraged data interval.

Sports Med 2010; 40 (2)

Robergs et al.

104

and from 1.62 to 0.44 minutes, respectively. In addition, both approaches involve no attempt to determine what specific component of variability is attenuated and what information should remain. The net result from time and breath averaging is an uneven degree of processing across the data set. In addition, how well do time and breath averages remove the ventilationinduced variability, and leave what may actually . reflect muscle- or cardiopulmonary-derived VO2 characteristics? When does breath and time averaging become too crude and actually remove important data and trends that have physiological meaning to oxygen supply and demand, .and perhaps even.to the ability to validly detect a VO2 plateau and VO2max? 4.1.2 Digital Filtering

Use of techniques such as digital filtering are common in engineering and biomechanics, but are not . typically used in processing breath-bybreath VO2 data. This is unfortunate, as the data in figures 1–3 revealed that there. are multiple explanations for the variability of VO2 data when acquired breath-by-breath, with the majority of this variability accounted for by ventilation and related parameters. These multiple contributors . to the final VO2 data signals have the potential to be independently removed if they occur at different rates, or frequencies, . to the variability introduced to whole body VO2 by muscle metabolism or cardiopulmonary physiology. To illustrate the multiple components that re. side within breath-by-breath VO2 at an absolute steady-state exercise intensity, we present data in figure 6a–e based on the assumptions and simulation of a relatively rapid low magnitude oscil. lation in muscle VO2, a slow medium magnitude cardiovascular oscillation that will influence gas delivery and removal from muscle, and a very slow high magnitude ventilatory oscillation profile. Each component (muscle, cardiovascular and ventilatory) has a unique frequency with a different time offset, which when combined and sampled randomly reveal the overall variability . that we might see when quantifying in VO 2 . VO2 from expired gas analysis indirect calorimetry. Figure 6d is not that dissimilar from the ª 2010 Adis Data Information BV. All rights reserved.

real data presented in figure 1a. Furthermore, and as expected based on the simulated data, the digital filtering removed all of the high magnitude ventilatory variability, most of the cardiovascular variability, and left a dataset pre. dominantly influenced in theory by muscle VO2 kinetics (figure 6e). The most . valid method to process breathby-breath VO2 data would be to specifically remove as much extraneous variability as possible, while minimally altering the variability inherent within the remaining data signal. The problem here though is that in reality we do not know . what the muscle VO2 kinetics are across the signal sampling range of 0.2–1.5 Hz (typical range for breathing frequency across an incremental protocol to volitional fatigue). Furthermore, this muscle response may be irrelevant given the mixing of metabolic gases in the central mixed venous blood volume, which would provide a physiological smoothing function driven by ventilation. As such, when using expired gas analysis indirect calorimetry, the cause of the remaining variability is likely to be cardiopulmonary rather than muscular. 4.1.3 Data Distortion

Although it would be convenient to assume . that a digital filter is the perfect answer to VO2 data processing, the fact is that no method of data processing is perfect. For the digital filter, the lower cut-off frequency selected for the processing can have a profound influence on the resulting data trend. A cut-off filter too high does not remove enough variability. A cut-off filter too low can alter, or distort, the data trend. Data distortion also results from time averages, as shown in figure 7a, with no distortion for breath averages if this is done incrementally and not as a block average (figure 7b). To show the results . of data distortion from digital filtering of VO2 data from incremental exercise, figure 7c presents the raw and resulting digitally filtered data for different low cut-off frequencies. Such low cut-off frequencies can be interpreted as allowing only variability to remain in the data at a frequency lower than the cutoff value. As such, all variability resonating at Sports Med 2010; 40 (2)

. VO2 Data Processing

105

a

VO2 (L/min)

4 3 2 d 1

4

0 3 VO2 (L/min)

b

VO2 (L/min)

4 3

2

1 2 1

0

0

4

e

c 3 VO2 (L/min)

VO2 (L/min)

4 3 2

2

1 1 0

0 0

50

100 150 Datapoints

200

250

0

50

100 150 Datapoints

200

250

Fig. 6. Simulated data to reveal multiple components at different frequencies that combine at random to reveal a breath-by-breath oxygen . . consumption (VO2) profile. The simulated data are for (a) muscle VO2, (b) cardiovascular function and (c) ventilation. (d) Represents the random presentation of data from the three signals. (e) Represents the digitally filtered dataset using a low pass (0.1 Hz) third order Butterworth filter (see Appendix).

frequencies higher than the cut-off are excluded and the datapoints of these higher frequency variability responses are changed to fit a lower frequency smaller magnitude profile tolerated by the low frequency cut-off value. The data in figure 7c reveal that low cut-off frequencies above 0.08 . Hz do not provide adequate reduction in VO2 data variability, but clearly fit the raw data well. On the other extreme, cut-off frequencies lower than 0.03 Hz provide dramatic reductions in data variability, yet induce a clear distortion of ª 2010 Adis Data Information BV. All rights reserved.

the data seen as a downward-right adjustment of the data and trend. The obvious question is when does the digital filtering begin to alter the true trend of the raw data? We decided to base our procedures on the methods used within the field of biomechanics. For example, Winter[32] has recommended that the optimal cut-off frequency can be determined from the line of best fit resulting from a linear regression applied to a subset of the range of residuals versus a range of low cut-off frequencies. Sports Med 2010; 40 (2)

Robergs et al.

3.00

3.25

2.75

3.00 2.75

4.0

4.0

2.50 14.0 14.5 15.0 15.5 16.0 16.5 Time (min)

3.5

b

2.50 2.25 2.00 11.0 11.5 12.0 12.5 13.0 13.5 14.0

3.5

17.0

Time (min)

3.0 VO2 (L/min)

3.0 VO2 (L/min)

3.50 VO2 (L/min)

a

VO2 (L/min)

106

2.5 2.0 1.5

2.5 2.0 1.5 1.0

1.0 VO2 60-sec average

0.5

VO2 linear regression 21-breath average

0.5

0

0 5

6

7

8

9

10 11 12 13 14 15 16 17

5

6

7

8

Time (min)

9

10 11 12 13 14 15 16 17 Time (min)

3.0 2.5 2.0 1.5 1.0

c 0.5 4.0

0 9.0

3.5

9.5

VO2 (L/min)

10.0 10.5 11.0 11.5 12.0 Time (min)

3.0 2.5 2.0

0.005 Hz 0.01 Hz 0.03 Hz 0.08 Hz 0.15 Hz Linear regression

1.5 1.0 0.5 0 5

6

7

8

9 10 11 12 13 14 15 16 17 Time (min)

Fig. 7. Data distortion resulting from processing with (a) a 60-second time average, (b) a 21-breath running average and (c) multiple examples of a .low frequency cut-off digital filter (Butterworth third order). The data used .in these figures are from a linear portion of oxygen consumption (VO2) data from a cycle ergometry ramp incremental protocol to maximum VO2.

The residuals are computed from the expected data value (from line of best fit) and that obtained after applying the low frequency cut-off digital filter. Presumably, the residuals error from digital filtering at higher cut-off frequencies induces a small negative change in the mean residual, with this error increasing as the lower cut-off frequency is decreased. Winter[32] has explained that this increasing error fits a linear function until ª 2010 Adis Data Information BV. All rights reserved.

the lower cut-off frequency is decreased to a certain frequency, after which dramatic increases in error result. It was proposed that the y-intercept of this linear regression can be used to extend out horizontally, and where it intersects the residuals plot reveals the recommended lower cutoff frequency. . We present this graphic for the . linear VO2 segment of data from a test to VO2max for a Sports Med 2010; 40 (2)

. VO2 Data Processing

107

a 3.5 3.0 VO2 (L/min)

2.5 2.0 1.5 1.0 0.5 0 0

2

4

6

8

10

12

14

b 2.5 20% error

2.0 VO2 (L/min)

representative subject during cycle ergometry in figure 8. The data in figure 8 clearly show that the residuals error response over a range of cut-off frequencies is nonlinear, and is accurately modelled by a two-segment exponential decay function. As no initial linear relationship exists, the recommendations of Winter[32] for objectively detecting the lower . cut-off frequency does not have validity for VO2 data. An alternate approach is to predetermine what is an acceptable level of signal variability reduction based on residuals analysis from a line of best fit. Previously, we demonstrated that measures of ventilation, FECO2 and FEO2 account for approximately 96–98% of variability in breath. by-breath VO2. Conversely, prior research has shown variance explanation closer to 90%.[22,29-31] Given the lack of research on this topic, and especially research using multiple prediction variables

1.5 1.0 0.5 0 0

1.8

2

4

6

8

10

12

14

c 1.6

2.5 0.05

5% error 2.0

1.4 VO2 (L/min)

0.04

Mean residual

1.2 0.03 1.0

1.5 1.0 0.5

0.02 0.8

0 0.01

0

0.6 0

0.4

0.10

0.13

0.16

0.19

0.22

0.25

0.2

2

4

6 8 Time (min)

10

12

14

Fig. . 9. Data for (a) measured and predicted oxygen consumption (VO2) based on the independent variables ventilation and oxygen pulse, and expired gas fractions for oxygen. The residuals error was then decreased by (b) 80% and (c) 95% to reveal smoothed graphical presentations of the data set.

0 0

0.05

0.10 0.15 0.20 Low cut-off frequency (Hz)

0.25

Fig. 8. The increase in mean residuals resulting from a comparison of digital filtering at cut-off frequencies ranging from 0.001 to 0.25 Hz. The data for this presentation are . from a subject with a linear increase in oxygen consumption (VO2) during incremental exercise to . maximum VO2. The residuals represent the difference between the digital . processed datapoints and the linear regression prediction of the VO2. The inner plot shows the curvilinear nature of the relationship throughout even for the higher values of the cut-off frequency range.

ª 2010 Adis Data Information BV. All rights reserved.

that also adjust for expired gas fractions (as we do in this report), at this time it seems prudent to develop a processing strategy that removes approximately 90% . of the original breath-by-breath variability in VO2. To illustrate this recommendation, figure 9a presents the raw . (dataset used for figure 1a) and predicted VO2 data from the regression prediction with the independent Sports Med 2010; 40 (2)

Robergs et al.

108

Raw data 10% residuals error c 4.5

4.0

4.0

3.5

3.5

3.0

3.0

VO2 (L/min)

VO2 (L/min)

a 4.5

2.5 2.0 1.5

2.5 2.0 1.5

1.0

1.0

0.5

0.5

0

0 4

6

8

10

12

14

16

18

20

6

8

10

12

14

16

18

20

4

6

8

10 12 14 Time (min)

16

18

20

d

4.5

4.5

4.0

4.0

3.5

3.5

3.0

3.0

VO2 (L/min)

VO2 (L/min)

b

4

2.5 2.0 1.5

2.5 2.0 1.5

1.0

1.0

0.5

0.5 0

0 4

6

8

10 12 14 Time (min)

16

18

20

. Fig. 10. Data for (a) measured and predicted oxygen consumption (VO2) based on 90% residuals error as described in the text. The same raw . VO2 data are presented as (b) 30-second averages, (c) 15-breath averages, and (d) after a 0.04 Hz digital filter.

variables of ventilation and FEO2. We then adjusted for 80% and 95% of the residual error (20% and . 5% of each residual, respectively) for each VO2 datapoint and plotted each result in figure 9b and c. To assess how such residuals error reduction . alters VO2 data from incremental exercise, we then applied a 90%.residual error reduction to the linear . segment of VO2 data acquired from a test to VO2max (figure 10 for data from figure 3). The data processing required to approximate 90% error reduction (SScorr = SS raw · 0.1, where SScorr = sum of squares corrected, and SS = sum of squares) [figure 10a] results from a time average of 30 seconds (figure 10b), a 15-breath running average (figure 10c), and a low cut-off frequency digital filter of 0.04 Hz (figure 10d). ª 2010 Adis Data Information BV. All rights reserved.

5. Recommendations for Processing Data from Indirect Calorimetry Based on the prior content, we are now ready to recommend strategies for processing data acquired from indirect calorimetry. If a time-averaged system has to be used, we recommend no longer than a 30-second time average (figure 4b) where the data are aligned to the central time of the interval period, which is 15 seconds, and thereby require time representation of 0, 0.25, 0.75, 1.25, 1.75, 2.25 minutes, etc. We also recommend that exercise physiologists who currently use expired mixing chamber systems, with no choice of other sampling and processing options, strive to equip themselves with software that will support acquisition and data processing Sports Med 2010; 40 (2)

. VO2 Data Processing

as breath averages. While a 30-second average provides reasonable reductions in data variability, it provides unreasonable decreases in data frequency, which will detract from how the data can be used to assess important physiology measurements and trends. For breath-by-breath systems and averaging systems suited to breath averages, we recommend a 15-breath running average, aligned to the time of the central breath, which is the eighth breath (figure 10c). Although we identified several theoretical problems with a breath average, the alternative of a digital filter requires a degree of mathematical computation and software dependence that simply does not exist in software of all commercial systems of indirect calorimetry. Furthermore, given that the 15-breath average induces minimal data loss (lose initial seven and last seven datapoints), has no data and trend distortion, can be accomplished with the software of many commercial indirect calorimetry systems, and, if not, can be easily applied to datasets with post-acquisition spreadsheet computation, it is a reasonable expectation that all scientists and practitioners can do this data processing. For scientists able to implement digital filters in their data processing, we recommend a low cut-off frequency digital filter of 0.04 Hz. 5.1 Recommendations for Detecting the Highest Value Datapoint

Once the recommended data processing strategies are used, then the task of detecting the peak or maximal value of any variable is simple. The highest, peak or maximal value becomes the highest processed . Thus, for time. datapoint. or VO averaged systems, VO2max 2 peak would be . for. the test. the highest 30-second VO2 average . or VO2 peak For breath-by-breath data, VO2max . would be the highest 15-breath VO2 average for the test. For exercise physiologists who can apply a 0.04 Hz low frequency cut-off digital. filter, the . largest single datapoint is VO2max or VO2 peak. Exercise physiologists need to agree on a consistent process for analysing data from indirect calorimetry. We feel that our recommendations are valid, based on empiricism, and if adopted ª 2010 Adis Data Information BV. All rights reserved.

109

can improve the quality of research and data interpretation in exercise physiology, sport and exercise science, and both basic and applied physiology. We encourage researchers to further assess our recommendations and rationale, to scientifically test our methods and recommendations, and thereby contribute to the refinement of these recommendations. As with any science, we acknowledge the need to continually improve on methods of data acquisition and interpretation, and we see this commentary as a major step forward in this . continually evolving process. Once VO2 data is more consistently processed and presented in peer-review research, perhaps we will then be in a better position to more validly comment . on such topics. as the validity of or need for the VO2 plateau at VO2max.[33,34,36,37] Acknowledgements No sources of funding were used to assist in the preparation of this review. The authors have no conflicts of interest that are directly relevant to the content of this review.

Appendix A Original Data Acquisition from Indirect Calorimetry

All original data presented in this manuscript were collected using a custom-developed indirect calorimetry system. Subjects ventilated air through a standard one-way mouthpiece configuration (Hans Rudolph, Shawnee, KS, USA). Expired air passed through a unidirectional flow turbine (KL Engineering, Van Nuys, CA, USA) connected to the expired port of the mouthpiece, and then inseries to a compliant 3 L mixing bag (Erich Jaeger, Friedberg, Germany). Mixed expired air was sampled from the mixing bag and analysed for oxygen and carbon dioxide content using electronic analysers (AEI Technologies, Pittsburg, PA, USA) with a time delay of 2.5 seconds. Analogue signals from the turbine and analysers were acquired in real time using commercial hardware (National Instruments, Austin, TX, USA), and processed in real time using custom-developed software (LabVIEW, National Instruments, Austin, TX, USA). Sports Med 2010; 40 (2)

Robergs et al.

110

Steady-State Simulated Data

. It was assumed that breath-by-breath VO2 data is influenced by three inherent sine-wave rhythms: muscle metabolism, cardiovascular function and ventilation. The respective signal features of these rhythms were as follows. Muscle: sine-wave frequency = 0.5 Hz, phase = 0, amplitude = 0.1 L/min, mean = 2.17 L/min, sampling frequency = 5 Hz, total samples = 250. Cardiovascular: sine-wave frequency = 0.225 Hz, phase = 45, amplitude = 0.25 L/min, mean = 2.17 L/min, sampling frequency = 5 Hz, total samples = 250. Ventilation: sine-wave frequency = 0.1 Hz, phase = 90, amplitude = 0.8 L/min, mean = 2.17 L/min, sampling frequency .= 5 Hz, total samples = 250. The final simulated VO2 response was derived by interleaving the three data arrays (muscle, cardiovascular, ventilation) and then completing 250 random samples from this larger array by computer-generated random sampling (LabVIEW, National Instruments, Austin, TX, USA).

References 1. Lusk G. The elements of the science of nutrition. Philadelphia (PA): WB Saunders, 1928 2. Hill AV. Muscular activity. London: Tindall and Cox, 1925 3. Hill AV, Lupton H. Muscular exercise, lactic acid, and the supply and utilization of oxygen. Q Q J Med 1923; 16: 135-71 4. Bassett DR, Howley ET. Limiting factors for maximum oxygen uptake and determinants of endurance performance. Med Sci Sports Exerc 2000; 32: 70-84 5. Bassett DR, Howley ET. Maximal oxygen uptake: ‘‘classical’’ versus ‘‘contemporary’’ viewpoints. Med Sci Sports Exerc 1997; 2: 591-603 6. Bergh U, Ekblom B, Astrand P. Maximal oxygen uptake ‘‘classical’’ versus ‘‘contemporary’’ viewpoints. Med Sci Sports Exerc 2000; 32: 85-8 7. Howley ET, Bassett DR, Welch HG. Criteria for maximal oxygen uptake: review and commentary. Med Sci Sports Exerc 1995; 27: 1292-301 8. Noakes TD. Implications of exercise testing for prediction of athletic performance: a contemporary perspective. Med Sci Sports Exerc 1988; 20: 319-30 9. Noakes TD. Challenging beliefs: ex Africa semper aliquid novi. Med Sci Sports Exerc 1997; 9: 571-90 10. Noakes TD. Maximal oxygen uptake: ‘‘classical’’ versus ‘‘contemporary’’ viewpoints: a rebuttal. Med Sci Sports Exerc 1998; 30: 1381-98 11. Robergs RA. An exercise physiologist’s ‘‘contemporary’’ interpretations of the ‘‘ugly and creaking edifices’’ of the . VO2max concept. JEPonline 2001; 4 (1): 1-44

ª 2010 Adis Data Information BV. All rights reserved.

12. Cumming GR, Borysyk LM. Criteria for maximum oxygen uptake in men over 40 in a population survey. Med Sci Sports Exerc 1972; 4: 18-20 13. Froelicher Jr VF, Brammell H, Davis G, et al. A comparison of three maximal treadmill exercise protocols. J Appl Physiol 1974; 36: 720-5 14. McArdle WD, Katch FI, Pechar GS. Comparison of continuous . and discontinuous treadmill and bicycle tests for maxVO2. Med Sci Sports Exerc 1973; 5: 156-60 15. Mitchell JH, Blomqvist G. Maximal oxygen uptake. New Engl J Med 1971; 284: 1018-22 16. Mitchell JH, Sproule BJ, Chapman CB. The physiological meaning of the maximal oxygen intake test. J Clin Invest 1958; 37: 538-46 17. Taylor HL, Buskirk E, Henschel A. Maximal oxygen intake as an objective measure of cardio-respiratory performance. J Appl Physiol 1955; 8: 73-80 18. American College of Sports Medicine. ACSM’s guidelines for exercise testing and prescription. Philadelphia (PA): Lippincott Williams and Wilkins, 2005 19. Wyndham CH, Strydom NB, Maritz JS, et al. Maximal oxygen intake and maximum HR during strenuous work. J Appl Physiol 1959; 14: 927-36 20. Beaver WL, Wasserman K, Whipp BJ. On-line computer analysis and breath-by-breath graphical display of exercise function tests. J Appl Physiol 1973; 34: 128-32 21. Beaver WL, LaMarra N, Wasserman K. Breath-by-breath measurement of true alveolar gas exchange. J Appl Physiol 1981; 51 (6): 1662-75 22. Lamarra N, Whipp BJ, Ward SA, et al. Effect of inter-breath fluctuations on characterizing exercise gas exchange kinetics. J Appl Physiol 1987; 62 (5): 2003-12 23. Roecker K, Prettin S, Sorichter S. Gas exchange measurements with high temporal resolution: the breath-by-breath approach. Int J Sports Med 2005; 26: S11-8 24. Whyte GP, Sharma S, George K, et al. Exercise gas exchange responses in the differentiation of pathologic and physiologic left ventricular hypertrophy. Med Sci Sports Exerc 1999; 31 (9): 1237-41 25. Day JR, Rossiter HB, Coats EM, et al. The maximally . attainable VO2 during exercise in humans: the peak vs. maximum issue. J Appl Physiol 2003; 95 (5): 1901-7 26. Duncan GE, Howley ET, Johnson BN. Applicability of . VO2max criteria: discontinuous versus continuous protocols. Med Sci Sports Exerc 1997; 29: 273-8 27. Dwyer DB. A standard method for determination of maximal . aerobic power from breath-by-breah VO2 data obtained during a continuous ramp test on a bicycle ergometer. JEPonline 2004; 7 (5): 1-9 28. Pearce DH, Milhorn HT, Holoman GH, et al. Computerbased system for analysis of respiratory responses to exercise. J Appl Physiol 1977; 42: 968-75 29. Potter CR, Childs DJ, Houghton W, et al. Breath-to-breath ‘‘noise’’ in the ventilatory and gas exchange responses of children to exercise. Eur J Appl Physiol Occup Physiol 1999; 80 (2): 118-24 30. Myers J, Walsh D, Buchanan N, et al. Can maximal cardiopulmonary capacity be recognized by a plateau in oxygen uptake? Med Sci Sports Exerc 1989; 96: 1312-6

Sports Med 2010; 40 (2)

. VO2 Data Processing

31. Myers J, Walsh D, Sullivan M, et al. Effect of sampling on variability and plateau in oxygen uptake. J Appl Physiol 1990; 68: 404-10 32. Winter DA. Biomechanics and motor control of human movement. 3rd ed. Hoboken (NJ): Wiley Publishers, 2005 33. Astorino TA, Robergs RA, . Ghiasvand F, et al. Incidence of the oxygen plateau at VO2max during exercise testing to volitional fatigue. JEPonline 2000; 3 (4): 1-12 34. Doherty M, Nobbs L, Noakes TD. Low frequency of the ‘‘plateau phenomenon’’ during maximal exercise in elite British athletes. Eur J Appl Physiol 2003; 89: 619-23 35. Freedson P, Kline G, Porcari J, et al. Criteria for defining . VO2max: a new approach to an old problem [abstract]. Med Sci Sports Exerc 1986; 18: S36

ª 2010 Adis Data Information BV. All rights reserved.

111

. 36. Weir JP, Koerner S, Mack B, et al. VO2 plateau detection in cycle ergometry. JEPonline 2004; 7 (2): 55-62 . 37. Howley ET. VO2max and the plateau: needed or not? Med Sci Sports Exerc 2007; 39 (1): 101-2 38. Macfarlane DJ. Automated metabolic gas analysis systems: a review. Sports Med 2001; 31 (12): 841-61

Correspondence: Robert A. Robergs, Head of Program: Sport and Exercise Science, School of Biomedical and Health Sciences, Building 20, The University of Western Sydney – Campbelltown Campus, Locked Bag 1797, Penrith South DC, NSW 1797, Australia. E-mail: [email protected]

Sports Med 2010; 40 (2)

Sports Med 2010; 40 (2): 113-139 0112-1642/10/0002-0113/$49.95/0

REVIEW ARTICLE

ª 2010 Adis Data Information BV. All rights reserved.

Guidelines for Glycerol Use in Hyperhydration and Rehydration Associated with Exercise Simon Piet van Rosendal,1 Mark Andrew Osborne,2 Robert Gordon Fassett1,3 and Jeff Scott Coombes1 1 School of Human Movement Studies, The University of Queensland, Brisbane, Queensland, Australia 2 Queensland Academy of Sport, Brisbane, Queensland, Australia 3 Royal Brisbane and Women’s Hospital, Brisbane, Queensland, Australia

Contents Abstract. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. Considerations for Glycerol Use . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1 Glycerol Dose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2 Volume of Fluid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.3 Type of Fluid. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.4 Timing of Fluid with Glycerol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.5 Duration of Hyperhydration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.6 Guidelines for Pre-Exercise Glycerol Hyperhydration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. Glycerol Ingestion during Exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1 Glycerol Ingestion during Exercise, after Pre-Exercise Hyperhydration . . . . . . . . . . . . . . . . . . . . . . 2.2 Glycerol Ingestion during Exercise, without Pre-Exercise Hyperhydration. . . . . . . . . . . . . . . . . . . . 2.3 Guidelines for Glycerol Ingestion during Exercise. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. Glycerol as a Rehydrating Agent. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4. Areas for Further Investigation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5. Side Effects from Glycerol Consumption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6. Summary and Recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Abstract

113 115 115 115 127 130 131 131 132 132 133 134 134 135 135 136

Dehydration in athletes alters cardiovascular and thermoregulatory function and may inhibit endurance exercise capacity if fluid loss exceeds 2% of bodyweight (BW). If this level of dehydration cannot be prevented when starting from a state of euhydration, then athletes may create a state of hyperhydration by consuming extra fluid prior to exercise. From this hyperhydrated situation, individuals have a greater capacity to tolerate fluid loss before becoming dehydrated. Furthermore, excess pre-exercise fluid intake enhances thermoregulatory ability, as well as increasing plasma volume to maintain cardiac output. However, hyperhydrating before exercise is difficult, because a large fluid intake is typically accompanied by diuresis. Glycerol-containing beverages create an osmotic gradient in the circulation favouring fluid retention, thereby facilitating hyperhydration and protecting against dehydration. Many studies have shown that increases in body water by 1 L or more are achievable through glycerol hyperhydration. This article analyses the evidence for glycerol use in

van Rosendal et al.

114

facilitating hyperhydration and rehydration, and provides guidelines for athletes wishing to use this compound. An analysis of the studies in this area indicates that endurance athletes intending to hyperhydrate with glycerol should ingest glycerol 1.2 g/kg BW in 26 mL/kg BW of fluid over a period of 60 minutes, 30 minutes prior to exercise. The effects of glycerol on total body water when used during rehydration are less well defined, due to the limited studies conducted. However, ingesting glycerol 0.125 g/kg BW in a volume equal to 5 mL/kg BW during exercise will delay dehydration, while adding glycerol 1.0 g/kg BW to each 1.5 L of fluid consumed following exercise will accelerate the restoration of plasma volume. Side effects from glycerol ingestion are rare, but include nausea, gastrointestinal discomfort and light-headedness. In summary, glycerol ingestion before, during or following exercise is likely to improve the hydration state of the endurance athlete.

Mechanisms that regulate body water are complex. Fluid intake factors such as volume, temperature and composition of ingested fluid, together with gastrointestinal absorption rate, need to be balanced against fluid loss factors such as sweat and renal excretion rates.[1] These can be extensively influenced by environmental conditions, with heat and humidity significantly increasing the rate of fluid loss and altering the distribution of body fluid to aid in heat dissipation.[1] Fluid losses, or shifts between compartments with dehydration, reduce plasma volume and cardiac stroke volume, with concomitant increases in heart rate.[2-4] Peripheral blood supply may be reduced to maintain central blood pressure, leading to a reduction in heat dissipation and increased core temperature.[5,6] These physiological changes may contribute to exertional heatstroke, which may be fatal.[7-9] Furthermore, endurance performance may be impaired when fluid losses exceed approximately 2% bodyweight (BW) during exercise,[10,11] or when subsequent exercise commences when an athlete is still hypohydrated by 2% BW or greater.[12-16] Throughout this review, ‘hyperhydration’ is defined as body water excess beyond normal fluctuations and is characterized by a urinespecific gravity under 1.010.[17] ‘Euhydration’ refers to normal body water and ‘hypohydration’ implies body water deficits beyond normal fluctuations (urine specific gravity over 1.023).[18] Furthermore, ‘dehydration’ refers to losing and ‘rehydration’ to gaining body water. ª 2010 Adis Data Information BV. All rights reserved.

When used predominantly as a preparatory mechanism for subsequent endurance exercise, especially in thermally stressful environments, it is accepted that pre-exercise hyperhydration will delay, prevent or attenuate the effects of dehydration.[19,20] However, ingesting a large bolus of fluid before exercise, even in dehydrated subjects, is typically ineffective at inducing hyperhydration, due to a rapid decrease in antidiuretic hormone (arginine vasopressin), leading to augmented diuresis.[21,22] Therefore, any substance that increases fluid retention before, during or after exercise may have beneficial effects on fluid homeostasis and assist the endurance athlete. Glycerol is a metabolite released during the breakdown of triglycerides, and is distributed in low concentrations throughout all body cells.[23] Its osmotic properties have generated interest in hydration research where the primary focus has been pre-exercise glycerol hyperhydration. Glycerol ingestion with fluid during exercise has also been investigated as a means of attenuating dehydration, while most recently glycerol has been investigated as an agent to assist rehydration. Several previous reviews[23,24] have discussed the pharmacokinetics and mechanisms by which glycerol assists in fluid retention and endurance performance.[25] The primary aim of this review is to provide athletes with specific guidelines for the use of glycerol for pre-exercise hyperhydration, or in beverages consumed during exercise (with and without pre-exercise hyperhydration) and during post-exercise recovery. There are a number of Sports Med 2010; 40 (2)

Guidelines for Glycerol Use

factors that need to be considered when formulating guidelines for athletes wishing to explore the use of glycerol in preparation for exercise. These are discussed with reference to previous studies (table I) that have investigated these issues. To more accurately compare and contrast these studies, they were evaluated using a scale that assessed a number of factors associated with the minimization of bias in areas such as subject selection, performance and data analysis (table II). The results of this process are shown in table III, and more emphasis has been placed on studies with less experimental bias when formulating the guidelines. The scoring system itself was developed using items from a number of extensively evaluated and validated tools[48-51] used to assess the quality of randomized, controlled clinical trials. Items were chosen from those used in the Jadad scoring system,[48] the PEDro scale[49] and the Delphi List,[50] in addition to recommendations contained in the CONSORT statement.[51] Finally, items five and 16 were added because of their perceived importance in studies used to assess exercise performance. It should be noted that the final version of the current scale has yet to be rigorously evaluated and validated on its own merit. From their assessment of construct validity, Jadad et al.[48] indicate that scores of 4/6 and 2/3 on their scales separate the bulk of studies into poor and excellent categories. Based on this, many subsequent studies have used a cut-off of >60% to classify a study as excellent.[52,53] From table III it can be seen that only a few publications contained a high number of important elements. Indeed, only five studies scored >60%, each of which found glycerol hyperhydration to be beneficial. 1. Considerations for Glycerol Use The following sections discuss factors that are fundamental to the use of glycerol in pre-exercise hyperhydration. 1.1 Glycerol Dose

Riedesel et al.[1] investigated the dose-response relationship for glycerol doses of 0.5, 1.0 or ª 2010 Adis Data Information BV. All rights reserved.

115

1.5 g/kg BW. All doses increased water retention; however, the hyperhydration achieved with the 0.5 g/kg glycerol dose was not significantly different from water intake alone. Subsequent analysis supports these findings that glycerol doses higher than 0.5 g/kg BW are required to maximize fluid retention. Robergs and Griffin[23] indicate that a glycerol dose of 1.0–1.5 g/kg BW will be required to elevate plasma glycerol levels to 15 mmol/L. This is the concentration above which blood glycerol levels stabilize, thereby maximizing the osmotic gradient for fluid retention.[23] In the Riedesel et al.[1] study, the level of fluid retention was not further increased following consumption of 1.5 g/kg BW over the 1.0 g/kg BW condition. This disproportion between glycerol dose and fluid retention may have occurred because of the direct relationship between plasma concentration and glycerol catabolism/excretion, which increases glycerol removal from the circulation with higher glycerol doses. Urinary glycerol excretion was doubled and tripled in the 1.5 g/kg condition compared with the 1.0 g/kg condition after 2 and 4 hours, respectively.[1] Consequently, the concentration of glycerol remaining in the body and acting to increase the osmotic gradient to assist in fluid retention was similar between the 1.0 and 1.5 g/kg trials.[1] The top five scoring studies from the quality analysis used glycerol doses of 1.0[35,40] or 1.2 g/kg BW.[34,36,38] All three studies using a dose of 1.2 g/kg BW had fluid retention levels >1 L after 2 hours, compared with 350–500 mL when 1.0 g/kg BW was used. This represents the volume by which total body water is increased above euhydration, which may then be used to convey thermoregulatory and performance benefits. Therefore 1.2 g/kg BW is the recommended dose, which is slightly higher than the average dose of 1.1 g/kg BW given across all pre-exercise hyperhydration studies.[25] 1.2 Volume of Fluid

The total volume of fluid used in glycerol hyperhydration studies ranges from 20 to 29 mL/kg BW. Goulet et al.[25] conducted a meta-analysis on fluid retention associated with glycerol hyperhydration Sports Med 2010; 40 (2)

116

ª 2010 Adis Data Information BV. All rights reserved.

Table I. Summary of studies using glycerol in beverages for hyperhydration, during exercise or rehydration Study, year

Subjects

Treatments

Fluid vol with glycerol (mL/kg BW)

Mean total fluid vol (mL)

Hydration regimen

Fluid with glycerol

Major findings

Exercise protocol

In GIH, ~780 (55%), 850 (60%), 920 mL (65%) of ingested fluid retained after 120 minutes for 0.5, 1.0 and 1.5 g/kg BW, respectively. GIH ﬂ urine vol through › plasma osmolality, no effect on plasma vol, little additional hyperhydration with 1.5 g/kg compared with 1.0 g/kg, especially after 2 h

No exercise component

Fluid retention ~850 mL (53%) after 120 min. Similar (% fluid retained) to results from 0.5 g/kg glycerol with fluid consumed within 40 min. Indicates little additional benefit from consuming fluid over extended period

No exercise component

Glycerol hyperhydration without exercise Riedesel et al.,[1] 1987 Series I

Riedesel et al.,[1] 1987 Series II

4 M + 3 F, healthy adults

Glycerol Placebo

0.5

21.4

1412

Initial glycerol bolus at time 0, then fluid consumed over 40 min; no exer

H2O with 0.1% NaCl

5 M + 3 F, healthy adults

Glycerol Placebo

1.0

21.4

1412

Initial glycerol bolus at time 0, then fluid consumed over 40 min; no exer

H2O with 0.1% NaCl

4 M + 3 F, healthy adults

Glycerol Placebo

1.5

21.4

1412

Initial glycerol bolus at time 0, then fluid consumed over 40 min; no exer

H2O with 0.1% NaCl

5 M + 5 F, healthy adults

Glycerol Placebo

1.0

25.7

1593

Initial glycerol bolus at time 0, with full fluid vol consumed over 210 min; no exer

H2O with 0.1% NaCl

Continued next page

van Rosendal et al.

Sports Med 2010; 40 (2)

Glycerol dose (g/kg BW)

Study, year

Subjects

Treatments

Glycerol dose (g/kg BW)

Fluid vol with glycerol (mL/kg BW)

Mean total fluid vol (mL)

Hydration regimen

Fluid with glycerol

Major findings

Exercise protocol

Fruend et al.,[26] 1995

11 M, healthy adults

Glycerol H2O Control

1.5 g/L TBW (= 0.9)

37.0 mL/kg TBW (= 22.2)

1765

Initial glycerol bolus at time 0, then fluid consumed over 30 min; no exer

H2O (flavoured)

In GIH, ~1250 mL (70%) and ~1200 mL (68%) of ingested fluid retained after 90 and 120 min. GIH ﬂ urine flow rates, ﬂ free H2O clearance rates and › fluid retention (~250 mL at 90 min and ~500 mL at 120 min) vs WIH (p < 0.05). ADH rose (p > 0.05) with glycerol at the same time urine flow and free H2O clearances differed, indicating a possible relationship

No exercise component

Melin et al.,[27] 2002 Koulmann et al.,[28] 2000

8M

Glycerol Control

1.1

21.4

1562

Initial glycerol bolus at time 0, then fluid consumed over 90 min; no exer

Mineral H2O (0.1% NaCl)

In GIH, ~1200 mL (77%) of ingested fluid retained after 2 h. This half-persisted (~560 mL) a further 90 min later. GIH › plasma osmolality. No effect on renin, aldosterone, ADH or ANP

No exercise component

1.0 initial + 0.1 every h after 2 h

Initial bolus 3.3, 24.7 total in first h, total of 28.4 after 4 h

1729 in 1 hb 1988 after 4 hb

Initial glycerol bolus at time 0, then fluid consumed over 60 min. Exer began 90 min after final fluid intake. Additional glycerol and fluid each h after 2 h

OJ + H2O

In GIH, ~1470 mL (85%) and ~1380 mL (80%) of ingested fluid retained after 90 and 150 min vs ~1300 mL (75%) and ~860 mL (50%) in WIH. GIH ﬂ urine vol before exercise, no effect on haemoglobin, haematocrit or serum

Treadmill walking at 60% . VO2max, with 5 min rest every 30 min (lab based)

Guidelines for Glycerol Use

ª 2010 Adis Data Information BV. All rights reserved.

Table I. Contd

Pre-exercise glycerol hyperhydration Lyons et al.,[20] 1990a

4 M + 2 F, healthy adults

Glycerol Placebo 1 (large fluid) Placebo 2 (small fluid)

117

Sports Med 2010; 40 (2)

Continued next page

118

ª 2010 Adis Data Information BV. All rights reserved.

Table I. Contd Study, year

Subjects

Treatments

Glycerol dose (g/kg BW)

Fluid vol with glycerol (mL/kg BW)

Mean total fluid vol (mL)

Hydration regimen

Fluid with glycerol

Major findings

Exercise protocol

electrolyte concentrations. GIH › sweat rates, ﬂ rectal temperature during exer 10 M, nonacc

4% CHOelectrolyte drink with 1% glycerol H2O 5% CHOelectrolyte drink

Unknown (estimated at 0.125 g/h over 60 h)

26

1742 initial then encouraged to drink 1 L/h over 60 h

Initial glycerol solution consumed over 90 min, then additional glycerol solution and fluid consumption variable over 60 h; exercise variable over 60 h

Ad libitum 4% CHOelectrolyte drink with 1% glycerol

H2O alone appeared to provide adequate hydration during submaximal exer in simulated desert conditions. Glycerol showed a nonsignificant trend to › sweat rates

Treadmill walking: 3 · 40 min protocols at 4.8 km/h daily (carrying army pack weighing ~16.5 kg) [lab based]

Montner et al.,[30] 1996 Series I

11, endurance trained

Glycerol H2O

1.2 (1.0 initial + 0.2 at 60 min) as 20% solution

26

1749

Initial 20% (5 mL/kg BW) glycerol solution consumed over 30 min, then additional glycerol dose at 60 min mark; total fluid ingestion over 90 min; exer began 60 min after final fluid intake

H2O (aspartameflavoured)

In GIH, ~800 mL (45%) of ingested fluid retained pre-exer (60 min after final fluid consumption) vs ~70 mL in WIH (p < 0.05). Pre-exer urine vol ﬂ 666 mL with GIH vs WIH (p < 0.05). Glycerol ingestion ﬂ heart rate but no effect on rectal temperature. GIH › endurance time (p < 0.05) in both Series I and Series II

Cycle to exhaustion at 61% Wmax (lab based)

Montner et al.,[30] 1996 Series II

5 M + 2 F, endurance trained

Glycerol H2O

1.2 (1.0 initial + 0.2 at 60 min)

26 + 3 every 20 min during exercise

1749 + 605.6 mL/h during exercise

Same pre-exer hyperhydration regimen as for Series I.

H2O (aspartameflavoured)

The additional fluid consumed during exer prolonged endurance

Cycle to exhaustion at 61% Wmax (lab based)

Continued next page

van Rosendal et al.

Sports Med 2010; 40 (2)

Meyer et al.,[29] 1995

Study, year

Subjects

Treatments

Glycerol dose (g/kg BW)

Fluid vol with glycerol (mL/kg BW)

Mean total fluid vol (mL)

Hydration regimen

Fluid with glycerol

Major findings

Exercise protocol

time in both GIH & WIH trials

Additional fluid (CHOelectrolyte solution) consumed during exer

8 M, endurance trained, heat-acc

Euhydration (control) Glycerol Glycerol + rehyd H2O H2O + rehyd

~1.0 (1.2 g/kg LBM)

~24.1 (29.1 mL/kg LBM); additional fluid consumed during exer

1862

Initial glycerol bolus at time 0, then fluid consumed over 30 min; exer began 30 min after final fluid intake

H2O (aspartameflavoured)

No difference in physiological or thermoregulatory responses between treatments, no difference in total urine vols between GIH & WIH

Treadmill walking at 1.56–1.65 m/s at 4–9% grade . (= 45% VO2max) [lab based]

Latzka et al.,[32] 1998

8 M, endurance trained, heat-acc

Euhydration (control) Glycerol H 2O

~1.0 (1.2 g/kg LBM)

~24.1 (29.1 mL/kg LBM)

1862

Initial glycerol bolus at time 0, then fluid consumed over 30 min; exer began 30 min after final fluid intake

H2O (aspartameflavoured)

No differences between GIH and WIH for › total body water. No difference in physiological or thermoregulatory responses between treatments. Glycerol › endurance time 14.5% compared with control (33.8 vs 29.5 min). Both GIH and WIH ﬂ heart rate over control

Treadmill walking to exhaustion at 1.56–1.65 m/s at 4–9% grade wearing chemical protective clothing (~55% . VO2max) [lab based]

Hitchins et al.,[33] 1999

8 M, endurance trained, heat-acc

Glycerol Placebo

1.0

22

1628

Glycerol solution consumed over 30 min; exer began 120 min after final fluid intake

Diluted CHOelectrolyte solution

In GIH, ~650 mL (48%) and 500 mL (38%) of ingested fluid retained after 90 and 120 min. GIH ingestion › fluid retention by 250 mL (90 min) and 600 mL

Cycle for 30 min fixed power output, + 30 min self-paced variable power output (lab based)

Continued next page

119

Sports Med 2010; 40 (2)

Latzka et al.,[31] 1997

Guidelines for Glycerol Use

ª 2010 Adis Data Information BV. All rights reserved.

Table I. Contd

120

ª 2010 Adis Data Information BV. All rights reserved.

Table I. Contd Study, year

Subjects

Treatments

Glycerol dose (g/kg BW)

Fluid vol with glycerol (mL/kg BW)

Mean total fluid vol (mL)

Hydration regimen

Fluid with glycerol

Major findings

Exercise protocol

(120 min) vs WIH (p < 0.05), › performance by 5% in variable workload phase 4M + 2 F

Glycerol pre+ 5% glucose during exer; Glycerol pre+ 0.5% glycerol, 5% glucose during exer; Glycerol pre+ 1.5% glycerol, 5% glucose during exer; H2O pre- + 5% glucose during exer

1.2 (1.0 initial + 0.2 at 60 min)

26

1768

Initial glycerol solution consumed over 30 min, then additional glycerol dose at 60 min mark. Total fluid ingestion over 120 min from the start of the glycerol solution intake; exer began immediately after final fluid intake

H2O

In GIH, ~1000 mL (57%) of ingested fluid retained after 2 h. Fluid retention › with GIH vs WIH (~600 mL; p < 0.05), serum osmolality › with glycerol, no effect on ADH. Continued glycerol ingestion during exercise › stroke vol, ﬂ heart rate

Semirecumbent cycling at 44% . VO2max for 110 min (lab based)

Anderson et al.,[35] 2001

6 M, endurance trained

Glycerol H2O

1.0

20 + CHOelectrolyte drink during exercise

1440

Glycerol solution consumed over 15 min; exer began 120 min after final fluid intake

Low joule cordial mixed with H2O

In GIH, ~350 mL (25%) of ingested fluid retained after 2 h. GIH ﬂ pre-exercise urine vol vs WIH (~400 mL; p < 0.05), › forearm blood flow, ﬂ heart rate during exercise, ﬂ rectal temperature late in exercise, ﬂ skin temperature late in exercise, ﬂ ANP, › performance by 5%

Cycle at 98% LT for 90 min, followed by max effort for 15 min (lab based)

Continued next page

van Rosendal et al.

Sports Med 2010; 40 (2)

Montner et al.,[34] 1999a

Subjects

Treatments

Glycerol dose (g/kg BW)

Fluid vol with glycerol (mL/kg BW)

Mean total fluid vol (mL)

Hydration regimen

Fluid with glycerol

Major findings

Exercise protocol

Coutts et al.,[36] 2002

7 M + 3 F, endurance trained

Glycerol Placebo

1.2

25

1955

Glycerol solution consumed over 60 min; exer began 70 min after final fluid intake

Diluted CHOelectrolyte solution

In GIH, ~920 mL (47%) of fluid was retained after 2 h on the warm day and ~1080 mL (55%) on the hot day. GIH › fluid retention 350 mL ( › plasma vol, ﬂ urine) vs WIH (p < 0.05). Glycerol reduced the › in completion time between hot and warm conditions for ODT, no difference in sweat rates

ODT (1.5 km swim, 40 km cycle, 10 km run) [field based]

Magal et al.,[37] 2003

11 M, endurance trained

Three phases (i) 1.0 per trial: (ii) – (i) hyperhydration (iii) 0.5 with/without glycerol (ii) dehydration (iii) rehyd with/without glycerol

(i) 22 (iii) 10 (iii) 11

(i) 1703 (iii) 774 (iii) 851

Glycerol solution consumed over 15 min, then fluid consumed over the next 135 min. Exer began immediately after final fluid intake. Additional fluid consumed during exer. During rehyd, glycerol solution consumed over 15 min with total fluid consumption over 90 min

(i) H2O (flavoured) (ii) CHOelectrolyte solution (iii) H2O (flavoured)

In GIH, ~1100 mL (65%) of ingested fluid retained after 2.5 h. GIH › fluid retention ( ﬂ urine production by ~900 mL) vs WIH (p < 0.05), › plasma vol (~7%), no performance benefits

Specific skill and agility tests followed by 75 min tennis match (field based)

Continued next page

121

Sports Med 2010; 40 (2)

Study, year

Guidelines for Glycerol Use

ª 2010 Adis Data Information BV. All rights reserved.

Table I. Contd

122

ª 2010 Adis Data Information BV. All rights reserved.

Table I. Contd Subjects

Treatments

Glycerol dose (g/kg BW)

Fluid vol with glycerol (mL/kg BW)

Mean total fluid vol (mL)

Hydration regimen

Fluid with glycerol

Major findings

Exercise protocol

Marino et al.,[38] 2003

6 M + 1 F, endurance trained

Glycerol Placebo

1.2

21

1655

Glycerol solution consumed over 150 min; exer began immediately after final fluid consumption

Concentrated OJ mixed with H2O

In GIH, ~1350 mL (82%) of ingested fluid retained after 2.5 h. GIH ﬂ urine output (~118 mL) vs WIH (p < 0.05), › % change in blood vol after 60 min, › heart rate during high-intensity efforts, › sweat rates. No difference between trials for total distance cycled, rectal temperature, mean skin temperature, power produced, perceived exertion, lactate or glucose

Cycle 60 min with aim to complete greatest distance possible. 1 min sprints at 10, 20, 30, 40, 50, 60 min marks (lab based)

Wingo et al.,[17] 2004

12 M, endurance trained, heat-acc

H2O + glycerol pre- + H2O during exer H2O prebut not during exer; H2O pre+ during exer;

1.0

2.8% BW

2153 (+ up to 1200 mL per 10mile loop (3600 mL total ad libitum)c

Glycerol solution consumed over 120 min; exer began 35 min after final fluid intake

H2O (flavoured)

In GIH, ~1350 mL (63%) of ingested fluid retained after 130 min. GIH ﬂ pre-exercise urine vol (~200 mL) vs WIH (p < 0.05), ﬂ post-exer thirst, ﬂ dehydration postexercise, ﬂ postexercise environmental symptoms questionnaire score, ﬂ time for final 10 miles by 5 min (p > 0.05); heart rate and rectal temperature not altered during exercise

3 · 10 mile loop mountain bike race (8 min break between loops) [field based]

Continued next page

van Rosendal et al.

Sports Med 2010; 40 (2)

Study, year

Subjects

Treatments

Glycerol dose (g/kg BW)

Fluid vol with glycerol (mL/kg BW)

Mean total fluid vol (mL)

Hydration regimen

Fluid with glycerol

Major findings

Exercise protocol

Goulet et al.,[39] 2006

6 M, endurance trained

Glycerol H2O

1.2

26 mL/kg BW before exer + 500 mL/h during exer

1781 + 830 during exer

Glycerol and fluid intake over 110 min (glycerol solution given at 0, 40 and 80 min marks, H2O given at 20 and 60 min marks); exer began 10 min after final fluid consumption

H2O (aspartameflavoured)

GIH › TBW ~800 mL (45% of ingested fluid) after 110 min. GIH ﬂ urine production by 271 mL pre-exercise (not significant) and 246 mL during exercise (p < 0.05) vs WIH. No effect on sweat rate, rectal temperature, perceived exertion, endurance performance (time to exhaustion) or peak power output

Cycle at 65% . VO2max for 120 min, followed by 5 min break then an incremental cycle to exhaustion (lab based)

Easton et al.,[40] 2007

12 M, endurance trained

Placebo/glycerol Placebo/placebo Creatine/glycerol Creatine/placebo

1.0

~28.6/dayb

2000 mLd

On day of trial, glycerol solution consumed over 60 min, then additional fluid over the next 180 min; further 60 min until exer begand

H2O

GIH › H2O retention vs WIH by 500 mL in placebo trial & by an additional 240 mL in creatine trial (p < 0.05). Glycerol ﬂ heart rate, rectal temperature and perceived exertion. No performance benefit

Cycle at 63% Wmax for 40 min, followed by 16.1 km (10 mile) time trial (lab based)

Nishijima et al.,[41] 2007 Experiment 1

10 M, healthy adults

Glycerol 1 Glycerol 2 Placebo

1.2

25

1651

Placebo and glycerol 1 solution (glycerol mixed in the full vol of fluid) consumed over 60 min. Glycerol 2 was

Diluted CHOelectrolyte solution

In GIH, ~600–800 mL (36–48%) of ingested fluid retained after 3 h. GIH 1 ﬂ urine vol by 671 mL vs WIH (p < 0.05). GIH 2 ﬂ urine vol by 843 mL vs WIH (p < 0.05) but GIH 1 & GIH 2 were not

No exer component

Continued next page

123

Sports Med 2010; 40 (2)

Study, year

Guidelines for Glycerol Use

ª 2010 Adis Data Information BV. All rights reserved.

Table I. Contd

124

ª 2010 Adis Data Information BV. All rights reserved.

Table I. Contd Study, year

Subjects

Treatments

Glycerol dose (g/kg BW)

Fluid vol with glycerol (mL/kg BW)

Mean total fluid vol (mL)

Hydration regimen

Fluid with glycerol

6 middledistance runners

Glycerol Placebo

1.2

Dini et al.,[42] 2007a

14 M, national level oarsmen

Glycerol pre+ H2O during exer Glycerol pre+ glycerol during exer H2O pre- + H2O during exer

1.0 pre28.5 pre- + 4.5 during exer + 1.0 during exer

25

Exercise protocol

significantly different (p > 0.05)

1.0 g/kg BW glycerol in 8 mL/kg BW fluid bolus within 30 min, then additional fluid over the next 60 min (with additional glycerol 0.2 g/kg BW ingested with fluid 60 min after starting hydration) Nishijima et al.,[41] 2007 Experiment 2

Major findings

Placebo solution consumed over 90 min. Glycerol was the glycerol 2 protocol from experiment 1

Diluted CHOelectrolyte solution

In GIH, ~850 mL (51%) of ingested fluid retained after 3 h. GIH › BW by 790 mL vs WIH (p < 0.05), and › average power by 9% (p > 0.05) vs WIH

Cycle for 40 min fixed power output, + 30 min self-paced variable power output (lab based)

2500 pre- + 400 during exer

Pre-exercise glycerol solution consumed over 90 min, then 180 min until exer began. Additional 2 · 200 mL solutions consumed at ~30 and 60 min marks during exer

H2O

Following the exercise test, fluid retention › 630 mL in GIH compared with WIH (p < 0.05). Glycerol to the rehydration beverages during exercise significantly ﬂ fluid loss (~300–525 mL), and improved work capacity at the anaerobic threshold, compared with both of the other hydration regimens

89 min rowing protocol: 6 · 3 min blocks, with 1 min rest between each, starting at 250 W and › to 400 W in 30 W increments; repeated 3 times with 15 min break between each (lab based)

Continued next page

van Rosendal et al.

Sports Med 2010; 40 (2)

1651

Study, year

Subjects

Treatments

Glycerol dose (g/kg BW)

Fluid vol with glycerol (mL/kg BW)

Mean total fluid vol (mL)

Hydration regimen

Fluid with glycerol

Major findings

Exercise protocol

Goulet et al.,[43] 2008

5 M + 1 F, endurance trained

Glycerol Control (euhydration)

1.2

26 pre- + 12.5 during exer

1776 pre- + 853 during exer

At 0 and 20 min, subjects drank 9 and 6 mL/kg BW fluid, each with 0.6 g/kg BW glycerol, then drank 6 mL/kg BW H2O at 40 and 60 min marks; further 30 min until exercise

H2O (aspartameflavoured)

In GIH, ~1100 mL (62%) of ingested fluid retained 30 min after final fluid consumption. Glycerol › endurance time and peak power and ﬂ HR and thirst. No effect on rectal temperature (tended to stay lower with glycerol), sweat rate, thermal stress or RPE

Cycle for 120 min at 65% . VO2max with 5 · 2 min intervals at 80% . VO2max performed at 12, 32, 52, 72 and 92 min marks); then incremental test to exhaustion (lab based)

Guidelines for Glycerol Use

ª 2010 Adis Data Information BV. All rights reserved.

Table I. Contd

Glycerol consumption during exercisea 4 M + 5 F, healthy adults

H2O 10% glycerol; 6% CHO; 6% CHO + 4% glycerol

~0.94 (1.2 g/kg LBM)

3/kg LBM

647

Glycerol solution consumed over initial 60 min of exer

CHOelectrolyte solution

Glycerol ﬂ the % change in plasma vol and ﬂ overall mean ratings of perceived thirst but no substantial urine, metabolic, hormonal, cardiovascular or thermoregulatory advantages to the consumption of solutions containing 4% or 10% glycerol during exercise

Cycle at 51.8% . VO2peak for 90 min (lab based)

Siegler et al.,[45] 2008e

10 M, experienced soccer players

Glycerol (5.2% solution) Placebo

~0.7

13.4

1000

500 mL of glycerol or placebo solution consumed within 30 min before training, then a further 500 mL consumed after 30 min of training

CHOelectrolyte solution

Glycerol solution attenuated the drop in plasma vol and BW during the training session. No effect on HR during exercise or fatigue test

Variable intensity training drills followed by intermittent endurance test (field based)

Continued next page

125

Sports Med 2010; 40 (2)

Murray et al.,[44] 1991

126

ª 2010 Adis Data Information BV. All rights reserved.

Table I. Contd Study, year

Subjects

Treatments

Glycerol dose (g/kg BW)

Fluid vol with glycerol (mL/kg BW)

Mean total fluid vol (mL)

Hydration regimen

Fluid with glycerol

Major findings

Exercise protocol

Glycerol use in rehydrationf Scheett et al.,[46] 2001

8 M, non-acc

Glycerol Placebo

1.0

3% BW (100% fluid lost during dehydration)

2487

Glycerol solution consumed over initial 30 min, then fluid consumed over the next 150 min; exer began immediately after final fluid consumption

1st 30% aspartameflavoured H2O, remaining 70% H2O

Glycerol ingestion ﬂ urine production, › plasma vol and BW restoration, better rehyd index with glycerol. Also, ﬂ rectal temperature (but no thermoregulatory benefits) and significantly longer exer time to exhaustion (12.6%) with glycerol

Cycle to exhaustion at . 50% VO2peak, cadence 60 rpm (lab based)

Kavouras et al.,[21] 2005

8 M, endurance trained, heat-acc

Glycerol H2O No fluid

1.0

3% BW (75% of fluid lost during dehydration)

2103

Glycerol solution consumed over initial 15 min, then fluid consumed over the next 80 min; exer began 30 min after final fluid consumption

1st 1/3 (glycerol solution) aspartameflavoured H2O, remaining 2/3 was plain H2O

Plasma vol › more and remained higher with glycerol, cutaneous vascular conductance was › with glycerol but no other thermoregulatory mechanisms. No effect on fluid-regulating hormones with glycerol. Glycerol ingestion significantly › time to exhaustion (18%)

Cycle to exhaustion at . 74% VO2peak, cadence 80–100 rpm (lab based)

a

Lyons et al.,[20] Montner et al.[34] and Dini et al.[42] also gave glycerol during exercise following glycerol-induced hyperhydration.

b

Assuming average 70 kg subjects.

c The full 1200 mL of fluid was consumed at the end of the loop on only three occasions out of the 72 trials. Average fluid consumed ranged from 693 mL to 842 mL across the trials. Average total fluid consumed was 4458 mL for water trial, 4643 mL for glycerol trial and 2240 mL for no fluid during exercise trial. Easton et al.[40] 2 · 500 mL daily of treatment solution for 6 days pre-trial + 500 mL of treatment solution with 2 · 500 mL H2O on day of trial. Siegler et al.[45] also gave 500 mL of glycerol solution before exercise.

f

Magal et al.[37] also gave glycerol during rehydration following hyperhydration and dehydration in their three-part study.

ADH = antidiuretic hormone; ANP = atrial natriuretic peptide; BP = blood pressure; BW = bodyweight; CHO = carbohydrate; exer = exercise; F = females; GIH = pre-exercise glycerol hyperhydration; H2O = water; heat-acc = heat acclimated; HR = heart rate; lab = laboratory; LBM = lean body mass; LT = lactic threshold; M = males; non-acc = non-heat acclimated; ODT = Olympic distance triathlon (1.5 km swim, 40 km bicycle, 10 km run); OJ = orange juice; rehyd = rehydration; RPE = rating of perceived exertion; rpm = revolutions per minute; . TBW = total body water; VO2max = maximum oxygen consumption; vol = volume; WIH = pre-exercise water hyperhydration; Wmax = reported as maximal workload; › indicates increase; ﬂ indicates decrease.

van Rosendal et al.

Sports Med 2010; 40 (2)

d e

Guidelines for Glycerol Use

127

Table II. Scale to evaluate original investigations for factors associated with the minimization of bias in selection of subjects, performance and analysis of resultsa 1

A clear description of the inclusion and exclusion criteria was provided

Yes &

No/not sure &

NA &

2.

The trials were randomized

Yes &

No/not sure &

NA &

3.

The method used to generate the random allocation sequence, including details of any restrictions (e.g. blocking, stratification) was described

Yes &

No/not sure &

NA &

4.

Sample size was justified (e.g. by power calculation)

Yes &

No/not sure &

NA &

5.

Attempts were made to control and/or monitor pre-trial conditions (e.g. diet, exercise)

Yes &

No/not sure &

NA &

6.

Design incorporated measures of important baseline variables

Yes &

No/not sure &

NA &

7.

There was blinding of all subjects

Yes &

No/not sure &

NA &

8.

There was blinding of all investigators involved in the trials

Yes &

No/not sure &

NA &

9.

Both the method of blinding and the evaluation of the successfulness of blinding were described

Yes &

No/not sure &

NA &

10.

Details were provided regarding the inability of a subject to complete study requirements

Yes &

No/not sure &

NA &

11.

Statistical methods used to compare groups for primary outcome measure(s),b and methods for additional analyses, such as subgroup analyses and adjusted analyses, were described

Yes &

No/not sure &

NA &

12.

Both point measures and measures of variability for the primary outcome measure(s)a were provided

Yes &

No/not sure &

NA &

13.

The results of between-group statistical comparisons were reported for the primary outcome measure(s)b [e.g. an estimated effect size], and its precision (e.g. 95% CI)

Yes &

No/not sure &

NA &

14.

The method used to assess adverse effects was described

Yes &

No/not sure &

NA &

15.

Reproducibility of the primary outcome measure(s)b was reported

Yes &

No/not sure &

NA &

16.

If a performance test was used, a familiarization trial was conducted

Yes &

No/not sure &

NA &

a

Scoring: the final percentage score is determined by dividing the number of ‘yes’ scores by total number of applicable items (16 minus the number of NA items).

b

If primary outcome measure not stated then key measure used.

NA = not applicable.

and found that a fluid volume of 26 mL/kg BW would maximize fluid retention. The current recommendations will therefore be based upon this volume. Two other studies have investigated the inclusion of glycerol with common pre-exercise hydration regimens.[45,54] While these are not specifically categorized as pre-exercise hyperhydration studies because of the small total fluid intake, they do provide a practical application for athletes. The first study investigated glycerol (1.0 g/kg BW) consumption with a typical pre-race fluid intake.[54] There were no hydration or performance benefits, most likely due to the smaller total fluid intake (1250 mL) combined with the bulk of the fluid being ingested 4 hours before exercise, allowing considerable time for fluid to be cleared by the kidneys before exercise commenced.[54] However, when consuming 500 mL of a glycerol ª 2010 Adis Data Information BV. All rights reserved.

solution within the 30 minutes preceding a 60-minute training session, and another 500 mL half way through training, Siegler et al.[45] showed that glycerol solutions can attenuate the BW and plasma volume losses during exercise. Thus, if an athlete is unable to tolerate the large volumes of fluid (26 mL/kg BW) and glycerol required to induce pre-exercise hyperhydration, then they may consider consuming smaller volumes of fluid and glycerol closer to the commencement of activity. 1.3 Type of Fluid

Carbohydrate-electrolyte beverages (sports drinks) may provide better hydration potential than water alone during prolonged endurance exercise.[55,56] The major factors affecting the absorptive potential of fluids include: type and concentration of carbohydrate, solution osmolality, Sports Med 2010; 40 (2)

128

ª 2010 Adis Data Information BV. All rights reserved.

Table III. Evaluation of the studies investigating glycerol use for pre-exercise hyperhydration and for rehydration during and after exercise Stats Data Eligibility Randomizationa,b Power Pre-trial Baseline Blindingc,d,e NonAdverse Reproducibility Familiarization reported performance calc. conditions measures completers described reportingf,g effects test described described

% score Reference

Pre-exercise hyperhydration 1

1a

0

0

1

0

NA

0

2f,g

0

0

NA

29

1

0

1

a

0

0

1

0

NA

1

2f,g

0

0

NA

36

20

0

1

a

0

0

1

1

1

1

2

f,g

0

0

1

50

29

0

0

0

1

1

2c,d

NA

1

2f,g

0

1

NA

57

26

0

0

0

1

0

0

NA

1

2f,g

0

0

NA

29

47

f,g

c

c

0

0

0

1

0

1

NA

1

2

0

0

NA

36

47

0

1a

0

1

1

2c,d

NA

1

1f

0

0

0

47

30

0

1a

0

1

1

2c,d

NA

1

1f

0

0

0

47

30

0

1

a

1

2

c,d

1

1

2f,g

0

0

0

56

31

1

1

a

0

1

1

2

c,d

NA

1

2

f,g

0

0

0

60

32

0

1a

1

1

0

2c,d

NA

1

2f,g

0

1

NA

64

34

0

0

0

1

1

2c,d

NA

1

2f,g

0

0

1

53

33

0

2a,b

0

1

1

3c,d,e

NA

1

2f,g

1

0

0

73

35

0

1

a

1

c,d

NA

1

2f,g

1

0

1

67

36

0

1

a

0

1

0

0

NA

1

2

f,g

0

0

NA

36

27

0

1a

0

1

1

3c,d,e

NA

1

2f,g

0

1

1

73

38

0

1a

0

1

1

2c,d

NA

1

2f,g

0

0

1

60

37

0

1

a

0

0

1

2

c,d

NA

1

2f,g

0

0

0

47

17

0

1a

0

1

1

2c,d

NA

1

2f,g

1

0

0

60

39

1

1

a

0

1

1

3

1

1

2

f,g

0

1

1

81

40

0

1a

0

1

1

2c,d

0

1

2f,g

0

0

0

50

41

0

1

a

0

1

0

0

NA

1

2

f,g

0

0

0

33

42

0

1a

0

1

1

0

NA

1

2f,g

1

1

0

57

43

0

0

1

1

2

c,d,e

0

0

0

1

1

1c

NA

1

2f,g

1

0

1

53

44

0

1a

0

1

1

2c,d

NA

1

2f,g

0

0

1

60

45

Continued next page

van Rosendal et al.

Sports Med 2010; 40 (2)

Glycerol during exercise

ª 2010 Adis Data Information BV. All rights reserved.

Reporting group comparisons (item 13).

f

g

calc. = calculation; NA = not applicable; stats = statistics.

Blinding described (item 9).

Reporting point measures (item 12).

e

Blinding of subjects (item 7).

Blinding of investigators (item 8). d

Randomization described (item 3).

c

Randomization (item 2).

b

129

a

21

46

Scoring: Each column, except the randomization, blinding and data reporting columns, have a maximum score of 1 point. The randomization column has a maximum score of 2 points as it combines items 2 and 3 from table II. The blinding column has a maximum score of 3 points as it combines items 7, 8 and 9 from table II. The data reporting column has a maximum score of 2 points as it combines items 12 and 13 from table II. The following denotes which specific criteria were fulfilled for the columns combining more than 1 item:

60

53 0

0 0

0 0

0 2f,g 1 NA 2 1 1 0 1 1

2f,g 1 NA 2c,d 1 1 0

c,d a

1a 0

Glycerol during rehydration

Baseline Blindingc,d,e NonAdverse Reproducibility Familiarization Eligibility Randomizationa,b Power Pre-trial Stats Data calc. conditions measures completers described reportingf,g effects reported performance described described test

Table III. Contd

% score Reference

Guidelines for Glycerol Use

pH, electrolyte concentration, volume of fluid and palatability (flavour, temperature, colour).[56] Given that glycerol alters the osmolality and caloric density of solutions it is dissolved in, it is also pertinent to understand the effect it will have on fluid absorption. Two studies have demonstrated that glycerol may provide a pro-absorptive effect, evidenced through enhanced intestinal water and sodium absorption in rat intestinal perfusion models.[57,58] Wapnir et al.[58] found that glycerol solutions attenuated water secretion into the perfused segment of the small intestine and significantly reduced the outflow of sodium into the lumen, when compared with sports drinks without glycerol. Furthermore, glycerol solutions had a higher influx-to-efflux ratio resulting in a net influx of water (from the intestinal lumen to the circulation), while glucose solutions had a lower influx-to-efflux ratio resulting in a net efflux of water (into the intestinal lumen).[58] Allan et al.[57] tested the absorption of oral rehydration solutions containing 0.75 mmol/L sodium and the following combinations of glucose : glycerol: 75 : 0, 50 : 25, 37.5 : 37.5, 25 : 50, 10 : 65 and 0 : 75 (mmol/L). Substituting glycerol for glucose enhanced both net sodium and water absorption in rat perfusion models, with peak absorption of both sodium and water occurring with the glucose : glycerol ratio of 25 : 50. Thus, the combination of glycerol and glucose enhanced absorption of fluid from the intestinal lumen to the circulation compared with the solution containing glycerol without glucose, and vice versa.[57] While further investigation in humans is required, the similarity of the physiological process associated with sodium and water movement from the intestinal lumen in rats and humans suggests that the results from these trials can be applied to humans with some confidence,[58] although a published abstract indicates that this may not be the case.[59] In addition, the inclusion of sodium in sports drinks might improve gastrointestinal glucose and fluid absorption,[56] and offer some protection against hyponatraemia. Hyponatraemia typically results from over-hydration with hypotonic fluids during exercise (i.e. drinking water and losing sodium in the sweat).[60-63] The risk of Sports Med 2010; 40 (2)

van Rosendal et al.

130

hyponatraemia is very high when so much fluid is consumed that weight gain during exercise exceeds 2.5% BW.[63,64] It has been advocated that drinking sports drinks containing high levels of sodium during prolonged (>2 hours) exercise may help to stabilize the sodium content of the extracellular fluid and assist in the prevention of hyponatraemia.[65-67] One proposed negative effect of drinking sports drinks prior to exercise is the possible development of hypoglycaemia. However, a recent series of experiments showed that glucose (75 g) consumed 45 minutes prior to exercise resulted in the stabilization of plasma glucose at a level that was not considered hypoglycaemic and was unaffected by exercise intensity.[68] Subsequent trials showed that rebound hypoglycaemia was (i) more evident following glucose ingestion when compared with carbohydrates with a lower glycaemic index;[69] (ii) equally evident following glucose doses of 25, 75 and 200 g;[70] and (iii) more evident when glucose was consumed longer before exercise (75 > 45 > 15 minutes).[71] However, in all of these conditions, hypoglycaemia was only present for the initial 10–20 minutes during steady-state exercise and had no negative effects on subsequent time trial performance lasting approximately 40 minutes.[69-71] Hargreaves et al.[72] have stated that if ingesting carbohydrate pre-exercise, then a reasonable amount (100 g) should be consumed to provide a supply of glucose later in exercise to offset the suppression of fat oxidation. Most glycerol hyperhydration studies have used aspartame-flavoured water as the hyperhydrating beverage and have consistently induced hyperhydration. Given the rationale that carbohydrate-electrolyte drinks may provide a better hydration potential than water alone, as well as providing sodium to attenuate hyponatraemia, it could also be proposed that consuming carbohydrate-electrolyte beverages with glycerol could further enhance the hyperhydration benefits of glycerol. 1.4 Timing of Fluid with Glycerol

Two different fluid ingestion protocols are used to promote pre-exercise hyperhydration. ª 2010 Adis Data Information BV. All rights reserved.

The first involves quickly ingesting a small concentrated bolus of glycerol solution, then consuming the remaining fluid over a longer duration. The second involves mixing the glycerol within the full volume of fluid, thereby ingesting glycerol with fluid throughout the pre-exercise hyperhydration period. Study methodologies differ greatly in the duration over which these fluids are provided, making it difficult to determine the optimal protocol for fluid consumption relative to glycerol intake. Table I provides details of different protocols used. One study investigated the difference between these styles of fluid and glycerol delivery and found that the change in bodyweight and urine volume were similar with either mode.[41] However, the results might be confounded by the fact that the duration over which fluids were consumed was longer in the glycerol bolus trial (90 vs 60 minutes).[41] Thus, the optimal period for fluid consumption in relation to the glycerol bolus intake also needs to be considered. The second protocol used in the abovementioned study by Riedesel et al.[1] provided the glycerol bolus that was optimal in the first series of trials (1.0 g/kg BW), with a larger volume of fluid (25.7 mL/kg BW of 0.1% NaCl solution) over a longer duration (3.5-hour period). This fluid regimen reduced hyperhydration to a level similar to the ingestion of 0.5 g/kg BW of glycerol and fluid within 40 minutes (21.4 mL/kg BW of 0.1% NaCl solution) used in the first protocol. However, others have used similar methods with much greater success.[28,37] Mixing the glycerol through the entire hyperhydration beverage in order to spread the glycerol consumption over a longer period also results in fluid retention.[17,38,41] However, these longer protocols tend to measure hyperhydration immediately after fluid ingestion is completed, while the shorter fluid intakes tend to measure hyperhydration after lengthy equilibration periods. It is not surprising that the percentage of fluid retained is higher if you measure immediately after fluid consumption. Of the five studies scoring highest in the quality assessment, glycerol and fluid were given over 15,[35] 60,[36] 120,[34] 150[38] or 300 minutes.[40] Fluid retention Sports Med 2010; 40 (2)

Guidelines for Glycerol Use

was greatest when glycerol and fluid was ingested over 60–150 minutes. Noakes et al.[73] highlighted the importance of gastric volume in regulating gastric emptying. Briefly, the maximum rate at which fluid can be delivered from the stomach is significantly influenced by the volume of fluid in the stomach, and therefore by the rate of fluid consumption.[73] The data from Noakes et al. indicate that the rate of water delivery from a 7% carbohydrate solution will be approximately 400 mL per 10 minutes if a gastric volume of approximately 800 mL is maintained.[73] Therefore, an individual should aim to consume 600–800 mL within about 10 minutes of starting the hydration period, and then a further 400 mL each 10- to 15-minute period during the next 50 minutes. This will result in a volume of fluid equal to 26 mL/kg BW (approximately 1820 mL for a 70 kg individual) being consumed within 60 minutes and absorbed within approximately 90 minutes from the onset of fluid intake. If an individual is unable to stomach volumes of this magnitude, then a smaller volume of fluid should be consumed in each 10-minute period, focusing on drinking the fluid as rapidly as possible. 1.5 Duration of Hyperhydration

Glycerol ingestion has been associated with hyperhydration for periods of up to 4 hours.[1,17,20,26,30,33,35-38,74] Peak hyperhydration is determined by the relationship between glycerol and fluid absorption, and clearance. It is difficult to determine when peak hyperhydration occurs, because the timing of measurements of fluid retention varies greatly between studies. Several studies have mapped fluid retention for periods of up to 3 hours after ingestion, without the confounding of an exercise protocol beginning within this period. After 2 and 3 hours, 60–80%[1,20,26] and 45–60%,[1,20,26] respectively, of the ingested fluid was still retained in the glycerol trial (compared with 40–60% and 10–30%, respectively, in the water hyperhydration trial).[26] Because the absorption of a large volume of fluid will occur rapidly when fluid intake is rapid (as discussed in section 1.4),[73] exercise should ª 2010 Adis Data Information BV. All rights reserved.

131

commence 30 minutes after final fluid consumption. Waiting for an extended period will result in unnecessary water loss and reduce overall hyperhydration. Waiting 30 minutes after fluid intake should allow sufficient time for the sensation of stomach fullness to subside, while ensuring that exercise begins when hyperhydration is close to maximal. It has been demonstrated that it is possible to maintain hyperhydration using glycerol for up to 49 hours.[47] For prolonged hyperhydration, water needs to be consumed at either the same or a greater rate than it is lost. Glycerol may assist in this process by decreasing water loss. However, during this time, glycerol would also need to be ingested at rates in excess of its rate of catabolism/ excretion. Although the study by Koenigsberg et al.[47] has demonstrated prolonged hyperhydration with glycerol, until further research into sustained hyperhydration (i.e. >4 hours) is completed, the authors would strongly advise against this practice, as it will theoretically increase the risk of hyponatraemia. Whether this could be overcome by using a carbohydrate-electrolyte beverage has not been studied (see section 4). 1.6 Guidelines for Pre-Exercise Glycerol Hyperhydration

Pre-exercise glycerol hyperhydration will be most advantageous when sweat losses cannot be replaced during exercise. However, if euhydration can be maintained during exercise, then preexercise hyperhydration may not provide any additional advantage. If an athlete commences exercise in a hyperhydrated state, they must ensure that they do not over-drink during exercise. The additional fluid retained with pre-exercise hyperhydration will act to dilute plasma sodium before commencing exercise. If additional large volumes of hypotonic fluid are consumed during exercise, there exists an increase in the potential risk of dilutional hyponatraemia.[18] Given the hypotonic composition of sweat, the extra fluid stored with pre-exercise hyperhydration may be lost during exercise without a concomitant increase in the loss of sodium, assuming that aggressive fluid intakes are not made during exercise. In addition, Sports Med 2010; 40 (2)

van Rosendal et al.

132

Table IV. Guidelines for glycerol and fluid ingestion to promote preexercise glycerol hyperhydration Only undertake a pre-exercise glycerol hyperhydration protocol if the exercise is likely to induce a reduction in bodyweight (BW) >2% Consume a glycerol dose of 1.2 g/kg BW with a volume of fluid equal to 26 mL/kg BW. If this volume is too high for an individual, then consider personalizing the protocol by consuming a smaller volume of fluid and glycerol closer to the commencement of activity Consume the glycerol solution over a period of 60 min Use the normal choice of beverage. Carbohydrate-electrolyte beverages with a relatively high sodium content might provide an additional advantage, however this requires further investigation Commence exercise approximately 30 min after the total hyperhydration fluid volume has been consumed Consider the increased metabolic cost associated with undertaking weight-bearing exercise with elevated BW

the increased chance of having to void during competition should be considered.[18] From the preceding sections, guidelines for glycerol and fluid ingestion to establish preexercise hyperhydration have been formulated and are presented in table IV. 2. Glycerol Ingestion during Exercise The goal of ingesting fluid during exercise should be to replace enough of the fluid lost as sweat to avoid incurring a fluid deficit of >2% BW and/or an electrolyte imbalance that may lead to hyponatraemia.[18] The volume of fluid to be consumed will therefore depend on the sweat rate during the activity. This is determined by the mode and intensity of exercise, the environmental conditions, exercise duration and individual variation.[18] However, guidelines for glycerol consumption during exercise will also depend upon whether the athlete has hyperhydrated preexercise, as this will delay the body water deficit reaching 2% BW. The following sections discuss the use of glycerol during exercise based on whether pre-exercise hyperhydration has taken place. 2.1 Glycerol Ingestion during Exercise, after Pre-Exercise Hyperhydration

Glycerol pharmacokinetics indicate that only small additional doses during exercise are needed to maintain elevated plasma glycerol levels for several hours, when pre-exercise hyperhydraª 2010 Adis Data Information BV. All rights reserved.

tion has been undertaken. For example, when a glycerol dose of 1.2 g/kg BW is used to hyperhydrate pre-exercise, plasma glycerol levels should be elevated to saturation kinetics for around 3 hours.[75] Glycerol will then be half eliminated approximately 140 minutes later,[23,75] although this will be more rapid if exercise begins 30 minutes after fluid ingestion, due to increased metabolism. Three studies have investigated glycerol consumption with fluid during exercise following pre-exercise glycerol hyperhydration (table I).[20,34,42] Montner et al.[34] required subjects to ingest a total of 1.2 g/kg BW of glycerol with water (26 mL/kg BW) pre-exercise in three hyperhydration trials. Subjects ingested a further 5 mL/kg BW every 20 minutes (total 25 mL/kg BW) of solutions containing either (by volume) 5% glucose, 0.5% glycerol (0.125 g/kg BW) in 5% glucose, or 1.5% glycerol (0.375 g/kg BW) in 5% glucose, during 110 minutes of semi-recumbent cycling. While the maintenance of bodyweight during exercise with continual glycerol ingestion was slightly better than with glycerol hyperhydration alone, the benefit was not significant. However, glycerol ingestion during exercise did result in an increased stroke volume and decreased heart rate, indicating improved cardiovascular performance. In support of the above pharmacokinetic data, there was no difference between the 0.5% and 1.5% (0.125 or 0.375 g/kg BW) glycerol solutions.[34] Dini et al.[42] showed that the addition of glycerol to rehydration beverages during exercise significantly increased fluid retention (~525 mL) and improved work capacity at the anaerobic threshold, compared with other hydration regimens. In the Lyons et al.[20] study, the effect of glycerol ingestion during exercise is difficult to establish as no trial existed in which subjects were given glycerol before but not during exercise. Furthermore, no blood samples were taken between the hyperhydration period and the commencement of exercise, so it is impossible to distinguish between the effects of the pre-exercise glycerol bolus and the doses given during exercise. It is recommended that athletes only consume glycerol during exercise if they are going to incur a fluid deficit >2% BW. A 70 kg athlete Sports Med 2010; 40 (2)

Guidelines for Glycerol Use

(with a total body water volume ~42 L) can lose ~1.4 L of fluid (so that total body water would drop to 40.6 L) in order to reach the ‘dehydration threshold’ of a 2% reduction in BW. However, the extra fluid retained with hyperhydration is also available to be lost as sweat, thereby delaying the progression of dehydration. If the same 70 kg athlete hyperhydrated preexercise by ~900 mL (the average volume retained with pre-exercise glycerol hyperhydration for a 70 kg athlete[25]), then their total body water would increase to approximately 42.9 L. This additional fluid is also available to be lost as sweat before they reach their ‘dehydration threshold’ of a reduction in total body water to 40.6 L. Now the athlete can lose ~2.3 L of fluid (1.4 L + the additional 0.9 L stored through hyperhydration) before reaching the same relative level of dehydration compared with when they are normally euhydrated. Based on a sweat rate of 1.5 L/h, exercise durations of approximately 90 minutes would be needed to achieve this fluid deficit. For continued glycerol ingestion with fluid during exercise following pre-exercise hyperhydration, it is recommended the athlete consume a small amount of glycerol (0.125 g/kg BW) in a volume equal to 5 mL/kg BW when exercise is of sufficient duration to dehydrate them by >2% BW. The specific dose is based on the Montner et al.[34] study, as it scored highly (64%) on the scale of research quality compared with the Dini et al.[42] study (33%). Furthermore, it is imperative that athletes avoid drinking any more fluid than a volume that is sufficient to replace sweat loss, so there is no net weight gain during exercise. Thus, if athletes have hyperhydrated pre-exercise, and exercise duration is £75 minutes, then very little fluid would be needed during exercise, and the consumption of glycerol with any fluid is not recommended. The American College of Sports Medicine’s (ACSM) position stand on fluid replacement also advocates the consumption of beverages containing sodium and/or salted snacks to help stimulate thirst and retain fluids.[18] Carbohydrate in rehydration solutions, while not further facilitating rehydration, may slightly improve the intestinal uptake of sodium and water.[76] Replacement of electroª 2010 Adis Data Information BV. All rights reserved.

133

lytes, particularly sodium, is crucial and the inclusion of electrolytes in ingested fluids will help maintain plasma volume.[44,45] 2.2 Glycerol Ingestion during Exercise, without Pre-Exercise Hyperhydration

Athletes starting exercise in a normal euhydrated condition might also benefit from adding glycerol to fluid consumed during exercise, as a means to prevent or delay dehydration by enhancing retention of the ingested fluid. Murray and colleagues[44] are, to our knowledge, the only group to study the specific effects of glycerol on hydration during exercise without any pre-exercise glycerol and fluid ingestion. Siegler et al.[45] gave 500 mL 30 minutes preexercise and a further 500 mL during exercise. The small total volume of fluid consumed (647[44] to 1000 mL[45]) in each of these protocols prohibits them being described as hyperhydration studies. Between them, they provided solutions with glycerol concentrations ranging from 4% to 10% of the fluid volume (4% glycerol with 6% carbohydrate and 10% glycerol;[44] and 5.2% glycerol with 4% carbohydrate[45]). The resulting glycerol doses were 0.38 and 0.94 g/kg BW for the 4% and 10% solutions of Murray et al.[44] and 0.7 g/kg BW for the 5.2% beverage of Siegler et al.,[45] respectively. . Over 90 minutes of cycling at 50% VO2 peak (in 30C, 45% relative humidity environment), both the 4% and 10% glycerol solutions reduced thirst sensation and attenuated the decrease in plasma volume seen with the water placebo and sports drink solutions.[44] From 60 to 80 minutes during exercise the 10% glycerol solution provided better maintenance of plasma volume than did the 4% solution; however, they were similar at all other timepoints. The protocol of Siegler et al.[45] resulted in a 40% reduction in BW loss with the glycerol solution. This reflects equal or improved thermal tolerance in the glycerol trials with less dehydration, as no difference was observed between trials for variables such as core temperature and heart rate. Siegler et al.[45] also showed a 55% reduction in the change in plasma volume over 60 minutes of exercise compared Sports Med 2010; 40 (2)

van Rosendal et al.

134

with the sports drink. Thus, the addition of glycerol to beverages consumed during exercise has the potential to maintain fluid balance to a greater extent than sports drinks or water alone. Guidelines for glycerol use during exercise need to consider the duration of the event. Glycerol metabolism is relatively slow (~8–9 g/h).[77,78] No studies have investigated the continual ingestion of glycerol during longer (e.g. ultra-endurance) events. However, because glycerol is slowly metabolized, it may be expected that continual administration over prolonged periods (e.g. >4 hours) will lead to an accumulation of glycerol in the circulation resulting in multiple side effects, some of which may be health threatening.[44,45] Therefore, we recommend athletes consume glycerol 0.4 g/kg BW with fluid during each of the first 4 hours of exercise. This would provide a similar dose to that recommended during hyperhydration, with adjustment for increased metabolism with exercise. After this time, individuals should consume fluid alone where necessary. The volume consumed each hour will depend on sweat rate, exercise duration and opportunities to drink.[18,79] For an event such as the marathon, the ACSM recommend that smaller persons exercising at a lower intensity in cooler environments need to replace around 0.4 L/h, and this is increased to 0.8 L/h for heavier individuals exercising at higher intensities in warmer environments.[18] Finally, similar to all guidelines presented here, glycerol use should only be considered when dehydration is likely to exceed 2% BW. Thus, using the example of the 70 kg athlete starting exercise in a normal euhydrated condition (i.e. who hasn’t hyperhydrated pre-exercise), glycerol use is recommended if fluid losses are likely to exceed 1.4 L. Based on a sweat rate of 1.5 L/h, an exercise duration of approximately 60 minutes would be required to achieve this fluid deficit. 2.3 Guidelines for Glycerol Ingestion during Exercise

Guidelines for glycerol ingestion will depend on the extent of dehydration during the activity. Because sweat rates are highly variable, athletes should estimate their sweat rate in conditions simiª 2010 Adis Data Information BV. All rights reserved.

lar to those in which performance will take place, to assist in determining a hydration protocol. To do this, weight loss during a session mimicking typical performance should be monitored and corrected for fluid intake, then divided by the duration of the activity. For simplicity, 1 kg weight loss is considered equal to approximately 1 L of fluid. Glycerol use during exercise is then advocated by following the guidelines in table V. 3. Glycerol as a Rehydrating Agent To date, only three studies have explored the role of glycerol in rehydration (table I).[21,37,46] Most recently, Scheett et al.[46] and Kavouras et al.[21] dehydrated subjects via exercise (by -3% and -4% BW, respectively) before rehydrating (3% BW each) with or without glycerol (1 g/kg BW). The third study to investigate glycerol consumption with rehydration fluid was conducted by Magal et al.,[37] employing a protocol that had three stages: (i) hyperhydration with or without glycerol; (ii) exercise-induced dehydration; and (iii) rehydration with or without glycerol (0.5 g/kg BW).

Table V. Guidelines for the inclusion of glycerol in rehydration fluids given during exercise Glycerol use should be considered when exercise is of sufficient duration to dehydrate by >2% bodyweight (BW) If pre-exercise hyperhydration with glycerol has taken place, then consume 0.125 g/kg BW of glycerol in a volume equal to 5 mL/kg BW Drinking fluid at a rate greater than that required to replace sweat loss (leading to a net weight gain during exercise) should be avoided. Therefore, if an athlete is hyperhydrated before exercise lasting £75 min, very little fluid would be needed during exercise under most conditions, and the consumption of glycerol with any fluid is not recommended If no pre-exercise hyperhydration has taken place, then a larger dose of glycerol with fluid during exercise is warranted. Therefore, we recommend athletes consume 0.4 g/kg BW glycerol with fluid during each of the first 4 h of exercise After 4 h, individuals should consume fluid alone where necessary It is recommended that smaller persons exercising at a lower intensity in cooler environments need to replace around 0.4 L/h, and this goes up to 0.8 L/h for heavier individuals exercising at higher intensities in warmer environments Continual administration of glycerol over prolonged periods (e.g. >4 h) may lead to an accumulation of glycerol in the circulation. As such, continuing glycerol ingestion after 4 h is not advised

Sports Med 2010; 40 (2)

Guidelines for Glycerol Use

Each of these studies found that beverages containing glycerol were associated with significantly more rapid and complete restoration of plasma volume than water alone. In the Magal et al. study,[37] subjects became dehydrated by -2% and -3% BW for the glycerol and placebo trials, respectively, with corresponding plasma volume changes of -11% and -13%. Following rehydration, plasma volume remained at -4% compared with euhydrated baseline for the water alone trial, but increased to +2% for the glycerol trial. This restoration and expansion of plasma volume to greater than baseline levels in the glycerol trial occurred even though the subjects were still in a state of whole body dehydration (at the end of rehydration, subjects were still -1.5% BW for the glycerol trial and -2.5% BW for the water trial).[37] As such, plasma volume is restored before the interstitial and intracellular fluid compartments.[37] The effect of glycerol on total body water is less defined. In the Magal et al.[37] study, urine volume was higher (by ~50 mL) and percentage fluid retention was lower (~5%) with glycerol compared with water. In the Kavouras et al.[21] study, urine volume was 385 mL lower with glycerol (p > 0.05), and in the Scheet et al.[46] study there was no difference in urine volume between the conditions. However, the influence of glycerol on total fluid retention might be related to rehydration duration. Kavouras et al.[21] and Magal et al.[37] monitored passive rehydration over 80 and 90 minutes, respectively. The data of Scheet et al.[46] indicate that benefits might be more pronounced with longer rehydration periods. In the glycerol trial, approximately 100 mL of urine was produced during each of the 3 hours. In the water trial, <100 mL total was produced in the first 2 hours, then urine output increased dramatically and ~300 mL was produced in the third hour. If this pattern continued, fluid excretion in the water trials would rapidly increase compared with the glycerol trial. Replacing both the fluid and electrolyte deficits should be the goal after exercise.[18] Shirreffs et al.[76] showed that adequate sodium content is important to promote rehydration and to replenish sodium within the body, reducing the risk ª 2010 Adis Data Information BV. All rights reserved.

135

Table VI. Guidelines for fluid and glycerol consumption in rehydration Consume 1.5 L of fluid for each 1 kg of weight loss Add glycerol 1.0 g/kg bodyweight to each 1.5 L of fluid consumed, when a subsequent bout of exercise will be undertaken within a few hours (this will provide similar glycerol doses with fluid volumes as those used in hyperhydration) If there is a long duration between successive exercise bouts, then rehydrate with water and meals and follow the pre-exercise hyperhydration recommendations before the next exercise session

of hyponatraemia. If time permits, then postexercise meals or salty snacks should be included with fluid to replace electrolyte deficits.[18,80] In this case, water should suffice as the rehydration fluid. If meals or snacks cannot be included, then carbohydrate-electrolyte solutions with relatively high sodium content might be more beneficial than drinking water alone at preventing the dilution of sodium, particularly in extracellular fluid.[18] Specific guidelines are difficult because the volume of fluid to be consumed will depend on the fluid deficit from sweat loss. Broad guidelines for fluid and glycerol consumption during rehydration are presented in table VI. 4. Areas for Further Investigation We are recommending that individuals use glycerol with their beverage of choice. This is mainly due to the lack of studies that have compared the use of glycerol with different beverages (e.g. sports drinks vs water). Given that most athletes will use sports drinks due to the provision of additional carbohydrate and electrolytes, it is important to know whether consuming glycerol with a carbohydrate-electrolyte beverage would be more beneficial than glycerol with water. A randomized, controlled trial comparing glycerol in water with glycerol in a sports drink should be conducted. 5. Side Effects from Glycerol Consumption The incidence of side effects associated with glycerol consumption in hyperhydration and rehydration is very low. Table I discusses the 26 studies Sports Med 2010; 40 (2)

van Rosendal et al.

136

that apply to the ingestion of glycerol with fluid consumed before, during or after exercise in warm/hot conditions. A further two investigations have looked at glycerol hyperhydration before exposure to cold environments.[81,82] Three of these 28 studies reported side effects after rapidly administering the glycerol as a concentrated bolus followed by fluid ingestion.[31,32,35] In two of these, a total of four subjects were nauseous after glycerol ingestion, resulting in the cancellation of the trial on that day,[31,32] while in the other, two subjects developed diarrhoea in the 24 hours after the trial.[35] A further three studies reported a low incidence of gastrointestinal distress (bloatedness) or light-headedness that did not affect participation.[36,40,44] Side effects are more frequent when glycerol is consumed rapidly. This was considered when formulating the guidelines to consume solutions over 60 minutes. Many authors have used the doses of glycerol and fluid that we have recommended, over the proposed time frames, without incidence. However, should an athlete be unable to tolerate these protocols, it is recommended that they lower the concentration of glycerol, and/or consume the solution over a period of 90 minutes instead of 60 minutes. Of the two studies using glycerol during exercise without pre-exercise hyperhydration, side effects were seen in one,[44] but not the other.[45] Murray et al.[44] gave a higher dose, with less fluid, which when combined with delayed gastric emptying with exercise may account for the symptoms present. Three other studies hyperhydrated with glycerol pre-exercise and then provided small doses of glycerol with fluid while exercising, and also reported no side effects.[20,30,42] Again, these factors were considered when formulating the guidelines. If the recommended doses cause gastrointestinal upset, then the concentration of glycerol should be lowered. While glycerol has also been used clinically to reduce cerebral oedema, intracranial hypertension and intraocular pressure by drawing water out of the CSF and brain,[83-85] there is very little chance of these events occurring when glycerol is consumed with large volumes of fluid. However, the risk would increase if glycerol accumulates following multiple large glycerol doses over ª 2010 Adis Data Information BV. All rights reserved.

extended periods, especially when dehydrated. Thus, the recommendations are for glycerol use for periods of up to 4 hours. Finally, there are certain populations for which glycerol ingestion is not advised due to actions on liver gluconeogenesis, kidney filtration, cardiovascular homeostasis and hydration homeostasis. These include pregnant females and individuals with diabetes, renal disease, migraine and headache disorders, cardiovascular disease and liver disorders.[23,83] 6. Summary and Recommendations This paper provides guidelines for the use of glycerol before, during and after exercise. Preexercise glycerol hyperhydration is likely to benefit the athlete during exercise in which a 2% BW reduction will occur, by attenuating dehydration. Furthermore, glycerol consumption with fluid during exercise and in rehydration is likely to maintain cardiovascular function and enhance the restoration of body water deficits. However, as with any proposed ergogenic aid, athletes are advised to trial using glycerol during training before deciding to use it during competition. If athletes fall into any of the ‘at-risk’ populations, it is recommended that glycerol is avoided or used only in consultation with their physician. Furthermore, all athletes are advised to discontinue using glycerol if they continually experience any associated side effects. If side effects do present, athletes are advised to trial different glycerol concentrations and fluid volumes to ascertain whether they may be able to develop an individualized protocol to still gain some benefit. Acknowledgements The authors would like to acknowledge the University of Queensland Graduate School Research Travel Grant for financial assistance. The authors have no conflicts of interest that are directly relevant to the content of this review.

References 1. Riedesel ML, Allen DY, Peake GT, et al. Hyperhydration with glycerol solutions. J Appl Physiol 1987 Dec; 63 (6): 2262-8

Sports Med 2010; 40 (2)

Guidelines for Glycerol Use

2. Gonzalez-Alonso J, Mora-Rodriguez R, Below PR, et al. Dehydration reduces cardiac output and increases systemic and cutaneous vascular resistance during exercise. J Appl Physiol 1995 Nov; 79 (5): 1487-96 3. Nadel ER, Fortney SM, Wenger CB. Effect of hydration state of circulatory and thermal regulations. J Appl Physiol 1980 Oct; 49 (4): 715-21 4. Sawka MN, Young AJ, Francesconi RP, et al. Thermoregulatory and blood responses during exercise at graded hypohydration levels. J Appl Physiol 1985 Nov; 59 (5): 1394-401 5. Gonzalez-Alonso J, Calbet JA, Nielsen B. Muscle blood flow is reduced with dehydration during prolonged exercise in humans. J Physiol 1998 Dec 15; 513 (Pt 3): 895-905 6. Gonzalez-Alonso J, Calbet JA. Reductions in systemic and skeletal muscle blood flow and oxygen delivery limit maximal aerobic capacity in humans. Circulation 2003 Feb 18; 107 (6): 824-30 7. Hubbard RW. The role of exercise in the etiology of exertional heatstroke. Med Sci Sports Exerc 1990 Feb; 22 (1): 2-5 8. Coris EE, Ramirez AM, Van Durme DJ. Heat illness in athletes: the dangerous combination of heat, humidity and exercise. Sports Med 2004; 34 (1): 9-16 9. Armstrong LE, Casa DJ, Millard-Stafford M, et al. American College of Sports Medicine position stand: exertional heat illness during training and competition. Med Sci Sports Exerc 2007 Mar; 39 (3): 556-72 10. Walsh RM, Noakes TD, Hawley JA, et al. Impaired highintensity cycling performance time at low levels of dehydration. Int J Sports Med 1994 Oct; 15 (7): 392-8 11. Below PR, Mora-Rodriguez R, Gonzalez-Alonso J, et al. Fluid and carbohydrate ingestion independently improve performance during 1 h of intense exercise. Med Sci Sports Exerc 1995 Feb; 27 (2): 200-10 12. Armstrong LE, Costill DL, Fink WJ. Influence of diureticinduced dehydration on competitive running performance. Med Sci Sports Exerc 1985 Aug; 17 (4): 456-61 13. Cheuvront SN, Carter 3rd R, Castellani JW, et al. Hypohydration impairs endurance exercise performance in temperate but not cold air. J Appl Physiol 2005 Nov; 99 (5): 1972-6 14. Craig EN, Cummings EG. Dehydration and muscular work. J Appl Physiol 1966 Mar; 21 (2): 670-4 15. Pichan G, Gauttam RK, Tomar OS, et al. Effect of primary hypohydration on physical work capacity. Int J Biometeorol 1988 Sep; 32 (3): 176-80 16. Nybo L, Jensen T, Nielsen B, et al. Effects of marked hyperthermia with and without dehydration on VO(2) kinetics during intense exercise. J Appl Physiol 2001 Mar; 90 (3): 1057-64 17. Wingo JE, Casa DJ, Berger EM, et al. Influence of a preexercise glycerol hydration beverage on performance and physiologic function during mountain-bike races in the heat. J Athl Train 2004 Jun; 39 (2): 169-75 18. Sawka MN, Burke LM, Eichner ER, et al. American College of Sports Medicine position stand: exercise and fluid replacement. Med Sci Sports Exerc 2007 Feb; 39 (2): 377-90 19. Wagner DR. Hyperhydrating with glycerol: implications for athletic performance. J Am Diet Assoc 1999 Feb; 99 (2): 207-12

ª 2010 Adis Data Information BV. All rights reserved.

137

20. Lyons TP, Riedesel ML, Meuli LE, et al. Effects of glycerolinduced hyperhydration prior to exercise in the heat on sweating and core temperature. Med Sci Sports Exerc 1990 Aug; 22 (4): 477-83 21. Kavouras SA, Armstrong LE, Maresh CM, et al. Rehydration with glycerol: endocrine, cardiovascular and thermoregulatory responses during exercise in the heat. J Appl Physiol 2006; 100 (2): 442-50 22. Figaro MK, Mack GW. Regulation of fluid intake in dehydrated humans: role of oropharyngeal stimulation. Am J Physiol 1997 Jun; 272 (6 Pt 2): R1740-6 23. Robergs RA, Griffin SE. Glycerol: biochemistry, pharmacokinetics and clinical and practical applications. Sports Med 1998 Sep; 26 (3): 145-67 24. Nelson JL, Robergs RA. Exploring the potential ergogenic effects of glycerol hyperhydration. Sports Med 2007; 37 (11): 981-1000 25. Goulet ED, Aubertin-Leheudre M, Plante GE, et al. A metaanalysis of the effects of glycerol-induced hyperhydration on fluid retention and endurance performance. Int J Sport Nutr Exerc Metab 2007 Aug; 17 (4): 391-410 26. Freund BJ, Montain SJ, Young AJ, et al. Glycerol hyperhydration: hormonal, renal, and vascular fluid responses. J Appl Physiol 1995 Dec; 79 (6): 2069-77 27. Melin B, Jimenez C, Koulmann N, et al. Hyperhydration induced by glycerol ingestion: hormonal and renal responses. Can J Physiol Pharmacol 2002 Jun; 80 (6): 526-32 28. Koulmann N, Jimenez C, Regal D, et al. Use of bioelectrical impedance analysis to estimate body fluid compartments after acute variations of the body hydration level. Med Sci Sports Exerc 2000 Apr; 32 (4): 857-64 29. Meyer LG, Horrigan Jr DJ, Lotz WG. Effects of three hydration beverages on exercise performance during 60 hours of heat exposure. Aviat Space Environ Med 1995 Nov; 66 (11): 1052-7 30. Montner P, Stark DM, Riedesel ML, et al. Pre-exercise glycerol hydration improves cycling endurance time. Int J Sports Med 1996 Jan; 17 (1): 27-33 31. Latzka WA, Sawka MN, Montain SJ, et al. Hyperhydration: thermoregulatory effects during compensable exercise-heat stress. J Appl Physiol 1997 Sep; 83 (3): 860-6 32. Latzka WA, Sawka MN, Montain SJ, et al. Hyperhydration: tolerance and cardiovascular effects during uncompensable exercise-heat stress. J Appl Physiol 1998 Jun; 84 (6): 1858-64 33. Hitchins S, Martin DT, Burke L, et al. Glycerol hyperhydration improves cycle time trial performance in hot humid conditions. Eur J Appl Physiol Occup Physiol 1999 Oct; 80 (5): 494-501 34. Montner P, Zou Y, Robergs R, et al. Glycerol hyperhydration alters cardiovascular and renal function. J Exerc Physiol (Online) 1999; 2 (1): 1-10 35. Anderson MJ, Cotter JD, Garnham AP, et al. Effect of glycerol-induced hyperhydration on thermoregulation and metabolism during exercise in heat. Int J Sport Nutr Exerc Metab 2001 Sep; 11 (3): 315-33 36. Coutts A, Reaburn P, Mummery K, et al. The effect of glycerol hyperhydration on Olympic distance triathlon performance in high ambient temperatures. Int J Sport Nutr Exerc Metab 2002 Mar; 12 (1): 105-19

Sports Med 2010; 40 (2)

138

37. Magal M, Webster MJ, Sistrunk LE, et al. Comparison of glycerol and water hydration regimens on tennis-related performance. Med Sci Sports Exerc 2003 Jan; 35 (1): 150-6 38. Marino FE, Kay D, Cannon J. Glycerol hyperhydration fails to improve endurance performance and thermoregulation in humans in a warm humid environment. Pflugers Arch 2003 Jul; 446 (4): 455-62 39. Goulet ED, Robergs RA, Labrecque S, et al. Effect of glycerol-induced hyperhydration on thermoregulatory and cardiovascular functions and endurance performance during prolonged cycling in a 25 degrees C environment. Appl Physiol Nutr Metab 2006 Apr; 31 (2): 101-9 40. Easton C, Turner S, Pitsiladis YP. Creatine and glycerol hyperhydration in trained subjects before exercise in the heat. Int J Sport Nutr Exerc Metab 2007 Feb; 17 (1): 70-91 41. Nishijima T, Tashiro H, Kato M, et al. Alleviation of exercise-induced dehydration under hot conditions by glycerol hyperhydration. Int J Sport Health Sci 2007; 5: 32-41 42. Dini M, Corbianco S, Rossi B, et al. Hyperhydrating with glycerol: effects on thermoregulation, hydration and athletic performance during specific exergonic exercise in a warm-humid environment. Sport Sci Health 2007; 2: 1-7 43. Goulet ED, Rousseau SF, Lamboley CR, et al. Pre-exercise hyperhydration delays dehydration and improves endurance capacity during 2 h of cycling in a temperate climate. J Physiol Anthropol 2008 Sep; 27 (5): 263-71 44. Murray R, Eddy DE, Paul GL, et al. Physiological responses to glycerol ingestion during exercise. J Appl Physiol 1991 Jul; 71 (1): 144-9 45. Siegler JC, Mermier CM, Amorim FT, et al. Hydration, thermoregulation, and performance effects of two sport drinks during soccer training sessions. J Strength Cond Res 2008 Sep; 22 (5): 1394-401 46. Scheett TP, Webster MJ, Wagoner KD. Effectiveness of glycerol as a rehydrating agent. Int J Sport Nutr Exerc Metab 2001 Mar; 11 (1): 63-71 47. Koenigsberg PS, Martin KK, Hlava HR, et al. Sustained hyperhydration with glycerol ingestion. Life Sci 1995; 57 (7): 645-53 48. Jadad AR, Moore RA, Carroll D, et al. Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trial 1996 Feb; 17 (1): 1-12 49. Maher CG, Moseley AM, Sherrington C, et al. A description of the trials, reviews, and practice guidelines indexed in the PEDro database. Phys Ther 2008 Sep; 88 (9): 1068-77 50. Verhagen AP, de Vet HC, de Bie RA, et al. The Delphi list: a criteria list for quality assessment of randomized clinical trials for conducting systematic reviews developed by Delphi consensus. J Clin Epidemiol 1998 Dec; 51 (12): 1235-41 51. Moher D, Schulz KF, Altman DG. The CONSORT statement: revised recommendations for improving the quality of reports of parallel-group randomised trials. Lancet 2001 Apr 14; 357 (9263): 1191-4 52. Moher D, Pham B, Jones A, et al. Does quality of reports of randomised trials affect estimates of intervention efficacy reported in meta-analyses? Lancet 1998 Aug 22; 352 (9128): 609-13

ª 2010 Adis Data Information BV. All rights reserved.

van Rosendal et al.

53. Pengel LH, Barcena L, Morris PJ. The quality of reporting of randomized controlled trials in solid organ transplantation. Transpl Int 2009; 22 (4): 377-84 54. Inder WJ, Swanney MP, Donald RA, et al. The effect of glycerol and desmopressin on exercise performance and hydration in triathletes. Med Sci Sports Exerc 1998 Aug; 30 (8): 1263-9 55. Coombes JS, Hamilton KL. The effectiveness of commercially available sports drinks. Sports Med 2000 Mar; 29 (3): 181-209 56. Von Duvillard SP, Braun WA, Markofski M, et al. Fluids and hydration in prolonged endurance performance. Nutrition 2004 Jul-Aug; 20 (7-8): 651-6 57. Allen LA, Wingertzahn MA, Teichberg S, et al. Proabsorptive effect of glycerol as a glucose substitute in oral rehydration solutions. J Nutr Biochem 1999 Jan; 10 (1): 49-55 58. Wapnir RA, Sia MC, Fisher SE. Enhancement of intestinal water absorption and sodium transport by glycerol in rats. J Appl Physiol 1996 Dec; 81 (6): 2523-7 59. Lamb DR, Lightfoot WS, Myhal M. Prehydration with glycerol does not improve cycling performance vs 6% CHO-electrolyte drink [abstract]. Med Sci Sports Exerc 1997; 29 (5 Suppl.): 249 60. Speedy DB, Faris JG, Hamlin M, et al. Hyponatremia and weight changes in an ultradistance triathlon. Clin J Sport Med 1997 Jul; 7 (3): 180-4 61. Speedy DB, Noakes TD, Rogers IR, et al. Hyponatremia in ultradistance triathletes. Med Sci Sports Exerc 1999 Jun; 31 (6): 809-15 62. Noakes TD, Sharwood K, Speedy D, et al. Three independent biological mechanisms cause exercise-associated hyponatremia: evidence from 2,135 weighed competitive athletic performances. Proc Nat Acad Sci USA 2005 Dec 20; 102 (51): 18550-5 63. Noakes T. Hyponatremia in distance runners: fluid and sodium balance during exercise. Curr Sports Med Rep 2002 Aug; 1 (4): 197-207 64. Irving RA, Noakes TD, Buck R, et al. Evaluation of renal function and fluid homeostasis during recovery from exerciseinduced hyponatremia. J Appl Physiol 1991 Jan; 70 (1): 342-8 65. Murray B, Eichner ER. Hyponatremia of exercise. Curr Sports Med Rep 2004 Jun; 3 (3): 117-8 66. Vrijens DM, Rehrer NJ. Sodium-free fluid ingestion decreases plasma sodium during exercise in the heat. J Appl Physiol 1999 Jun; 86 (6): 1847-51 67. Coyle EF. Fluid and fuel intake during exercise. J Sports Sci 2004 Jan; 22 (1): 39-55 68. Achten J, Jeukendrup AE. Effects of pre-exercise ingestion of carbohydrate on glycaemic and insulinaemic responses during subsequent exercise at differing intensities. Eur J Appl Physiol 2003 Jan; 88 (4-5): 466-71 69. Jentjens RL, Jeukendrup AE. Effects of pre-exercise ingestion of trehalose, galactose and glucose on subsequent metabolism and cycling performance. Eur J Appl Physiol 2003 Jan; 88 (4-5): 459-65 70. Jentjens RL, Cale C, Gutch C, et al. Effects of pre-exercise ingestion of differing amounts of carbohydrate on subsequent metabolism and cycling performance. Eur J Appl Physiol 2003 Jan; 88 (4-5): 444-52

Sports Med 2010; 40 (2)

Guidelines for Glycerol Use

71. Moseley L, Lancaster GI, Jeukendrup AE. Effects of timing of pre-exercise ingestion of carbohydrate on subsequent metabolism and cycling performance. Eur J Appl Physiol 2003 Jan; 88 (4-5): 453-8 72. Hargreaves M, Hawley JA, Jeukendrup A. Pre-exercise carbohydrate and fat ingestion: effects on metabolism and performance. J Sports Sci 2004 Jan; 22 (1): 31-8 73. Noakes TD, Rehrer NJ, Maughan RJ. The importance of volume in regulating gastric emptying. Med Sci Sports Exerc 1991 Mar; 23 (3): 307-13 74. Goulet E, Gauthier P, Labrecque S, et al. Glycerol hyperhydration, endurance performance, and cardiovascular and thermoregulatory responses: a case study of a highly trained triathlete. J Exerc Physiol 2002 May; 5 (2): 19-28 75. Sommer S, Nau R, Wieland E, et al. Pharmacokinetics of glycerol administered orally in healthy volunteers. Arzneimittelforschung 1993 Jul; 43 (7): 744-7 76. Shirreffs SM, Armstrong LE, Cheuvront SN. Fluid and electrolyte needs for preparation and recovery from training and competition. J Sports Sci 2004 Jan; 22 (1): 57-63 77. Massicotte D, Scotto A, Peronnet F, et al. Metabolic fate of a large amount of 13C-glycerol ingested during prolonged exercise. Eur J Appl Physiol 2006; 96: 322-9 78. Burelle Y, Massicotte D, Lussier M, et al. Oxidation of [(13)C]glycerol ingested along with glucose during prolonged exercise. J Appl Physiol 2001 May; 90 (5): 1685-90 79. Noakes TD. Drinking guidelines for exercise: what evidence is there that athletes should drink ‘‘as much as tolerable’’,

ª 2010 Adis Data Information BV. All rights reserved.

139

80.

81.

82.

83.

84.

85.

‘‘to replace the weight lost during exercise’’ or ‘‘ad libitum’’? J Sports Sci 2007 May; 25 (7): 781-96 Shirreffs SM, Taylor AJ, Leiper JB, et al. Post-exercise rehydration in man: effects of volume consumed and drink sodium content. Med Sci Sports Exerc 1996 Oct; 28 (10): 1260-71 Arnall DA, Goforth Jr HW. Failure to reduce body water loss in cold-water immersion by glycerol ingestion. Undersea Hyperb Med 1993 Dec; 20 (4): 309-20 O’Brien C, Freund BJ, Young AJ, et al. Glycerol hyperhydration: physiological responses during cold-air exposure. J Appl Physiol 2005 Aug; 99 (2): 515-21 Frank MS, Nahata MC, Hilty MD. Glycerol: a review of its pharmacology, pharmacokinetics, adverse reactions, and clinical use. Pharmacotherapy 1981 Sep-Oct; 1 (2): 147-60 McCurdy DK, Schneider B, Scheie HG. Oral glycerol: the mechanism of intraocular hypotension. Am J Ophthalmol 1966 May; 61 (5): 1244-9 Tourtellotte WW, Reinglass JL, Newkirk TA. Cerebral dehydration action of glycerol: I, historical aspects with emphasis on the toxicity and intravenous administration. Clin Pharmacol Ther 1972 Mar-Apr; 13 (2): 159-71

Correspondence: Dr Simon Piet van Rosendal, School of Human Movement Studies, University of Queensland, St Lucia, QLD, 4072, Australia. E-mail: [email protected]

Sports Med 2010; 40 (2)

REVIEW ARTICLE

Sports Med 2010; 40 (2): 141-161 0112-1642/10/0002-0141/$49.95/0

ª 2010 Adis Data Information BV. All rights reserved.

Evaluation of Instruments for Measuring the Burden of Sport and Active Recreation Injury Nadine E. Andrew,1 Belinda J. Gabbe,1,2 Rory Wolfe1 and Peter A. Cameron1,2 1 Department of Epidemiology and Preventive Medicine, Monash University, Melbourne, Victoria, Australia 2 National Trauma Research Institute, The Alfred Hospital, Melbourne, Victoria, Australia

Contents Abstract. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. Assessing the Suitability of an Outcome Assessment Instrument . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1 Population Demographics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2 Stakeholders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.3 Generic versus Specific . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.4 Reliability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.5 Validity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.6 Responsiveness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.7 Accessibility and Acceptability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. Search Strategy and Selection Criteria for Reviewed Outcome Measures . . . . . . . . . . . . . . . . . . . . . . 3. Health Status and Health-Related Quality-of-Life Measures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1 The Quality of Well-Being . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 The EuroQol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3 The Assessment of Quality of Life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4 Short Form-36. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5 Short Form-12. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.6 Sickness Impact Profile-136 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.7 Sickness Impact Profile-68 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4. Functional Outcome Measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1 The Functional Independence Measure and the Functional Assessment Measure . . . . . . . . . . . 4.2 The Functional Capacity Index (FCI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3 Glasgow Outcome Scale and Glasgow Outcome Scale-Extended . . . . . . . . . . . . . . . . . . . . . . . 4.4 Musculoskeletal Functional Assessment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.5 The Short Musculoskeletal Functional Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5. Physical Activity Measures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1 The Short International Physical Activity Questionnaire. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2 Paffenbarger Physical Activity Questionnaire . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3 Godin Leisure-Time Exercise Questionnaire. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Abstract

141 143 143 143 143 144 144 144 144 144 146 146 146 146 150 150 151 151 151 151 152 152 153 153 154 154 154 155 155

Sport and active recreation injuries are common. Participants are generally young, healthy and physically active individuals and as a result their injuries can have long-ranging effects for both the individuals and society.

Andrew et al.

142

Accurate and appropriate measurement of the outcomes of sport and active recreation injuries is essential for understanding the time frame and quality of recovery, and quantifying the burden of these injuries. The WHO has developed a framework that can be used for studying health-related outcomes called the International Classification of Function (ICF). As such, the ICF is a useful tool for assessing the suitability of outcome measures for general sport and active recreation populations. This article provides a review of outcome measures that are potentially suitable for use in a general sport and active recreation injury population, assessed within the framework of the ICF. An extensive literature search was performed to identify instruments used in sport and active recreation (and general) injury populations that would be suitable for measuring the outcomes and burden of sport and recreation injuries and return to physical activity. The search identified six health status and health-related quality-of-life (HR-QOL) measures and five functional outcome measures. Of the outcome measures reviewed, the Short Form-36 was the most commonly used and covered many of the areas relevant to a sport and active recreation population. The comprehensiveness of the Sickness Impact Profile36 meant that it contained many relevant items; however, its usefulness is limited by its high level of responder burden. The Musculoskeletal Functional Assessment provided a detailed measure of function, appropriate to a sport and active recreation population, and the Glasgow Outcome Scale-Extended can provide a suitable global measure of function. The Short International Physical Activity Questionnaire is a potential means of measuring return to physical activity for this group. There are no outcome measures specifically designed to measure outcomes in a general sport and active recreation population. There are, however, existing measures that when used in combination have the potential to provide a comprehensive assessment of injury outcomes in this group. Future research should focus on validating existing measures suitable for a sport and active recreation population as well as developing an ICF sport and active recreation core set of items. An ICF core set would assist researchers and clinicians in selecting the combination of outcome measures most appropriate to their needs as well forming the basis for the development of a specific sport and active recreation outcome measure.

Sport and active recreation injuries are common,[1-3] and have the potential to result in longterm physical and mental health consequences ranging from an inability to return to pre-injury levels of sporting activity to severe disability requiring long-term treatment and care.[1,2] Nevertheless, few studies have assessed injury outcomes in an adult sport and active recreation population and how best to measure outcomes in this group remains unclear. Consensus on this issue is necesª 2010 Adis Data Information BV. All rights reserved.

sary to progress our understanding of the quality of recovery following sport and recreation-related injury, information critical for establishing the burden of sport and active recreation-related injury and informing injury prevention and safety promotion initiatives. Sport and active recreation injuries can range from isolated ligament injuries to multiple traumas, with varying consequences. The WHO defines disability as an umbrella term that includes Sports Med 2010; 40 (2)

Measuring the Burden of Sports Injury

impairment, activity limitation and participation restrictions, and acknowledges that disability involves a complex interaction between features of a person’s body and their environment.[3] The WHO has developed the International Classification of Function (ICF), which provides a unified, scientific framework that can be used for studying health and health-related outcomes. The ICF covers the key domains of body functions, body structures, activity limitation and participation restrictions and the environment.[4] Each domain is important for an injured participants’ successful return to sport and active recreation. Outcome measures that cover the key domains of the ICF should provide the level of comprehensiveness needed to measure outcomes in the diverse sport and active recreation population. The WHO is currently developing core sets of ICF items, developed through systematic review processes and extensive consultation with health professionals and patients, for various patient groups. A core set is yet to be developed for a sport and active recreation population. The ICF level two classifications that appear to be most relevant to sport and active recreation are those relating to mental functions, pain, neuromusculoskeletal and movement-related functions, muscle functions, mobility, community, social and civic life, and environmental attitudes. This article provides an overview of commonly used and potentially useful measures of injury outcome with a specific focus on their validity, reliability, utility and usefulness for measuring outcomes of sport and active recreation injuries in adults, with reference to the ICF. 1. Assessing the Suitability of an Outcome Assessment Instrument 1.1 Population Demographics

Adults who participate in sport and active recreation differ from the general population. They tend to be young, physically active and healthy, with a high level of employment.[1,2,5] They also have higher levels of physical function, psychological function and perceived health than non-participants.[6-8] Hence, the morbidity conseª 2010 Adis Data Information BV. All rights reserved.

143

quences of their injuries can be high. Furthermore, sport and active recreation injuries that result in long-term reductions of physical activity levels in this already active group will potentially increase their risk of developing chronic diseases later in life.[9-12] Not only is it important to measure longterm outcomes for this group, but the measures used need to capture the full range of consequences, including changes in physical activity levels. 1.2 Stakeholders

Sport and active recreation involves a number of stakeholders, for whom the use of standardized outcome measures is important. These include patients, clinicians, researchers, coaches, trainers, sporting clubs, peak sporting bodies, health policy makers, funding bodies, health insurers and sports insurers. The selection of an outcome measure will depend on the purposes for which the outcome data will be used. Measures that are easy to understand, administer, score and interpret are more likely to aid in bridging the interface between the health sector and the sports sector. 1.3 Generic versus Specific

Outcome measures can be divided into generic and specific measures. Generic instruments allow broad comparisons to be made across different populations and often address multiple aspects of health.[13,14] They can be divided into health status and health-related quality-of-life measures (HRQOL) and functional measures. The former focus on multiple dimensions of health such as physical, mental, social and functional health,[15] whereas functional measures focus on aspects of health related to the individual’s ability to perform selected activities. Specific instruments focus on health domains that are important to particular diseases or populations. They are more responsive than generic instruments,[13,14] but do not allow comparison across disease or injury groups. For a sport and active recreation population these comparisons are important for establishing the burden of injury, prioritizing research and funding, and guiding policy development,[16] especially given Sports Med 2010; 40 (2)

Andrew et al.

144

the low priority sport and recreation injuries currently receive and the paucity of information available on their outcomes.[17] 1.4 Reliability

Outcome measures are used to measure changes over time due to treatment or natural history[18] and therefore must fulfill the key psychometric properties of reliability, validity and responsiveness. Reliability is the ability of a measure to produce the same result when reapplied under conditions in which a change would not be expected.[13,18] High reliability means that we can be confident that any changes are a true reflection of change and not due to error. For measures that require administration by an interviewer, reliability between examiners or interrater reliability is also required. An intra-class correlation coefficient (ICC) is used to measure agreement for continuous measures, the Kappa statistic or weighted Kappa (Kw) is used for categorical measures and a Spearman correlation (p) can be used for ordinal variables. The lower limit of the 95% confidence interval (CI) should be ‡0.75.[18] 1.5 Validity

Key aspects of validity are content and construct validity. Good content validity means that the measure contains items that are relevant to the demographics and desired outcomes of a sport and active recreation population, will have minimal ceiling and floor effects, and does not restrict the range of measurement so as to capture the wide range of outcomes experienced by this group.[13] Construct validity enables quantification of the logical relationship between the instrument and other measurable characteristics of the patient group or other outcome measures.[13,18] For example, you would expect a good correlation or convergent validity between measures that are measuring similar aspects of disability but a low correlation or divergent validity between measures that are measuring different aspects of disability. Good agreement is generally accepted if 0.40£ r ‡0.60.[19-21] ª 2010 Adis Data Information BV. All rights reserved.

1.6 Responsiveness

Responsiveness is the ability of the instrument to detect change over time and is established by correlating change scores with changes in other related measures or variables. Instruments that are responsive at the high end of function will be most appropriate as it is important to detect the subtle improvements that are important to sports participation, particularly in the later stages of recovery.[19] Responsiveness is usually measured using a standard response mean (SRM). An SRM of 0.5–0.7 represents a moderate effect and ‡0.8 a large effect.[19] 1.7 Accessibility and Acceptability

Measures need to be acceptable to both patients and researchers/clinicians. Completion time and the degree of difficulty of an instrument can impact on the completeness of the data obtained.[16,22] Completion time is important in sport and active recreation populations as younger, adult populations often have work and family commitments, which means that they have less free time than older, retired individuals.[16] Self-administration by mail or web delivery offers advantages in terms of costs and resources but often at the expense of quality and completeness.[13] Web delivery can provide a more reliable contact point than mail; however, it is limited to participants with web access. Those administered in person by trained interviewers can be more accurate and complete but are resource intensive.[13] Administration by phone interview can be an acceptable compromise.[13] Licensing costs associated with some outcome measures can also impact on their uptake. 2. Search Strategy and Selection Criteria for Reviewed Outcome Measures Searches of Ovid, MEDLINE, PubMed and Cumulative Index to Nursing and Allied Health Literature (CINAHL) databases were performed using the key words ‘sport’, ‘recreation’, ‘athletic’, ‘athlete’, ‘injury’, ‘disability’, ‘impairment’, ‘quality of life’, ‘function’ and ‘outcomes’. Searches were limited to English-language citations, from January 1950 to December 2007. A manual search was Sports Med 2010; 40 (2)

Measuring the Burden of Sports Injury

145

conducted of three sports medicine journals and three general injury journals that are known to publish papers using outcome measures in sport and recreation injuries from 2000 to 2007 inclusive. These were the British Journal of Sports Medicine, American Journal of Sports Medicine, Clinical Journal of Sports Medicine, Injury Prevention, International Journal of Injury Control, and Safety Promotion and Injury. The reference lists of key articles from the searches were checked for additional relevant articles. The searches revealed 133 articles that reported outcomes in adult patients injured as a result of

participation in sport or active recreation. Studies included a range of patients from elite athletes to active recreation participants. Fifty studies used disease-specific or joint-specific outcome measures.[23-73] As these studies focused on specific disorders such as anterior cruciate ligament injuries, the measures used were not appropriate for a general sport and active recreation population. Fifteen studies included a validated health status or functional outcome instrument[1,2,5,7,74-86] (table I). As our search strategy identified a small number of outcome measures for review, an additional MEDLINE search was performed to identify

Table I. Studies containing sport and active recreation populations that used generic outcome measures Outcome measure

Study

Participants

Study type

Injury types/diagnosis

Mean follow-up period

Peterson et al.[74]

Recreational athletes

Randomized control trial

Achilles tendinopathy

6, 12, 54 wk

Guskiewicz et al.[75]

Retired professional footballers

Retrospective cohort

Concussion

NA

Naal et al.[76]

Recreational athletes

Survey

Unicompartmental knee arthroscopy

18 mo

Anandacoomarasamy and Barnsley[77]

Elite and recreational athletes

Case control

Inversion ankle injuries

29 mo

Mazzocca et al.[78]

Collegiate and high school athletes

Case series

Anterior shoulder instability

37 mo

Debnath et al.[79]

Elite and recreational athletes

Case series

Spondylolysis

2y

Williams et al.[80]

Competitive and recreational athletes

Retrospective case series

Traumatic posterior shoulder instability

5.1 y

Finch et al.[5]

Competitive and recreational athletes

Prospective cohort

General outpatient injuries

6 wk

Wang et al.[7]

Elite athletes

Case control

Lumbar discectomy

3.1 y

Von Porat et al.[81]

Soccer players

Retrospective cohort

Anterior cruciate ligament injuries

14 y

Nicholas et al.[82]

Retired American football players

Cross-sectional study

General injury

NA

SF-12

Meller et al.[83]

Competitive and recreational athletes

Retrospective cohort

Recurrent anterior shoulder instability

24 mo

SIP-68

Dekker et al.[1]

Competitive and recreational athletes

Retrospective cohort

General inpatient injuries

1–4 y

Dekker et al.[2]

Competitive and recreational athletes

Retrospective cohort

General outpatient injuries

2–5 y

EuroQol

Turner et al.[84]

Professional footballers

Retrospective cohort

General injuries

4.7 y

GOSE

Lindsay et al.[85]

Competitive and recreational athletes

Retrospective cohort

Head injuries

6 mo

SMFA

Giza et al.[86]

Recreational footballers

Case series

Hip fracture-dislocations

12 mo

SF-36

GOSE = Glasgow Outcome Scale Extended; NA = not available; SF-12 = 12-item Short Form Health Survey; SF-36 = Short Form-36 Health Survey; SIP-68 = Sickness Impact Profile-68; SMFA = Short Musculoskeletal Functional Assessment.

ª 2010 Adis Data Information BV. All rights reserved.

Sports Med 2010; 40 (2)

Andrew et al.

146

common outcome measures used in general trauma populations, resulting in 24 studies and seven additional outcome measures. For many sports participants, an important measure of recovery is return to pre-injury physical activity levels, highlighting the importance of valid measurement of this aspect of recovery. A few studies have used a graded system or a point of reference to assess changes in sporting or activity levels,[87-89] but none were validated measures. Physical activity questionnaires developed for use in general populations could be suitable for measuring return to physical activity following injury. Many of these questionnaires have undergone psychometric testing and have the advantage of capturing sport-related outcomes in situations where, for example, a person injured whilst cycling may be able to return to cycling but no longer be able to participate in their pre-injury level of football. Therefore, physical activity questionnaires that measure aspects of physical activity appropriate for a sport and active recreation population such as type, duration, frequency and intensity of activities within a suitable time frame were also reviewed. Outcome measures identified from the above searches were reviewed for their use in general sport and active recreation populations. 3. Health Status and Health-Related Quality-of-Life Measures The HR-QOL and health status measures identified in sport and recreation studies were the Short Form-36 Health Survey (SF-36),[5,7,74-82] the 12-item Short Form Health Survey (SF-12),[83] the Sickness Impact Profile-68 (SIP-68)[1,2] and the EuroQol (EQ-5D).[84] Other relevant measures used in general injury populations were the Sickness Impact Profile (SIP),[21,90-97] the Quality of Well-Being (QWB),[98,99] and the Assessment of Quality of Life (AQoL).[100,101]

(table II). The QWB can be used as a health utility measure in which quality-adjusted life-years (QALYs) can be calculated,[121,122] and is best suited to policy analysis and economic studies.[121] QALYs are the number of years lost from one’s life expectancy as a result of dysfunction[123] and is an important concept when measuring the burden of injury associated with sport and active recreation injuries. The symptombased approach of the QWB means the primary focus is on the ICF domains of body functions and structures, and to a lesser extent on activities and participation. The highest level of physical activity measured by the QWB refers to walking and stairs, reducing its usefulness for an active sport and recreation population. 3.2 The EuroQol

The EQ-5D has a measure of HR-QOL over five domains (table II) and a global health measure in the form of a visual analogue scale (VAS). It was designed for population health surveys and can be used to calculate QALYs.[124] The most desirable features of the EQ-5D are its brevity and simplicity and its ability to be used in a wide variety of conditions.[124] The first three domains are related to activities and participation. Two of these do not address items specific to a sport and active recreation population, although the section on usual activities allows inclusion of sport and recreation activities. The other domains in the EQ-5D relate to body functions pertaining to mental health and pain. The EQ-5D was shown to have ceiling effects in a general population[125] reflecting its focus on low levels of function. The open-ended nature of the VAS in section two, which allows respondents to rate their health from a worst to best imaginable health state, may provide a global measure of HR-QOL appropriate for use in a sport and active recreation population. 3.3 The Assessment of Quality of Life

3.1 The Quality of Well-Being

The QWB assesses over 25 symptoms and records functional limitations within the domains of mobility, physical activity and social activity ª 2010 Adis Data Information BV. All rights reserved.

The AQoL contains 15 items over five domains (table II), each with four responses, increasing its sensitivity over the QWB and the EQ-5D.[126] Most of the items in the AQoL relate to ICF body Sports Med 2010; 40 (2)

No. of items

Domains

Time to complete (min)

Injury populations where it has been validated

Advantages and disadvantages

QWB

14 plus assessment of >25 symptoms

Mobility Physical activity Social activity symptoms

10–30

Nil

Can be converted to QALYs Useful for policy development Only measures function or symptom-based problems Does not address high levels of function

EuroQol

6 plus a visual analogue scale

Mobility Self-care Usual activity Pain/discomfort Anxiety/depression

1

Nil

Brief and simple to administer Has wide applicability Can be used to calculate QALYs Large ceiling effects likely in a sport and active recreation population

AQoL

15

Illness Independent living Social relationships Physical senses Psychological well-being

5

General injury population[101]

Increased sensitivity over the EuroQol and QWB Can be used to calculate QALYs Likely to have large ceiling effects Only measures low levels of function

SF-36

36

Physical function Role-physical Bodily pain Social functioning General health Vitality Role-emotional Mental health

5–10

Athletic injuries[6,75,102,103] Traumatic brain injury[104,105] Multi trauma + head injury[106] Orthopaedic injury[107,108]

Only generic measure with some validation in a sport and recreation population Does not measure change in sport and recreation Poor responsiveness for mental health subscale

SF-12

12

Physical function Role-physical Bodily pain Social functioning General health Vitality Role-emotional Mental health

2

General trauma[109]

Brief and simple to administer Omits items most relevant to a sport and active recreation population

SIP-136

136

Sleep/rest Emotional behaviour Body care and movement Household management Mobility

20–30

Rehabilitation patients[110] General trauma[20,21] Lower extremity trauma[111]

Comprehensive measure Has questions specific to active recreation Ceiling effects demonstrated Long completion time

Continued next page

147

Sports Med 2010; 40 (2)

Outcome instrument

Measuring the Burden of Sports Injury

ª 2010 Adis Data Information BV. All rights reserved.

Table II. Summary of outcome measures

Outcome instrument

148

ª 2010 Adis Data Information BV. All rights reserved.

Table II. Contd No. of items

Domains

Time to complete (min)

Injury populations where it has been validated

Advantages and disadvantages

Social interaction Ambulation Alertness/behaviour Communication Recreation and pastimes Eating Work 68

Somatic autonomy Mobility control Communication and psychological autonomy Social interaction Emotional behaviour Mobility range

10–15

Head injury[105]

Reduced responder burden compared to the SIP-136 Omits many of the questions considered to be most relevant to a sport and active recreation population

FIM and FAM

18 (FIM) +12 additional (FAM)

Motor Cognitive Behavioural Communication Community functioning

FIM + FAM = 35

General trauma[112] Head injury[113,114]

Designed for use in inpatient rehabilitation programmes Only measures low levels of function Ceiling effects likely to be a problem

FCI

10

Excretory function Eating Sexual function Ambulation Hand/arm movement Bending/lifting Speech Auditory function Visual function

8

General trauma[21] Lower limb trauma[115]

Designed specifically for trauma patients Does not contain items specifically relevant to a sport and active recreation population Focuses on low levels of function

GOS and GOSE

7

Level of consciousness Independence in the home Independence outside the home Work Social and leisure activities Family and friends Return to normal life

<10

Head injury[116,117]

Simple to administer Global measure of function only Low sensitivity

Continued next page

Andrew et al.

Sports Med 2010; 40 (2)

SIP-68

No. of items

Domains

Time to complete (min)

Injury populations where it has been validated

Advantages and disadvantages

MFA

101

Activities using arms or legs Activities using hands Work around home Self care Sleep and rest Leisure and recreation Relationships Thinking Life changes and feelings Work

15

Orthopaedic trauma[20,118]

Good measure of injury impact on active recreation Designed for groups relevant to a sport and recreation population May not be appropriate for non-extremity injuries Potential ceiling effects

SMFA

46

Daily activities Emotional status Arm and hand function Mobility category Bother index

5–10

Orthopaedic trauma[119]

Reduced responder burden compared to MFA Developed in clinically relevant group Omits many items relevant to a sport and active recreation population

Short IPAQ

7

Work Domestic and gardening Leisure time Transport

5

Lower limb arthritis[120]

Measures physical activity across multiple domains High uptake for population surveillance in multiple countries Limited use in outcome studies Yet to be used in injury populations

PPAQ

3

Sport and leisure Stairs Blocks walked

Variable

Nil

Allows for detailed recording of sport and recreation activities Does not cover domains outside sport and recreation Time period over which sport and recreation is recorded would need to be modified

GLETQ

4

Leisure time

Variable

Nil

Easy and brief to administer Reduced accuracy as it only counts episodes of exercise of >15 min duration Does not cover domains outside leisure time

AQoL = Assessment of Quality of Life; FAM = Functional Assessment Measure; FCI = Functional Capacity Index; FIM = Functional Independence Measure; GLETQ = Godin LeisureTime Exercise Questionnaire; GOS = Glasgow Outcome Scale; GOSE = Glasgow Outcome Scale Extended; IPAQ = Short International Physical Activity Questionnaire; MFA = Musculoskeletal Function Assessment; PPAQ = Paffenbarger Physical Activity Questionnaire; QALYs = quality-adjusted life-years; QWB = Quality of Well-Being; SF-12 = 12-item Short Form Health Survey; SF-36 = Short Form-36 Health Survey; SIP-68 = the Sickness Impact Profile-68; SIP-136 = Sickness Impact Profile-136; SMFA = Short Musculoskeletal Function Assessment.

149

Sports Med 2010; 40 (2)

Outcome instrument

Measuring the Burden of Sports Injury

ª 2010 Adis Data Information BV. All rights reserved.

Table II. Contd

Andrew et al.

150

functions such as senses, sleep, emotions and pain. Environmental factors relating to medications, devices and treatment are addressed, as are activities and participation, but with a focus on low levels of functioning. The AQoL can also be used to calculate QALYs using the items contained in the last four domains (table II). The AQoL has been validated in an injury population.[101] However, its exclusion of activity and participation items and levels of function most relevant to a sport and active recreation population are likely to result in underestimation of the impact of injury in this group. This is reflected in a study in which 45% of a normal population reported scores in the highest decile.[127] A later version of the AQoL (AQoL Mark 2) has been extended to 20 items. High level mobility functions such as running are included and vitality is assessed. Community roles are addressed with sporting groups included in the examples. Though this version is better suited to a sport and active recreation or injury population, it has yet to be validated or used in these groups. 3.4 Short Form-36

The SF-36 has been widely used in sport and active recreation populations.[5,7,74-80] It contains 36 items over eight domains (table II) and provides a separate score for each subscale and mental component summary (MCS) and physical component summary (PCS) scores. Many of the items in the SF-36 are applicable to a sport and active recreation population, especially in the ICF domains of activities and participation relating to mobility, recreation and leisure and mental functions. The lack of cognitive and upper limb subscales could result in underestimation of the impact of injury in these areas. There is limited assessment of the psychometric properties of the SF-36 in sport and active recreation populations. A study of elite athletes found that serious injury was a predictor of lower PCS, MCS and subscale scores, and that mild injury was a predictor of lower PCS scores.[6] Another study on retired professional footballers found that those with clinical depression had lower MCS and PCS scores compared with those without depression.[75] ª 2010 Adis Data Information BV. All rights reserved.

These results, however, cannot readily be extrapolated to recreational athletes. Criterion validity was assessed in knee-injured sporting populations.[102,103] Good correlations (r = 0.57–0.72) were found between physical function measured by the SF-36 and knee function tests, with divergent validity demonstrated between the other subscales of the SF-36 and the knee function tests.[102,103] Ceiling effects were noted in the SF36 role physical subscale.[102] Good criterion validity was also demonstrated in general injury and traumatic brain injury (TBI) patients.[21,104,128] The reliability of the SF-36 demonstrated high variability between subscales (ICC = 0.04–0.77) in a sport and active recreation population with patella dislocation;[102] however, there was a median interval of 21 days between tests and factors particular to patella dislocation such as fluctuations of symptoms could have affected the results, as could the inclusion of children.[102] In TBI patients, good reliability was demonstrated across all subscales of the SF-36 for one study[104] and in less than half the subscales in another.[105] In sport and active recreation populations, improvements in the PCS and MCS have been demonstrated 2 years after surgery,[79] and changes in physical function, role function, bodily pain and social function subscales have been demonstrated as early as 5–6 weeks after injury.[5] High SRMs (0.5–1.1) were reported for the physical subscales but not the mental health subscales[128] in general injury patients. It is likely that the physical components of the SF-36 are more responsive than the mental components in this group. The SF-36 is suitable for use in a sport and active recreation population. Its main limitations are the potential for ceiling effects in some subscales and the lack of responsiveness of the mental health subscales. The suitability of the SF-36 for upper limb injuries requires further investigation, and further psychometric testing is required in sport and active recreation populations.[129] 3.5 Short Form-12

The SF-12 was derived by selecting 12 items that provided a >90% correlation with the SF-36 and covered each of the eight subscales. Selection Sports Med 2010; 40 (2)

Measuring the Burden of Sports Injury

was based on predictions from US population data and validated in chronic medical patients.[130] Though the SF-12 appears to be a valid alternative to the SF-36 in general, trauma and medical populations,[109,130] its omission of items relevant to sport and active recreation populations such as vigorous activities and walking long distances are likely to underestimate the impact of injury in this group. Therefore, the SF-36 would be preferable to the SF-12 in sport and active recreation populations. 3.6 Sickness Impact Profile-136

The SIP-136 has not been used in specific sport or active recreation studies; however, it has been used in studies of injuries commonly seen in sporting populations.[131,132] The SIP-136 assesses sickness-related behaviours and is designed to have broad applicability across a variety of illnesses and demographics. The SIP-136 contains 136 questions over 12 domains giving it increased content validity over shorter measures,[133] and contains a large number of concepts that can be linked to the ICF.[134] Those most relevant to a sport and active recreation population are related to energy, psychomotor function, exercise tolerance, muscle function and physical recreation.[135,136] The SIP136 can be scored to give physical, psychosocial well-being and individual category scores (table II). The psychometric properties of the SIP-136 have been established in a range of patients, including rehabilitation patients, but have not been assessed in a sport and active recreation population. One study, however, found that certain aspects of the SIP-136 had a much higher degree of relevance to sporting populations, especially those aspects relating to pain and recreational activities.[132] In trauma patients, good convergent validity (r > 0.60) was demonstrated between the PCS of the SIP-136 and the Functional Capacity Index (FCI),[21] and a moderate correlation (r = 0.41) was demonstrated between the SIP-136 and clinical measures of physical impairment.[111] A ceiling but not a floor effect was also demonstrated in trauma patients.[20] The SIP-136 was able to discriminate between treatment groups in conservatively managed ª 2010 Adis Data Information BV. All rights reserved.

151

ankle sprains[131] and was responsive to changes that occurred after anterior cruciate ligament reconstruction at 3 and 12 months.[132] The SIP-136 was also sensitive to improvements in function in trauma patients with lower limb fractures over 6, 12 and 30 months.[91,111] The comprehensiveness of the SIP-136 and its inclusion of relevant subscales make it suitable for a sport and active recreation population. The main disadvantage of the SIP-136 is its long completion time, and further psychometric testing is required in a sport and active recreation population. 3.7 Sickness Impact Profile-68

The SIP-68 has been used in two studies involving general sport and active recreation populations[1,2] and has reliability, validity and responsiveness similar to that of the SIP-136 in rheumatology and medical patients.[137-139] The main disadvantage of the SIP-68 is that it omits many of the questions that are most relevant to a sport and active recreation population.[91,111,132] Fourteen of the 24 questions in the SIP-136 considered to be most relevant to a sport and active recreation population are omitted,[132] limiting the potential usefulness of this instrument in a sport and active recreation population. 4. Functional Outcome Measures Functional measures identified in our search and used in sport and active recreation populations were the Glasgow Outcome Scale (GOS)[85] and the Short Musculoskeletal Functional Assessment (SMFA).[86] Other suitable measures used in general injury studies include the Glasgow Outcome Scale Extended (GOSE),[140] Musculoskeletal Functional Assessment (MFA),[118,141-143] Functional Independence Measure (FIM) and Functional Assessment Measure (FAM),[112,113,144] and the FCI.[21,145,146] 4.1 The Functional Independence Measure and the Functional Assessment Measure

The FIM is an 18-item scale designed to measure change over the course of inpatient rehabilitation programmes and has motor and cognitive Sports Med 2010; 40 (2)

Andrew et al.

152

components. The FAM consists of 12 additional items and was designed for use with the FIM. The questions focus on performance of activities relating to self-care and independence and do not contain items relevant to high levels of function. This is reflected in follow-up studies involving general trauma and TBI patients in which ceiling effects were reported for between 80% and 95% of subjects.[100,112-114] Therefore, the FIM and FAM are unlikely to measure outcomes meaningful to a sport and active recreation population. 4.2 The Functional Capacity Index (FCI)

The FCI was initially designed to predict 12-month outcomes for the injury descriptions contained in the abbreviated injury scale (AIS). The AIS is primarily a threat-to-life scale and does not accurately identify injuries that have high morbidity. The FCI was developed to rectify this.[147] The FCI questionnaire was later developed as an evaluative tool, utilizing these predictive weights in its scoring system.[21] The FCI has the advantage of being specifically designed for use in injury populations and has been validated in this group,[21] though further psychometric evaluation is required.[148] The FCI covers ten dimensions each with between three and seven categories of capacity that are weighted depending on their impact on everyday living[147] (table II). The FCI focuses on body functions with some dimensions such as ambulation and hand/arm movement also containing sub-categories relating to activities and participation and environmental factors. The FCI focuses on tasks necessary for everyday living and does not cover areas such as sport and active recreation.[21] As such, it is not suited to a sport and active recreation population. 4.3 Glasgow Outcome Scale and Glasgow Outcome Scale-Extended

The GOS was developed as a global measure of outcome following head injury, has been recommended for use in general injury studies,[149] and has been used in one study involving sport and active recreation participants.[85] The GOS is designed to reflect disability as defined by the ª 2010 Adis Data Information BV. All rights reserved.

WHO[150] and covers multiple aspects of the ICF relating mainly to activities and participation, across five domains.[116] The GOS is scored by allocating the patient to one of five broad categories ranging from dead to ‘‘resumption of normal activity despite minor deficits’’. The GOS contains a section on social and leisure activities but measures quantity, rather than quality, of participation.[150] This and the allowance of minor deficits in its highest category means that the impact of injury in a sport and active recreation population could be underestimated. Importantly, pre-injury status is considered when scoring the GOS. An extended version of the GOS was developed to increase sensitivity and reduce ceiling effects.[151,152] The upper three categories of good recovery, moderate disability and severe disability are subdivided to provide eight categories (table II). The GOSE has greater suitability over the GOS for sport and active recreation populations due to additional categories and allowing qualification of whether or not a patient has returned to ‘normal life’. Nevertheless, the GOSE has yet to be used in a sport and active recreation population. The psychometric properties of the GOSE have only been assessed in head injured populations. Its content validity is evidenced by good correlations with other functional measures (r = 0.46–0.89)[116,117,152] and the Beck depression inventory (r = 0.64) for all patients except the most severely disabled.[116] Modest associations were also demonstrated with various cognitive tests.[116,153] Administration by an interviewer using a structured interview is recommended for the GOSE to increase reliability.[150] Good intrarater reliability was found for face-to-face versus telephone interviews (Kw = 0.92).[154] Inter-rater reliability results were more variable. One study found low reliability (Kw = 0.56–0.57) both at discharge and 12 months after injury,[105] whereas other studies found acceptable levels (Kw = 0.84–0.98) through various modes of administration;[117,150,154] however, some studies were limited by small patient numbers.[154,155] Good reliability (Kw = 0.92) was also demonstrated for Sports Med 2010; 40 (2)

Measuring the Burden of Sports Injury

postal interviews.[155] Agreement on the GOSE was slightly better for those with severe injuries compared with those with minor injuries,[154] which could reduce its reliability in a sport and active recreation population. The GOSE was able to demonstrate change in a cohort of head-injured patients at 5–7 years after injury as compared with 12 months after injury.[117] Change was also demonstrated in a sample of TBI and general trauma patients at 3 and 6 months after injury,[151] and predicted increases in scores were also demonstrated over a 12-month period.[105] The categorical nature of the GOSE may reduce its sensitivity compared with continuous measures; however, this is yet to be established. Though predominantly used in TBI populations, the use of the GOSE in general trauma populations and its inclusion of relevant items suggests that it may be a suitable global measure of function for sport and active recreation populations. Further psychometric evaluation of the GOSE is required in this population, especially for minor injuries. Further evidence of reliability of the GOSE is needed. 4.4 Musculoskeletal Functional Assessment

The MFA is a self-reported measure developed to assess musculoskeletal disorders of the extremities, including fractures and soft tissue injuries, making the MFA particularly relevant to sport and active recreation populations.[1,2,156] The MFA includes 101 items over ten categories. The MFA contains many ICF sub-categories relevant to a sport and active recreation population. Activities and participation, and to a lesser extent body functions and body structures, are covered and include items such as running, changes in physical recreation activities and changes in physical fitness due to disability[157] (table II). Scoring allows for a total score as well as category sub-scores.[20] Despite use in injury studies,[118,141-143] the MFA has not been used in sport and active recreation injury studies. Validity of the MFA has been established in trauma patients.[20,118] Good correlations have been demonstrated between physician ratings of extremity function and MFA extremity function ª 2010 Adis Data Information BV. All rights reserved.

153

(r = 0.40–0.66), but not between other subscores.[20,118] Convergent validity was demonstrated between various clinical measures and the relevant lower extremity and upper extremity MFA items, and between self-ratings of health and changes in activity.[20,118] Construct validity relating to injury and demographic characteristics and predicted MFA scores were also demonstrated.[118] The total score of the MFA does not have floor or ceiling effects,[20,157] though ceiling effects were noted within individual categories.[20] The MFA has demonstrated good reliability (ICC = 0.70–0.92) for self-administration and inter-rater reliability, with the MFA more reliable in injury than in arthritis groups.[20,157] Good responsiveness (SRM = 0.74) has been demonstrated over a 6-month period for the overall MFA score, but was variable between categories with the categories of family relationships and mobility showing the lowest levels of responsiveness.[20] The MFA could be an appropriate outcome measure for musculoskeletal sport and active recreation injuries. The ability of the MFA to accurately assess function in non-musculoskeletal sport and active recreation injuries is unknown; however, the inclusion of a cognitive component and general function questions suggest that it is likely to be acceptable for broader injury groups. Psychometric analysis of the MFA in a sport and active recreation population is needed. 4.5 The Short Musculoskeletal Functional Assessment

The SMFA was developed to reduce respondent burden for the MFA whilst maintaining important items. It has been used in one sport and active recreation study.[86] Though the SMFA has been shown to be reliable, valid and responsive in patients with extremity disorders,[119] the questionnaire does not include many of the MFA items most relevant to a sport and active recreation context such as those related to running and the category relating to leisure and recreational activities. As such, the SMFA is likely to be a less appropriate measure Sports Med 2010; 40 (2)

Andrew et al.

154

of function than the MFA in a sport and active recreation context.

5. Physical Activity Measures 5.1 The Short International Physical Activity Questionnaire

The short International Physical Activity Questionnaire (IPAQ) measures physical activity over the previous 7 days or a typical week over four domains (table II). Time spent in highintensity, medium-intensity and walking activities and sitting is recorded and MET scores are obtained for each category where 1 MET is the resting metabolic rate during quiet sitting. A total score is derived as well as separate scores for each category except sitting. The questionnaire was designed for physical activity surveillance across a variety of cultures in response to the need for a standardized physical activity measure. The short IPAQ has been validated in general populations across a number of countries. Validity has been assessed against accelerometers or motion detecting devices with only fair agreement (p = 0.30–0.39);[158-160] however, this may be due to accelerometers not measuring all aspects of physical activity and consequently underestimating physical activity in some people.[161] Good convergent validity (r ‡ 0.5) was demonstrated between the IPAQ and other physical activity questionnaires and physical activity logs.[160,162,163] Though not validated in injury populations, one study showed that patients with greater severity of osteoarthritis of the knee and hip had lower activity levels as measured by the short IPAQ.[120] A similar relationship may exist between disability due to injury and the short IPAQ. A large international study found good test-retest reliability (p = 0.74) for telephone-administered and self-administered questionnaires,[158] with lower reliability in rural and undeveloped areas. Another study found moderate reliability (ICC = 0.68);[164] however, two European studies found low reliability (ICC = 0.45–0.54).[160,165] Physical activity can vary from week to week, hence differences may be partly related to different administration periods as ª 2010 Adis Data Information BV. All rights reserved.

studies with longer re-administration periods reported lower reliability.[160,165] The short IPAQ has the advantage of measuring physical activity across a number of domains and is suitable for use in a variety of cultures. The responsiveness of the short IPAQ in an injury context is unknown and consequently the developers do not recommend its use in smallscale intervention studies.[166] The variability of the results obtained from reliability studies suggests that further evaluation is required. 5.2 Paffenbarger Physical Activity Questionnaire

The Paffenbarger Physical Activity Questionnaire (PPAQ) or College Alumnus physical activity questionnaire was developed for use in exercise and chronic disease epidemiology studies. The PPAQ measures calories expended in sport, leisure and recreational activities, as well as flights of stairs climbed and city blocks walked. Sport and recreation activities are listed as weeks in the past year that each activity was performed, whereas other areas are recorded for the previous week (table II). This allows for variation in sporting participation habits but in the context of injury will limit the time frames for which it can be used. The PPAQ does not measure physical activity across multiple domains; however, it does allow time spent in each sport or recreation activity to be listed separately.[167] Validity and reliability studies have involved general adult populations and university students. Only fair agreement was demonstrated between the PPAQ and accelerometer readings (r = 0.29–0.30)[168,169] and activity logs that measured total activity (r = 0.31).[168] However, when only the activity log items included in the PPAQ were compared, a high level of correlation was found (r = 0.60).[169] Another study found good agreement (r > 0.50) between the PPAQ and five of seven other physical activity questionnaires.[170] Two studies found good test-retest reliability when administered within a time frame of 1 month,[168,169] whereas a study that used a time frame of 7–12 weeks found poor test-retest reliability (r = 0.58). When the reliability results Sports Med 2010; 40 (2)

Measuring the Burden of Sports Injury

were recalculated using only participants who reported no change in their activity levels, the correlation increased significantly (r = 0.69).[171] The PPAQ appears to be valid and reliable in populations with similar demographics to a sport and active recreation population; however, the use of selected domains and activities means that some aspects of physical activity such as cycling to and from work or work-based activities may be missed. The 12-month time frame for recording sport and recreation activities is only appropriate for measuring recovery over very long periods. Nevertheless, the scoring system could be modified to cover activities over the last month or week. 5.3 Godin Leisure-Time Exercise Questionnaire

The Godin Leisure-Time Exercise Questionnaire (GLETQ) is a four-item questionnaire used to assess the number of times in an average week participants spend in strenuous, moderate and mild physical activity for more than 15 minutes[172] (table II). A score is obtained that can be converted to METs. Though not used specifically as an outcome measure, the GLETQ has been used to measure physical activity in patients with existing lower limb and spinal cord injuries.[173,174] The GLETQ has undergone minimal psychometric testing. Accelerometer correlations were fair (r = 0.32–0.35).[168,175] Correlations between the GLETQ and a 4-week activity diary were lower than the other questionnaires (r = 0.36).[168] High test-retest reliability was reported (r = 0.75–0.82) when administered within a 2-week period.[175,176] The GLETQ has demonstrated changes in activity levels across phases of treatment and recovery in breast cancer patients[177] and could show changes in activity associated with injury and its phases of recovery. The GLETQ measures aspects of physical activity relevant to a sport and active recreation population but it does not account for physical activity across other domains. The GLETQ measures exercise episodes per week greater than 15 minutes rather than actual time spent and thus may underestimate overall activity levels. The reliability of the GLETQ is good and its brevity is ª 2010 Adis Data Information BV. All rights reserved.

155

a desirable trait for a sport and active recreation population. 6. Conclusion Despite the plethora of outcome measures available, none have been specifically designed to measure injury outcomes in a general sport and active recreation population. In the absence of a purpose-designed instrument, there are existing generic measures that could, alone or in combination with others, be useful for measuring outcomes in this group. The SF-36 covers many of the areas of HR-QOL relevant to a sport and active recreation population and enables comparison with other disease and injury populations. Where a detailed measure of function is required, the MFA could be useful, while the GOSE may have merit as a global measure of function. Physical activity measures present a standardized method for measuring return to activity in sport and active recreation populations with the short IPAQ appearing to be the most comprehensive of this group. Ultimately, the choice of outcome measure will depend on the requirements of the users. So far there is no core set for sport and active recreation injury patients and though the ICF can provide a general framework in which to assess the appropriateness of existing measures, the development of a core set would assist researchers and clinicians in selecting the combination of outcome measures that would provide the most comprehensive assessment of disability and recovery in this group. Future research should focus on validating existing generic measures suitable for sport and active recreation populations as well as developing a measure specific to their requirements based on an ICF core set. Only through improved measurement of outcomes will gains be made in quantifying the burden of sport and active recreation injury outcomes. Acknowledgements Dr Belinda Gabbe was supported by a Career Development Award from the National Health and Medical Research

Sports Med 2010; 40 (2)

Andrew et al.

156

Council of Australia during the preparation of this review. The authors have no conflicts of interest that are directly relevant to the content of this review.

19. 20.

References 1. Dekker R, van der Sluis C, Groothoff J, et al. Long-term outcome of sports injuries: results after inpatient treatment. Clin Rehabil 2003; 17 (5): 480-7 2. Dekker R, Groothoff JW, van der Sluis CK, et al. Longterm disabilities and handicaps following sports injuries: outcome after outpatient treatment. Disabil Rehabil 2003; 25 (20): 1153-7 3. World Health Organisation. Disabilities, 2008 [online]. Available from URL: http://www.who.int/topics/dis abilities/en/ [Accessed 2008 Feb 12] 4. World Health Organisation. International classification of functioning, disability and health. Geneva: World Health Organisation, 2001 5. Finch C, Little C, Garnham A. Quality of life improvements after sports injury. Int J Inj Contr Saf Promot 2001; 8 (2): 113-5 6. McAllister DR, Motamedi AR, Hame SL, et al. Quality of life assessment in elite collegiate athletes. Am J Sports Med 2001; 29 (6): 806-10 7. Wang JC, Shapiro MS, Hatch JD, et al. The outcome of lumbar discectomy in elite athletes. Spine 1999; 24 (6): 570-3 8. Malmberg J, Miilunpalo S, Pasanen M, et al. Characteristics of leisure time physical activity associated with risk of decline in perceived health-a 10-year follow-up of middle aged and elderly men and women. Prev Med 2005; 41 (1): 141-50 9. Tanasescu M, Leitzmann M, Rimm E, et al. Exercise type and intensity in relation to coronary heart disease in men. JAMA 2002; 288 (16): 1994-2000 10. Yu S, Yarnell J, Sweetnam P, et al. What level of physical activity protects against premature cardiovascular death? The Caerphilly study. Heart 2003; 89 (5): 502-6 11. Sigal R, Wasserman D, Kenny G, et al. Physical activity/ exercise and type 2 diabetes. Diabetes Care 2004; 27 (10): 2518-39 12. Emmons K, McBride C, Puleo E, et al. Prevalence and predictors of multiple behavioural risk factors for colon cancer. Prev Med 2005; 40 (5): 527-34 13. Guyatt GH, Feeny DH, Patrick DL. Measuring healthrelated quality of life. Ann Intern Med 1993; 118 (8): 622-9 14. Wright JG. Outcomes research: what to measure. World J Surg 1999; 23 (12): 1224-6 15. Aaronson NK. Quantitative issues in health related quality of life assessment. Health Policy 1998; 10: 217-30 16. Bergner M, Rothman M. Health status measures: an overview and guide for selection. Annu Rev Public Health 1987; 8: 191-210 17. Evans T. Outcome measurement in athletic therapy: selecting the appropriate outcomes tool. Athl Ther Today 2004; 16 (6): 15-8 18. Pynsent PB. Choosing an outcome measure. In: Pynsent PB, Fairbank J, Carr A, editors. Outcome measures in

ª 2010 Adis Data Information BV. All rights reserved.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

orthopaedics and orthopaedic trauma. London: Arnold, 2004: 3-5 Cohen J. Statistical power analysis for the behavioural sciences. New York: Academic Press, 1977 Martin DP, Engelberg R, Agel J, et al. Comparison of the Musculoskeletal Function Assessment Questionnaire with the Short Form-36, the Western Ontario and McMasters Universities Osteoarthritis Index, and the Sickness Impact Profile Health-Status Measures. J Bone Joint Surg Am 1997; 79A: 1323-35 MacKenzie EJ, Sacco WJ, Luchter S, et al. Validating the Functional Capacity Index as a measure of outcome following blunt trauma. Qual Life Res 2002; 11: 797-808 Fitzpatrick R. Measures of health status, health-related quality of life and patient satisfaction. In: Pynsent PB, Fairbank J, Carr A, editors. Outcome measures in orthopaedics and orthopaedic trauma. London: Arnold, 2004: 56 Taylor D, Tenuta J, Uhorchak J, et al. Aggressive surgical treatment and early return to sports in athletes with grade III syndesmosis sprains. Am J Sports Med 2007; 35 (11): 1833-8 Saxena A, Eakin C. Articular talar injuries in athletes: results of microfracture and autogenous bone graft. Am J Sports Med 2007; 35 (10): 1680-7 Seroyer S, Tejwani S, Bradley J. Arthroscopic capsulolabral reconstruction of the type VIII superior labrum anterior posterior lesion: mean 2-year follow-up on 13 shoulders. Am J Sports Med 2007; 35 (9): 1477-83 Frohm A, Saartok T, Halvorsen K, et al. Eccentric treatment for patellar tendinopathy: a prospective randomised short-term pilot study of two rehabilitation protocols. Br J Sports Med 2007; 41 (7): e7 von Porat A, Henriksson M, Holmstrom E, et al. Knee kinematics and kinetics in former soccer players with a 16-year-old ACL injury: the effects of twelve weeks of knee-specific training. BMC Musculoskelet Disord 2007; 8: 35 Larrain M, Montenegro H, Mauas D, et al. Arthroscopic management of traumatic anterior shoulder instability in collision athletes: analysis of 204 cases with a 4- to 9-year follow-up and results with the suture anchor technique. Arthroscopy 2006; 22 (12): 1283-9 Bradley J, Baker Cr, Kline A, et al. Arthroscopic capsulolabral reconstruction for posterior instability of the shoulder: a prospective study of 100 shoulders. Am J Sports Med 2006; 34 (7): 1061-71 Baums M, Kahl E, Schultz W, et al. Clinical outcome of the arthroscopic management of sports-related ‘‘anterior ankle pain’’: a prospective study. Knee Surg Sports Traumatol Arthrosc 2006; 14 (5): 482-6 Ogon P, Maier D, Jaeger A, et al. Arthroscopic patellar release for the treatment of chronic patellar tendinopathy. Arthroscopy 2006; 22 (4): 462.e1-5 Tambe A, Godsiff S, Mulay S, et al. Anterior cruciate ligament insufficiency: does delay in index surgery affect outcome in recreational athletes. Int Ortho 2006; 30 (2): 104-9 Gudas R, Kalesinskas R, Kimtys V, et al. A prospective randomized clinical study of mosaic osteochondral autologous transplantation versus microfracture for the

Sports Med 2010; 40 (2)

Measuring the Burden of Sports Injury

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

treatment of osteochondral defects in the knee joint in young athletes. Arthroscopy 2005; 21 (9): 1066-75 Visnes H, Hoksrud A, Cook J, et al. No effect of eccentric training on jumper’s knee in volleyball players during the competitive season: a randomized clinical trial. Clin J Sport Med 2005; 15 (4): 227-34 Gobbi A, Nunag P, Malinowski K. Treatment of full thickness chondral lesions of the knee with microfracture in a group of athletes. Knee Surg Sports Traumatol Arthrosc 2005; 13 (3): 213-21 Ide J, Maeda S, Takagi K. Sports activity after arthroscopic superior labral repair using suture anchors in overhead-throwing athletes. Am J Sports Med 2005; 33 (4): 507-14 Young M, Cook J, Purdam C, et al. Eccentric decline squat protocol offers superior results at 12 months compared with traditional eccentric protocol for patellar tendinopathy in volleyball players. Br J Sports Med 2005; 39 (2): 102-5 Gobbi A, Domzalski M, Pascual J. Comparison of anterior cruciate ligament reconstruction in male and female athletes using the patellar tendon and hamstring autografts. Knee Surg Sports Traumatol Arthrosc 2004; 12 (6): 534-9 Enad J, El Attrache N, Tibone J, et al. Isolated electrothermal capsulorrhaphy in overhand athletes. J Shoulder Elb Surg 2004; 13 (2): 133-7 Reinold M, Wilk K, Hooks T, et al. Thermal-assisted capsular shrinkage of the glenohumeral joint in overhead athletes: a 15- to 47-month follow-up. J Orthop Sport Phys 2003; 33 (8): 455-67 Gobbi A, Tuy B, Mahajan S, et al. Quadrupled bonesemitendinosus anterior cruciate ligament reconstruction: a clinical investigation in a group of athletes. Arthroscopy 2003; 19 (7): 691-9 Kim S, Ha K, Park J, et al. Arthroscopic posterior labral repair and capsular shift for traumatic unidirectional recurrent posterior subluxation of the shoulder. J Bone Joint Surg 2003; 85-A (8): 1479-87 Gobbi A, Mahajan S, Zanazzo M, et al. Patellar tendon versus quadrupled bone-semitendinosus anterior cruciate ligament reconstruction: a prospective clinical investigation in athletes. Arthroscopy 2003; 19 (6): 592-601 Meighan A, Keating J, Will E. Outcome after reconstruction of the anterior cruciate ligament in athletic patients: a comparison of early versus delayed surgery. J Bone Joint Surg 2003; 85 (4): 521-4 Marcacci M, Zaffagnini S, Iacono F, et al. Intra- and extraarticular anterior cruciate ligament reconstruction utilizing autogeneous semitendinosus and gracilis tendons: 5-year clinical results. Knee Surg Sports Traumatol Arthrosc 2003; 11 (1): 2-8 Bonneux I, Vandekerckhove B. Arthroscopic partial lateral meniscectomy long-term results in athletes. Acta Orthop Belg 2002; 68 (4): 356-61 Krips R, van Dijk C, Lehtonen H, et al. Sports activity level after surgical treatment for chronic anterolateral ankle instability: a multicenter study. Am J Sports Med 2002; 30 (1): 13-9 Jerre R, Ejerhed L, Wallmon A, et al. Functional outcome of anterior cruciate ligament reconstruction in recrea-

ª 2010 Adis Data Information BV. All rights reserved.

157

49.

50.

51.

52.

53.

54.

55.

56.

57.

58.

59.

60.

61.

62.

63.

tional and competitive athletes. Scand J Med Sci Sports 2001; 11 (6): 342-6 Mishra D, Fanton G. Two-year outcome of arthroscopic bankart repair and electrothermal-assisted capsulorrhaphy for recurrent traumatic anterior shoulder instability. Arthroscopy 2001; 17 (8): 844-9 Larrain M, Botto G, Montenegro H, et al. Arthroscopic repair of acute traumatic anterior shoulder dislocation in young athletes. Arthroscopy 2001; 17 (4): 373-7 Echemendia R, Putukian M, Mackin R, et al. Neuropsychological test performance prior to and following sports-related mild traumatic brain injury. Clin J Sport Med 2001; 11 (1): 23-31 Uhorchak J, Arciero R, Huggard D, et al. Recurrent shoulder instability after open reconstruction in athletes involved in collision and contact sports. Am J Sports Med 2000; 28 (6): 794-9 Nakayama Y, Shirai Y, Narita T, et al. Knee functions and a return to sports activity in competitive athletes following anterior cruciate ligament reconstruction. J Nippon Med Sch 2000; 67 (3): 172-6 Wiger P, Brandsson S, Kartus J, et al. A comparison of results after arthroscopic anterior cruciate ligament reconstruction in female and male competitive athletes: a two- to five-year follow-up of 429 patients. Scand J Med Sci Sports 1999; 9 (5): 290-5 Yoneda M, Hayashida K, Wakitani S, et al. Bankart procedure augmented by coracoid transfer for contact athletes with traumatic anterior shoulder instability. Am J Sports Med 1999; 27 (1): 21-6 Testa V, Capasso G, Maffulli N, et al. Ultrasound-guided percutaneous longitudinal tenotomy for the management of patellar tendinopathy. Med Sci Sports Exerc 1999; 31 (11): 1509-15 O’Neill D. Arthroscopic Bankart repair of anterior detachments of the glenoid labrum: a prospective study. J Bone Joint Surg 1999; 81 (10): 1357-66 Takeda H, Watarai K, Ganev G, et al. Modified Bankart procedure for recurrent anterior dislocation and subluxation of the shoulder in athletes. Int Ortho 1998; 22 (6): 361-5 DeBerardino T, Arciero R, Taylor D. Arthroscopic treatment of soft-tissue impingement of the ankle in athletes. Arthroscopy 1997; 13 (4): 492-8 Novak P, Bach BJ, Hager C. Clinical and functional outcome of anterior cruciate ligament reconstruction in the recreational athlete over the age of 35. Am J Knee Surg 1996; 9 (3): 111-6 Marcacci M, Zaffagnini S, Visani A, et al. Arthroscopic reconstruction of the anterior cruciate ligament with Leeds-Keio ligament in non-professional athletes: results after a minimum 5 years’ follow-up. Knee Surg Sports Traumatol Arthrosc 1996; 4 (1): 9-13 Montgomery Wr, Jobe F. Functional outcomes in athletes after modified anterior capsulolabral reconstruction. Am J Sports Med 1994; 22 (3): 352-8 Verhaven E, DeBoeck H, Haentjens P, et al. Surgical treatment of acute type-V acromioclavicular injuries in athletes. Am J Sports Med 1993; 20 (4): 702-6

Sports Med 2010; 40 (2)

158

64. Aiello L, Iwamoto M, Guyer D. Penetrating ocular fishhook injuries: surgical management and long-term visual outcome. Opthamology 1992; 99 (6): 862-6 65. Argo D, Trenhaile SW, Savoie 3rd FH, et al. Operative treatment of ulnar collateral ligament insufficiency of the elbow in female athletes. Am J Sports Med 2006; 34 (3): 431-7 66. Mithoefer K, Peterson L, Mandelbaum BR, et al. Articular cartilage repair in soccer players with autologous chondrocyte transplantation: functional outcome and return to competition. Am J Sports Med 2005; 33 (11): 1639-46 67. Charron KM, Schepsis AA, Voloshin I. Arthroscopic distal clavicle resection in athletes: a prospective comparison of the direct and indirect approach. Am J Sports Med 2007; 35 (1): 53-8 68. Miller SF, Congeni J, Swanson K. Long-term functional and anatomical follow-up of early detected spondylolysis in young athletes. Am J Sports Med 2004; 32 (4): 928-33 69. Myklebust G, Holm I, Maehlum S, et al. Clinical, functional, and radiologic outcome in team handball players 6 to 11 years after anterior cruciate ligament injury: a follow-up study. Am J Sports Med 2003; 31 (6): 981-9 70. Coleman BD, Khan KM, Kiss ZS, et al. Open and arthroscopic patellar tenotomy for chronic patellar tendinopathy: a retrospective outcome study. Victorian Institute of Sport Tendon Study Group. Am J Sports Med 2000; 28 (2): 183-90 71. Peers KHE, Lysens RJJ, Brys P, et al. Cross-sectional outcome analysis of athletes with chronic patellar tendinopathy treated surgically and by extracorporeal shock wave therapy. Clin J Sport Med 2003; 13 (2): 79-83 72. Khan W, Fahmy N. The S-Quattro in the management of sports injuries of the fingers. Injury 2006; 37: 860-8 73. Mithoefer K, Williams R, Warren R, et al. High-impact athletics after knee articular cartilage repair: a prospective evaluation of the microfracture technique. Am J Sports Med 2006; 34 (11): 1413-8 74. Peterson W, Welp R, Rosenbaum D. Chronic achilles tendinopathy: a prospective randomised study comparing the therapeutic effect of eccentric training, the airheel brace and a combination of both. Am J Sports Med 2007; 35 (10): 1659-67 75. Guskiewicz KM, Marshall SW, Bailes J, et al. Recurrent concussion and risk of depression in retired professional football players. Med Sci Sport Exer 2007: 39 (6): 903-9 76. Naal FD, Fischer M, Preuss A, et al. Return to sport and recreational activity after unicompartmental knee arthroplasty. Am J Sports Med 2007; 35: 1688-95 77. Anandacoomarasamy L, Barnsley L. Long term outcomes of inversion ankle injuries. Br J Sports Med 2005; 39: e14 78. Mazzocca AD, Brown Jr FM, Carreira DS, et al. Arthroscopic anterior shoulder stabilization of collision and contact athletes. Am J Sports Med 2005; 33 (1): 52-60 79. Debnath UK, Freeman BJC, Gregory P, et al. Clinical outcomes and return to sport after the surgical treatment of spondylolysis in young athletes. J Bone Joint Surg 2003; 85-B (2): 244-9 80. Williams R, Strickland S, Cohen M, et al. Arthroscopic repair for traumatic posterior shoulder instability. Am J Sports Med 2003; 31: 203-9

ª 2010 Adis Data Information BV. All rights reserved.

Andrew et al.

81. von Porat A, Roos E, Roos H. High prevalence of osteoarthritis 14 years after an anterior cruciate ligament tear in male soccer players: a study of radiographic and patient relevant outcomes. Ann Rheum Dis 2004; 63: 269-73 82. Nicholas SJ, Nicholas JA, Nicholas C, et al. The health status of retired American football players: Super Bowl III revisited. Am J Sports Med 2007; 35 (10): 1674-9 83. Meller R, Krettek C, Gosling T, et al. Recurrent shoulder instability among athletes: changes in quality of life, sports activity, and muscle function following open repair. Knee Surg Sports Traumatol Arthrosc 2007; 15: 295-304 84. Turner A, Barlow J, Heathcote-Elliot C. Long term health impact of playing professional football in the United Kingdom. Br J Sports Med 2000; 34 (5): 332-6 85. Lindsay KW, McLatchie G, Jennett B. Serious head injury in sport. BMJ 1980; 281 (6243): 789-91 86. Giza E, Mithofer K, Matthews H, et al. Hip fracturedislocation in football: a report of two cases and review of the literature. Br J Sports Med 2004; 38 (e17): 1-2 87. Lee AJ, Garraway WM, Hepburn W, et al. Influence of rugby injuries on players’ subsequent health and lifestyle: beginning a long term follow up. Br J Sports Med 2001; 35 (1): 38-42 88. Gobbi A, Francisco R. Factors affecting return to sport after anterior cruciate ligament reconstruction with patellar tendon and hamstring graft: a prospective clinical investigation. Knee Surg Sports Traumatol Arthrosc 2006; 14 (10): 1021-8 89. Valderrabano V, Perren T, Ryf C, et al. Snowboarder’s Talus fracture: treatment outcome of 20 cases after 3.5 years. Am J Sports Med 2005; 33 (6): 871-80 90. Holtslag HR, van Beek EF, Lindeman E, et al. Determinents of long-term functional consequences after major trauma. J Trauma 2007; 62 (4): 919-27 91. Jurkovich G, Mock C, MacKenzie E, et al. The Sickness Imact Profile as a tool to evaluate outcome in trauma patients. J Trauma 1995; 39 (4): 625-31 92. DePalma JA, Fedorka P, Simko LC. Quality of life experienced by severely injured trauma survivors. AACN Clin Issues 2003; 14 (1): 54-63 93. Harris IA, Young JM, Rae H, et al. Predictors of general health after major trauma. J Trauma 2008; 64 (4): 969-74 94. Ponsford J, Hill B, Karamitsios M, et al. Factors influencing outcome after orthopedic trauma. J Trauma 2008; 64 (4): 1001-9 95. Post RB, van der Sluis CK, Ten Duis HJ. Return to work and quality of life in severely injured patients. Disabil Rehabil 2006; 28 (22): 1399-404 96. Sampalis JS, Liberman M, Davis L, et al. Functional status and quality of life in survivors of injury treated at tertiary trauma centers: what are we neglecting? J Trauma 2006; 60 (4): 806-13 97. Morris S, Lenihan B, Duddy L, et al. Outcome after musculoskeletal trauma treated in a regional hospital. J Trauma 2000; 49 (3): 461-9 98. Holbrook TL, Hoyt DB, Anderson JP. The impact of major in-hospital complications on functional outcomes and quality of life after trauma. J Trauma 2001; 1 (50): 91-5

Sports Med 2010; 40 (2)

Measuring the Burden of Sports Injury

99. Holbrook T, Hoyt D, Anderson J. The importance of gender on outcome after major trauma: functional and psychologic outcomes in women versus men. J Trauma 2001; 50 (2): 270-3 100. Urquhart D, Williamson O, Gabbe B, et al. Outcomes of patients with orthopaedic trauma admitted to Level 1 Trauma Centres. ANZ J Surg 2006; 76 (7): 600-6 101. Watson WL, Ozanne-Smith J, Richardsons J. An evaluation of the assessment of quality of life utility instrument as a measure of the impact of injury on health-related quality of life. Int J Inj Contr Saf Promot 2005; 12 (4): 227-39 102. Paxton EW, Fithian DC, Stone ML, et al. The reliability and validity of knee-specific and general health instruments in assessing acute patellar dislocation outcomes. Am J Sports Med 2003; 31 (4): 487-92 103. Marx RG, Jones EC, Answorth A, et al. Reliability, validity and responsiveness of four knee outcome scales for athletic patients. J Bone Joint Surg 2001; 83-A: 1459-69 104. Findler M, Cantor J, Haddad L, et al. The reliability and validity of the SF-36 health survey questionnaire for use with individuals with traumatic brain injury. Brain Inj 2001; 15 (8): 715-23 105. van Baalen B, Odding E, van Woensel MPC, et al. Reliability and sensitivity to change of measurement instruments used in a traumatic brain injury population. Clin Rehabil 2006; 20 (8): 686-700 106. MacKenzie EJ, McCarthy ML, Ditunno JF, et al. Using the SF-36 for characterising outcome after multiple trauma involving head injury. J Trauma 2002; 52 (3): 527-34 107. Michaels AJ, Madey SM, Krieg JC, et al. Traditional injury scoring underestimates the relative consequences of orthopaedic injury. J Trauma 2001; 50 (3): 389-96 108. MacDermid JC, Richards RS, Donner A, et al. Responsiveness of the short form-36, disability of the arm, shoulder, and hand questionnaire, patient-rated wrist evaluation, and physical impairment measurements in evaluating recovery after a distal radius fracture. J Hand Surg 2000; 25 (2): 330-40 109. Kiely JM, Brasel K, Weidner K, et al. Predicting quality of life six months after traumatic injury. J Trauma 2006; 61: 791-8 110. Bergner M, Bobbit RA, Pollard WE, et al. The Sickness Impact Profile: validation of a health status measure. Med Care 1976; 14 (1): 57-67 111. Butcher JL, MacKenzie EJ, Cushing B, et al. Long-term outcomes after lower extremity trauma. J Trauma 1996; 41 (1): 4-9 112. Baldry Currens JA. Evaluation of disability and handicap following injury. Injury 2000; 31 (2): 99-106 113. Hall K, Mann N, High W, et al. Functional measures after traumatic brain injury: ceiling effects of FIM, FIM+FAM, DRS and CIQ. J Head Trauma Rehabil 1996; 11: 27-39 114. Gurka J, Flemingham K, Baguley I, et al. Utility of the Functional Assessment Measure after discharge from inpatient rehabilitation. J Head Trauma Rehabil 1999; 14 (3): 247-56 115. McCarthy ML, MacKenzie EJ. Predicting ambulatory function following lower extremity trauma using the

ª 2010 Adis Data Information BV. All rights reserved.

159

116.

117.

118.

119.

120.

121.

122.

123.

124. 125.

126.

127.

128.

129.

130.

131.

132.

functional capacity index. Accident Anal Prev 2001; 33 (6): 821-31 Wilson JTL, Pettigrew LEL, Teasdale GM. Emotional and cognitive consequences of head injury in relation to the Glasgow Outcome Scale. J Neurol Neurosurg Psychiatry 2000; 69 (2): 204-9 Whitnall L, McMillan T, Murray GD, et al. Disability in young people and adults after head injury: 5-7 year follow up of a prospective cohort study. J Neurol Neurosurg Psychiatr 2006; 77: 640-5 Engelberg R, Martin D, Agel J, et al. Musculoskeletal Function Assessment Instrument: criterion and construct validity. J Orthop Res 1996; 14 (2): 182-92 Swiontkowski M, Engelberg R, Martin D, et al. Short Musculoskeletal Function Assessment questionnaire: validity, reliability, and responsiveness. J Bone Joint Surg Am 1999; 81 (9): 1245-60 Rosemann T, Kuehlein T, Laux G, et al. Osteoarthritis of the knee and hip: a comparison of factors associated with physical activity. Clin Rheumatol 2007; 26 (11): 1811-7 Kaplan R, Giants T, Sieber W, et al. The Quality of Well Being Scale: critical similarities and differences with SF-36. Int J Qual Health Care 1998; 10 (6): 509-20 Anderson JP, Holbrook TL. Quality of well-being profiles followed paths of health status change at micro- and meso-levels in trauma patients. J Clin Epidemiol 2007; 60 (3): 300-8 Sieber W, Groessl EJ, David KM, et al. Quality of WellBeing self-administered (QWB-SA) Scale. San Diego (CA): University of California, 2004 Nemeth G. Health related quality of life outcome instruments. Eur Spine J 2006; 15 Suppl. 1: S44-51 Brazier J, Jones N, Kind P. Testing the validity of the Euroqol and comparing it with the SF-36 Health Survey Questionnaire. Qual Life Res 1993; 2 (3): 169-80 Hawthorne G, Richardson J, Day NA. A comparison of the Assessment of Quality of Life (AQoL) with four other generic utility instruments. Ann Med 2001 Jul; 33 (5): 358-70 Hawthorne G, Osborne R. Population norms and meaningful differences for the Assessment of Quality of Life (AQoL) measure. Aust N Z J Public Health 2005 Apr; 29 (2): 136-42 Kopjar B. The SF-36 health survey: a valid measure of changes in health status after injury. Inj Prev 1996; 2 (2): 135-9 Dowrick A, Gabbe B, Williamson O, et al. Outcome instruments for the assessment of the upper extremity following trauma: a review. Injury 2005; 36: 468-76 Ware JE, Kosinski M, Keller S. A 12-item short-form health survey: construction of scales and preliminary tests of reliability and validity. Med Care 1996; 34 (3): 220-33 Leanderson J, Wredmark T. Teatment of acute ankle sprains: comparison of a semi-rigid ankle brace and compression bandage in 73 patients. Acta Orthop 1995; 66 (6): 529-31 Schenck RC, Blaschak MJ, Lance ED, et al. A prospective outcome study of rehabilitation programs and anterior cruciate ligament reconstruction. Arthroscopy 1997; 13 (3): 285-90

Sports Med 2010; 40 (2)

160

133. Ware Jr JE, Sherbourne CD. The MOS 36-item short-form health survey (SF-36). I. Conceptual framework and item selection. Med Care 1992; 30 (6): 473-83 134. Cieza A, Brokow T, Ewert T, et al. Linking health-status measurements to the international classification of function, disability and health. J Rehabil Med 2002; 34 (5): 205-10 135. Gilson BS, Gilson JS, Bergner M, et al. The Sickness Impact Profile: the development of an outcome measure of health care. Am J Public Health 1975; 65 (12): 1304-10 136. Bergner M, Bobbit RA, Gilson BS. The Sickness Impact Profile: development and final revision of a health status measure. Med Care 1981; 19 (8): 787-805 137. De Bruin AF, Diederiks JPM, de Witte LP, et al. Assessing the responsiveness of a functional status measure: the Sickness Impact Profile versus the SIP68. J Clin Epidemiol 1997; 50 (5): 529-40 138. De Bruin AF, Diederiks JPM, De Witte LP, et al. The development of a short generic version of the sickness impact profile. J Clin Epidemiol 1994; 47 (8): 407-18 139. De Bruin AF, Buys M, De Witte LP, et al. The sickness impact profile: SIP68, a short generic version. First evaluation of the reliability and reproducibility. J Clin Epidemiol 1994; 47 (8): 863-71 140. Gabbe B, Sutherland A, Williamson O, et al. Use of health care services 6 months following major trauma. Aus Health Rev 2007; 31 (4): 628-32 141. Stalp M, Koch C, Ruchholtz S, et al. Standardized outcome evaluation after blunt multiple injuries by scoring systems: a clinical follow-up investigation 2 years after injury. J Trauma 2002 Jun; 52 (6): 1160-8 142. Sutherland AG, Alexander DA, Hutchison JD. Recovery after musculoskeletal trauma in men and women. J Trauma 2005; 59 (1): 213-6 143. Sutherland AG, Alexander DA, Hutchison JD. The mind does matter: psychological and physical recovery after musculoskeletal trauma. J Trauma 2006; 61 (6): 1408-14 144. McKevitt E, Calvert E, Ng A, et al. Geriatric trauma: resource use and patient outcomes. Can J Surg 2003; 46 (3): 211-5 145. Schluter PJ, Cameron CM, Purdie DM, et al. How well do anatomical-based injury severity scores predict health service use in the 12 months after injury? Int J Inj Contr Saf Promot 2005; 12 (4): 241-6 146. Gotschall CS. The Functional Capacity Index, second revision: morbidity in the first year post injury. Int J Inj Contr Saf Promot 2005; 12 (4): 254-6 147. MacKenzie E, Damiano A, Miller T, et al. The development of the Functional Capacity Index. J Trauma 1996; 41 (5): 799-807 148. Gabbe B, Williamson O, Cameron P, et al. Choosing outcome assessment instruments for trauma registries. Acad Emerg Med 2005; 12 (8): 751-7 149. Neugebauer E, Bouillon B, Bullinger M, et al. Quality of life after multiple trauma: summary and recommendations of the consensus conference. Restor Neurol Neurosci 2002; 20 (3-4): 161-7 150. Wilson JTL, Pettigrew LEL, Teasdale GM. Structured interviews for the Glasgow Outcome Scale and the Extended Glasgow Outcome Scale: guidelines for their use. J Neurotrauma 1998; 15 (8): 573-82

ª 2010 Adis Data Information BV. All rights reserved.

Andrew et al.

151. Levin HS, Boake C, Song J, et al. Validity and sensitivity to change of the extended Glasgow Outcome Scale in mild to moderate traumatic brain injury. J Neurotrauma 2001; 18 (6): 575-84 152. Hudak AM, Caesar RR, Frol AB, et al. Functional outcome scales in traumatic brain injury: a comparison of the Glasgow Outcome Scale (Extended) and the Functional Status Examination. J Neurotrauma 2005; 22 (11): 1319-26 153. Clifton GL, Kruetzer JS, Choi SC, et al. Relationship between Glascow Outcome Scale and neuropsychological measures after brain injury. Neurosurg 1993; 33 (1): 34-9 154. Pettigrew LEL, Wilson JTL, Teasdale GM. Reliability of ratings on the Glasgow Outcome Scales from in-person and telephone structured interviews. J Head Trauma Rehabil 2003; 20 (2): 252-8 155. Wilson JTL, Edwards P, Fiddes H, et al. Reliability of postal questionnaires for the Glasgow Outcome Scale. J Neurotrauma 2002; 19 (9): 999-1006 156. Gabbe B, Finch C, Cameron P, et al. The incidence of serious injury and death during sport and recreation activities in Victoria, Australia. Br J Sports Med 2005; 39 (8): 573-7 157. Martin DP, Engelberg R, Agel J, et al. Development of a musculoskeletal extremity health status instrument: the Musculoskeletal Function Assessment Instrument. J Bone Joint Surg 1996; 14: 173-81 158. Craig C, Marshall A, Sjostrom M, et al. International Physical Activity Questionnaire: 12-country reliability and validity. Med Sci Sports Exerc 2003; 35 (8): 1381-95 159. Ekelund U, Sepp H, Brage S, et al. Criterion-related validity of the last 7-day, short form of the International Physical Activity Questionnaire in Swedish adults. Public Health Nutr 2006; 9 (2): 258-65 160. Mader U, Martin B, Schutz Y, et al. Validity of four short physical activity questionnaires in middle-aged persons. Med Sci Sports Exerc 2006; 38 (7): 1255-66 161. Pols MA, Peeters PHM, Kemper HCG, et al. Methodological aspects of physical activity assessment in epidemiological studies. Eur J Epidemiol 1998; 14: 63-70 162. Tehard B, Saris WHM, Astrup A, et al. Comparison of two physical activity questionnaires in obese subjects: The NUGENOB study. Med Sci Sports Exerc 2005; 37 (9): 1535-41 163. Macfarlane D, Lee C, Ho EY, et al. Convergent validity of six methods to assess physical activity in daily life. J Appl Physiol 2006; 101: 1328-34 164. Brown W, Trost S, Bauman A, et al. Test-retest reliability of four physical activity measures used in population surveys. J Sci Med Sport 2004; 7 (2): 205-15 165. Rutten A, Vuillemin A, Ooijendijk WTM, et al. Physical activity monitoring in Europe: the European Physical Activity Surveillance System (EUPASS) approach and indicator testing. Public Health Nutr 2003; 6 (4): 377-84 166. International physical activity questionnaire website. Guidelines for data processing and analysis of the International Physical Activity Questionnaire (IPAQ) [online]. Available from URL: http://www.ipaq.ki.se/scoring.pdf [Accessed 2008 Aug 6]

Sports Med 2010; 40 (2)

Measuring the Burden of Sports Injury

167. Paffenbarger R, Hyde R, Wing A, et al. Physical activity, all-cause mortality, and longevity of college alumni. New Engl J Med 1986; 314 (10): 605-13 168. Jacobs D, Ainsworth B, Hartman T, et al. A simultaneous evaluation of 10 commonly used physical activity questionnaires. Med Sci Sports Exerc 1993; 25 (1): 81-91 169. Ainsworth B, Leon A, Richardson M, et al. Accuracy of the College Alumnus Physical Activity Questionnaire. J Clin Epidemiol 1993; 46 (12): 1403-11 170. Albanes D, Conway J, Taylor P, et al. Validation and comparison of eight physical activity questionnaires. Epidemiology 1990; 1 (1): 65-71 171. Washburn R, Smith K, Goldfield S, et al. Reliability and physiologic correlates of the Harvard Alumni Activity Survey in a general population. J Clin Epidemiol 1991; 44 (12): 1319-26 172. Godin G, Shephard R. A simple method to assess exercise behaviour in the community. Can J Appl Sports Sci 1985; 10 (3): 141-6 173. Godin G, Colantonio A, Davis G, et al. Prediction of leisure time exercise behavior among a group of lower-limb disabled adults. J Clin Psychol 1986; 42 (2): 272-9

ª 2010 Adis Data Information BV. All rights reserved.

161

174. Noreau L, Shephard R, Simard C, et al. Relationship of impairment and functional ability to habitual activity and fitness following spinal cord injury. Int J Rehabil Res 1993; 16 (4): 265-75 175. Rauh M, Hovell M, Hofstetter C, et al. Reliability and validity of self-reported physical activity in Latinos. Int J Epidemiol 1992; 21 (5): 966-71 176. Reed J, Phillips D. Relationship between physical activity and the proximity of exercise facilities and home equipment used by undergraduate university students. J Am Coll Health 2005; 53 (6): 285-90 177. Valenti M, Porzio G, Aielli F, et al. Physical exercise and quality of life in breast cancer survivors. Int J Med Sci 2008; 5 (1): 24-8

Correspondence: Dr Belinda J. Gabbe, Department of Epidemiology and Preventive Medicine, Monash University, Alfred Hospital, Commercial Rd, Melbourne, VIC 3004, Australia. E-mail: [email protected]

Sports Med 2010; 40 (2)

Sports Med 2010; 40 (2): 163-178 0112-1642/10/0002-0163/$49.95/0

RESEARCH REVIEW

ª 2010 Adis Data Information BV. All rights reserved.

Match and Training Injuries in Rugby League A Review of Published Studies Doug A. King,1,2 Patria A. Hume,2 Peter D. Milburn3 and Dain Guttenbeil4 1 Emergency Department, Hutt Valley District Health Board, Lower Hutt, New Zealand 2 Sports Performance Research Institute New Zealand, School of Sport and Recreation, Auckland University of Technology, Auckland, New Zealand 3 School of Physiotherapy and Exercise Science, Gold Coast Campus Griffith University, Gold Coast, Queensland, Australia 4 New Zealand Rugby League Inc., Penrose, Auckland, New Zealand

Abstract

Rugby league is an international collision sport played by junior, amateur, semiprofessional and professional players. The game requires participants to be involved in physically demanding activities such as running, tackling, passing and sprinting, and musculoskeletal injuries are common. A review of injuries in junior and senior rugby league players published in Sports Medicine in 2004 reported that injuries to the head and neck and muscular injuries were common in senior rugby league players, while fractures and injuries to the knee were common in junior players. This current review updates the descriptive data on rugby league epidemiology and adds information for semiprofessional, amateur and junior levels of participation in both match and training environments using studies identified through searches of PubMed, CINHAL, Ovid, MEDLINE, SCOPUS and SportDiscus databases. This review also discusses the issues surrounding the definitions of injury exposure, injury rate, injury severity and classification of injury site and type for rugby league injuries. Studies on the incidence of injuries in rugby league have suffered from inconsistencies in the injury definitions utilized. Some studies on rugby league injuries have utilized a criterion of a missed match as an injury definition, total injury incidences or a combination of both timeloss and non-time-loss injuries, while other studies have incorporated a medical treatment injury definition. Efforts to establish a standard definition for rugby league injuries have been difficult, especially as some researchers were not in favour of a definition that was all-encompassing and enabled non-time-loss injuries to be recorded. A definition of rugby league injury has been suggested based on agreement by a group of international researchers. The majority of injuries occur in the match environment, with rates typically increasing as the playing level increases. However, professional level injury rates were reportedly less than semiprofessional participation. Only a few studies have reported training injuries in rugby league, where injury rates were reported to be less than match injuries. Approximately 16–30% of all

King et al.

164

rugby league injuries have been reported as severe, which places demands upon other team members and, if the player returns to playing too early, places them at an increased risk of further injuries. Early research in rugby league identified that ligament and joint injuries were the common injuries, occurring primarily to the knee. More recently, studies have shown a change in anatomical injury sites at all levels of participation. Although the lower limb was the frequent injury region reported previously, the shoulder has now been reported to be the most common injury site. Changes in injury site and type could be used to prompt further research and development of injury reduction programmes to readdress the issue of injuries that occur as a result of participation in rugby league activities. Further research is warranted at all participation levels of rugby league in both the match and training environments to confirm the strongest risk factors for injury.

Rugby league is a team contact sport played internationally. It consists of 13 players in each team, and is typically (but not always) played under a limited interchange rule where up to 12 interchanges of players are permitted in matches. Each team is permitted six tackles in possession of the ball with which they must advance into the opposition’s territory and score a try.[1-3] The ball must only be passed backwards but can be carried or kicked into the opposition’s territory.[1,3] At the completion of six tackles the ball is immediately given to the opposition team to commence their set of six tackles.[1,2] Most teams often kick the ball after the fifth tackle (ceding the possession to the opposition) to avoid being tackled a sixth time, in an attempt to gain further territorial advantage. The same players are therefore involved in both attack and defence. As with all sport, there is a risk of sustaining an injury when participating in rugby league activities. The game is intermittent in nature, requiring participants to compete in a challenging physical contest. Players often undergo frequent bouts of high-intensity activity (e.g. tackling, sprinting, running and passing) interspersed with short bouts of low-intensity activity (e.g. jogging, walking and standing).[4-7] As a result of the intermittent and contact nature of the game, the physiological demands of rugby league are complex. Players are required to have maximal aerobic power, speed, muscular strength, and power and agility appropriately developed to be able to compete in the match environment.[4-7] As a ª 2010 Adis Data Information BV. All rights reserved.

result of the physical and intense nature of the game, musculoskeletal injuries are common.[1,7,8] The rugby league team consists of two main groups of participants (forwards and backs) on the field and reserves available for interchange.[9-11] The demands on the participants vary according to the specific positions played,[9-11] with forwards (two props, one hooker, two second rowers and one lock) more predominately involved in large numbers of physical collisions and tackles.[12] Backs (one half-back, one stand-off/five-eighth, two centres, two wings and one fullback) spend more time in free running but are also involved in tackles and collisions.[7,10-12] The aim of this study was to review and update the descriptive data on rugby league injury epidemiology and add information for semiprofessional, amateur and junior levels of participation in both match and training environments. 1. Methods Searches of PubMed, CINHAL, Ovid, MEDLINE, SCOPUS, and SportDiscus databases were performed to identify studies published in English prior to October 2008. The computer databases provided access to sports-oriented and biomedical journals, serial publications, books, theses, conference papers and related research published since 1948. Terms utilized for the search of relevant research studies included ‘rugby league’ and ‘injury incidence’. Qualifying studies were mainly uncontrolled trials and the outcome Sports Med 2010; 40 (2)

Match and Training Injuries in Rugby League

165

variables were injury incidence, injury site, injury type, match time, seasonal variations and injury severity of rugby league players. 2. Findings Thirty-five published studies, mainly conducted in Australia and New Zealand, have reported the injury incidence of rugby league participants in matches (i.e. tournaments and competitions) and/or training (table I). Tournaments differ from matches in two main areas: (i) tournament matches can vary in the periods between matches (i.e. several matches in one day through to one

game a week); and (ii) they can comprise players of different participation levels either playing against each other or competing in the same team. Conversely, matches are typically undertaken on a regular schedule over a longer period for a specific participation level. Injury incidence has changed with changes to the rules of the game, and with the resulting changes to the activity and intensity at which the game is played. Although there have been World Cup tournaments for both male and female participants, there have been no published studies on the injury incidence in these competitions, and several levels of competition do not include females (table I).

Table I. Prospective rugby league epidemiological publications analysed by type of playing population for matches and training by country of origin Playing population level

Sex

Studies conducted on match injuries

Studies conducted on training injuries

Professional tournament

Male

None

None

Professional match

Male Male

England[13-21] Australia[2,22-24]

None None

Semiprofessional tournaments

None

None

None

Semiprofessional teams

Male Male

Australia[8,25] New Zealand[26]

Australia[8,27,28] Australia[8,27,28]

Sub-elite tournament

Male

None

None

Sub-elite teams

Male

Australia[29]

Australia[30,31]

Amateur tournament

Male Female

None New Zealand[32]

None None

Amateur teams

Male Male

Australia[33] New Zealand[34,35]

New Zealand[36] New Zealand[36]

Masters tournament

Male

None

None

Masters teams

Male

None

None

School-level tournaments

Male Female

None None

None None

School-level team

Male Female

None None

None None

School-level tournaments

Male Female

None None

None None

Junior

Male Male Female

Australia[37,38] New Zealand[39] None

None None None

Sevens

Male Male

Australia[40] New Zealand[41]

None None

Competition

Male Male

New Zealand[42] New Zealand[42]

None None

Review of injuries at various levels

Male Male Male Female

New Zealand[34] Australia[12,43] England[3] None

None None None None

ª 2010 Adis Data Information BV. All rights reserved.

Sports Med 2010; 40 (2)

King et al.

166

Worldwide, the majority of rugby league participants are amateur (i.e. they derive income from another source), therefore the majority of injuries that occur from rugby league participation are incurred by these participants, but the rate of injury differs for professional players. Although there have been many studies reporting both match and training injures in rugby league and conducted at various levels of participation, there are some participation levels, such as masters and school level, where no prospective injury epidemiological studies have been completed. There was also a paucity of studies on amateur and junior levels of participation. While there is an assumption that the results of epidemiological studies on professional participants translate to other cohorts of participants in other countries, these assumptions have yet to be tested.[44] 3. Injury Definitions for Rugby League A fundamental process, and typically the first step for the injury prevention process, is ongoing injury surveillance.[45-48] However, comparison of results from injury surveillance studies is often difficult due to the inconsistencies in the injury definitions utilized[46,47] and the methodologies undertaken,[45,46,49-51] and therefore the results and conclusions obtained can often have important differences.[1,12,13,23,45,46,50,51] To fully understand the extent and nature of rugby league injuries, it is necessary to consider the various injury definitions that have been used for collating and assessing rugby league injuries. The definition of a sports injury is frequently discussed and, to date, there is no universally accepted definition of a sports injury.[49,52-54] Sports injury definitions are typically provided as operational criteria for the recording and reporting of injuries rather than as a theoretical definition.[55] These definitions are usually broadly based around the concept that ‘‘bodily damage caused by a transfer or absence of energy’’ is what causes injuries to occur.[55] This concept is useful in clarifying whether an incident in rugby league should be recorded as an injury. Several team sports (cricket,[47] football/soccer[56] and rugby union[55]) have published consensus statements in ª 2010 Adis Data Information BV. All rights reserved.

an attempt to obtain more consistent and comparable results from studies undertaken in these sporting activities. Studies on the incidence of injuries in rugby league are no different and have suffered from inconsistencies in the injury definitions utilized (tables II–IV).[43,51] Variations reported for injury incidence are often the result of data obtained from relatively small numbers of players and teams,[3,14] and often over a limited time frame.[43] Some studies on rugby league injuries (as shown in tables II–IV) have utilized criteria of a missed match as an injury definition,[2,3,9,21,29,38,58-60] total injury incidences[8,13,15,22,26-28,31-33,35-37,39-42,57,61] or a combination of both time-loss and non-timeloss injuries,[8,13,22,26-28,32,35,39,41,52,57,61,62] while other studies have incorporated a medical treatment injury definition.[22,24] Consequently, the definition of an injury in rugby league has been widely discussed and disputed[52-54] and to date there is no uniformly accepted definition. The key issue in establishing the injury definition has been centred on which injuries should actually be recorded. Studies have reported that up to 90% of sports injuries have been recorded as non-time-loss injuries.[63] This has been similar in some studies in rugby league where non-time-loss injuries have been reported to be between 85%[57] and 93%[28] of match injuries, and up to 98%[28] of training injuries. The inclusion of non-time-loss injuries has been reported to bias the reported data towards non-time-loss injuries and, as a result of this bias, head and neck injuries make up a large proportion of the total injuries in rugby league.[43] The use of a time-loss definition directs attention to those injuries that are most likely to have direct consequences on players, their team and their club performance;[64] however, the use of this definition understates the effects that non-timeloss injuries have on the healthcare system.[14] The time-loss definition also contains certain inaccuracies, such as an amateur player who trains only twice a week has a greater likelihood of recovering from an injury before the next training session than a subelite or elite level player who trains daily.[65] An injured player may also participate in a training session but the participation Sports Med 2010; 40 (2)

Match and Training Injuries in Rugby League

167

Table II. Professional rugby league injury definitions used and the resulting average injury rate and rates for matches and training Study

Level of participation

Total injury rate

Missed match/ training injury rate

Injury definition used

Seward et al.[22]

Professional (8 teams, 3 grades, 1 year)

139 per 1000 player hours

44 per 1000 player hours

Injury that caused a player to be unavailable for selection in a match, or participation in a training session or any other injury that required specific medical treatment, other than routine conservative measures

Gibbs[2]

Professional (1 club, 3 grades over 3 years)

–

45 per 1000 player game hours

That occurring during a game that caused a player to miss a subsequent match

Estell et al.[57]

Professional (2 grades) and elite junior (4 teams) over 1 season

243 per 1000 player game hours

34 per 1000 player game hours

Pain, discomfort or disability arising during or immediately after, and as a result of, playing in a rugby league match

Stephenson et al.[15]

Professional (1 club, 2 grades over 4 seasons)

114 per 1000 hours of match play

HodgsonPhillips et al.[13]

Professional (1 team over 4 seasons)

346 per 1000 player hours

39 per 1000 player hours

Pain, discomfort, disability or illness (new or recurrent) that the player acknowledged after participating in a rugby league-related activity/game

Gissane et al.[3]

Professional (pooled data from 4 prospective studies: Seward et al.,[22] Gibbs,[2] Stephenson et al.,[15] Gissane et al.[17])

–

40 per 1000 player game hours

Injury requiring a player to miss the subsequent game

Gissane et al.[21]

Professional (1 club, 9 seasons)

–

46 per 1000 player hours

A physical impairment received during a competitive match that prevented a player from being available for selection to play in the next game

Orchard[58]

Professional (1 club, 2 grades over 6 seasons)

–

40 per 1000 player hours

Injury requiring a player to miss a subsequent game

O’Connor[59]

Professional (100 players from 13 clubs)

–

2 per 1000 player hours

An injury was recorded if there was: (a) pain and tenderness in the adductors or at the adductor bonetendon junction; (b) pain and weakness on resisted adduction; and (c) the player was unable to complete training or a game and missed the next training session

Orchard et al.[24]

Professional State of Origin from 2000 to 2006 (1 team, 3 games a year)

139 per 1000 player hours

44 per 1000 player hours

A match injury recurrence (for the State of Origin team) was defined as an injury to the same body part that had been medically assessed prior to the start of the match and which caused the player to miss subsequent games for his club after the Origin match

The onset of pain or a disability that occurred while playing rugby league football

– indicates no data available.

level may be restricted by the injury or they may be undergoing a modified training session.[65] Recent attempts by researchers to establish a standard definition for rugby league injuries have been difficult,[52,53] especially as some researchers were not in favour of a definition that was allencompassing and enabled non-time-loss injuries to be recorded. The following definition of rugby league injury was eventually agreed to by all ª 2010 Adis Data Information BV. All rights reserved.

but two of the six researchers, and is therefore suggested for use in forthcoming studies: ‘‘Any pain or disability that occurs during participation in a rugby league match or training activities that is sustained by a player, irrespective of the need for match or training time loss or for first aid or medical attention. An injury that results in a player requiring first aid or medical attention is referred to as a ‘medical attention injury’ and an Sports Med 2010; 40 (2)

King et al.

168

Table III. Semiprofessional rugby league injury definitions used and the resulting average injury rate and rates for matches and training Study

Level of participation

Total injury incidence

Missed match/training injury incidence

Injury definition used

Gabbett[27]

Semiprofessional (60 players over 1 season)

27 per 1000 training hours

9 per 1000 training hours

Any pain or disability suffered by a player that was subsequently assessed by the head trainer during a training session or immediately after the training session

Gabbett[8]

Semiprofessional (156 players over 2 seasons)

825 per 1000 playing hours; 45 per 1000 training hours

68 per 1000 playing hours; 1 per 1000 training hours

Any pain or disability suffered by a player during a match or training session, and subsequently assess by the head trainer during or immediately following the match or training session

Gabbett[31]

Sub-elite (220 players, 3 years)

78 – 157 per 1000 training hours

19–33 per 1000 training hours

Any pain or disability suffered by a player that was subsequently assessed by the head trainer during, or immediately following the training session

Gabbett[28]

Semiprofessional (79 players, 1 year)

106 per 1000 training hours; 917 per 1000 playing hours

2 per 1000 training hours; 65 per 1000 playing hours

Any pain or disability suffered by a player during a match or training session, and subsequently assessed by the head trainer during or immediately after the match or training session

Gabbett[9]

Semiprofessional (156 players, 2 seasons)

–

68 per 1000 playing hours

Any pain or disability suffered by a player during a match that resulted in the player missing a subsequent match

Gabbett[29]

Sub-elite (1 sub-elite club over three competitive seasons [a = unlimited interchange, b = limited interchange])

–

(a) 73 per 1000 playing hours; (b) 51 per 1000 playing hours

Any pain, disability or injury that occurred as a result of a competition game that caused the player to miss a subsequent game

Gabbett et al.[60]

Semiprofessional (1 club, 153 players over 4 years)

–

55 per 1000 playing hours

Any pain, disability or injury that occurred as a result of a competition match that caused a player to miss a subsequent match

King et al.[26]

Semiprofessional (8 teams, 240 players, 1 year)

115 per 1000 playing hours

78 per 1000 playing hours

Any pain or disability suffered by a player during a match that required advice and/or treatment

– indicates no data available.

injury that results in the player being unable to partake in full part of future training and/or match activities is referred to as a ‘time loss’ injury.’’ [52] An advantage of the all-encompassing injury definition enables comparison between rugby union[55] and soccer,[56] as the injury definitions are similar. The disadvantage is that there are possibly more transient/non-missed match injuries recorded in rugby league and compliance with a broad injury definition may be limited. This was addressed with the recommendation that both total injuries (all injuries recorded) and injuries that result in time loss/missed match/ training be reported, enabling inter-study comparisons to be undertaken in future research. For further information on the debate regarding ª 2010 Adis Data Information BV. All rights reserved.

definitions of injury please see Orchard and Hoskins[54] and Hodgson et al.[53] 3.1 Definition of Injury Exposure and Injury Rate for Rugby League

There is no set format for data collection for sports participation, although the reporting of injury incidence in sports is becoming more standardized, enabling comparison of results between sporting codes[66] and different environments (e.g. training, appearances and competition). Studies involving all levels of rugby league participation[1,12] have reported injury rates (tables II–IV) using both a denominator (number of athletes, games, appearances) and a numerator Sports Med 2010; 40 (2)

Match and Training Injuries in Rugby League

169

(exposure measure),[67,68] expressed, for example, as injury rate per 1000 playing hours. To calculate the injury risk exposure hours for a rugby league team, the number of players (13; NP) on the field at any time is multiplied by the game duration (80 minutes, or £1.33 hours at different participation levels; GD). The result is 17.3 player exposure hours per team per game. Game injury risk exposure hours for the team are calculated by multiplying the player exposure hours per team per game by the number of games

(NG); e.g. 23 games per season gives 13NP · 1.3GD · 23NG = 398 game injury risk exposure hours.[8,12,27,28,33,40,62,67-70] 3.2 Definition of Injury Severity for Rugby League

Assessment of sports injury severity is another aspect that has also not achieved consensus in the literature. A recommendation for classification of injury severity has been proposed that relates

Table IV. Amateur rugby league injury definitions used and the resulting average injury rate and rates for matches and training Study

Level of participation

Total injury incidence

Missed match/training injury incidence

Injury definition used

Norton et al.[42]

Amateur (24 teams over 1 season)

25 per 1000 hours of play; 0.03 per 1000 hours of training

–

Injury occurring during a match or training, for which medical attention was sought, or the player was unable to attend or take part in training or a match

Pringle et al.[61]

Amateur (1730 players age 6–15 y over 1 season)

25 per 1000 player hours

10 per 1000 player hours

A minor injury was defined as one where the player was still in discomfort immediately after the game, but was able to play the following week. A moderate injury was defined as one that prevented the player from participating in the following week’s game

Raftery et al.[37]

Amateur (253 junior teams over 1 year)

10 per 1000 playing hours

–

Any incident that required medical or paramedical review, missed participation at one training session or non-participation in one game

Gabbett[33]

Amateur (9 teams over 3 seasons)

161 per 1000 game hours

–

Injury that was subsequently assessed by the head trainer during or immediately after the match

Gabbett[40]

Amateur (168 players, 3 rugby league sevens tournaments)

284 per 1000 playing hours

–

Any pain or disability suffered by a player that was subsequently assessed by the head trainer during or immediately after a rugby league sevens match

King et al.[41]

Rugby league sevens (semiprofessional and amateur)

498 per 1000 playing hours

262 per 1000 playing hours

Any pain or disability experienced by a player during a match that required advice and/or treatment

King[39]

Junior (3 teams in under 16 and 1 team in under 18 competition)

217 per 1000 playing hours

129 per 1000 playing hours

Any pain or disability experienced by a player during a match that required advice and/or treatment

King et al.[35]

Amateur (1 team 50 players, 1 year)

701 per 1000 playing hours

194 per 1000 playing hours

Any physical or medical condition that occurred during participation in a rugby league match that required medical treatment or resulted in missed match participation

King et al.[32]

Amateur women’s tournament (5 teams over 3 days)

307 per 1000 playing hours

176 per 1000 playing hours

Any pain or disability experienced by a player during a match that required advice and/or treatment

King et al.[36]

Amateur training (1 team, 50 players, 1 year)

22 per 1000 training hours

17 per 1000 training hours

Any physical or medical condition that occurred during participation in rugby league training activities that required medical treatment or resulted in missed training participation

Gabbett[38]

Junior rugby league (80 players over four competitive seasons)

–

57 per 1000 playing hours

Any pain or disability experienced by a player during a match that resulted in the player missing a subsequent match

– indicates no data available.

ª 2010 Adis Data Information BV. All rights reserved.

Sports Med 2010; 40 (2)

King et al.

170

Table V. Injury severity classifications for rugby league Study

Injury severity non-missed participation

minor

moderate

major

King et al.[26,36,41]

Transient (0 games/training missed)

1 game/training week missed

2–4 games/training weeks missed

‡5 games/training weeks missed

Hodgson-Phillips et al.[14]

Transient (0 games missed)

1 game missed

2–4 games missed

‡5 games weeks missed

King[39]

Transient (0 games/training missed)

1 game/training week missed

2–4 games/training weeks missed

‡5 games/training weeks missed

Gabbett[8,12,27,33,38,40]

Transient (0 games/training missed)

1 game/training week missed

2–4 games/training weeks missed

‡5 games/training weeks missed

Self-care by participant

Healthcare professional

Assessed at a hospital ‡5 games or weeks missed

Stevenson et al.[72] Hodgson-Phillips et al.[13]

Transient (0 games missed)

Sandelin et al.[73] Gibbs[2]

Transient (0 games missed)

1 game missed

2–4 games missed

<1 week

1–4 weeks

‡4 weeks

1 game missed

2–4 games missed

‡5 games or weeks missed

– indicates no data available.

the injury severity to the amount of missed match or participation time as a result of the injury.[65] Although this classification has been used in the identification of injury severity, the time-loss difference has varied. For example, a minor injury has been classified as a loss of participation in sporting activities of 1–7 days,[65] yet another time-loss classification for minor injuries has been up to 28 days.[71] The commonly used definitions for injury severity in rugby league studies (table V) have been: 1. Transient (no games or training lost); 2. Minor (one game or training week missed); 3. Moderate (two to four games or training weeks missed); and 4. Major (five or more games or training weeks lost). In a recent endeavour to standardize the definitions for injury severity in rugby league,[52] two of the six researchers again did not agree with the inclusion of transient injuries.[53,54] Although time-loss is identified as the ‘gold standard’ in reporting rugby league injuries, transient injuries still create an impact on the financial resources of teams and participation of players.[14,74] Previous epidemiological injury studies have identified that non-time-loss injury incidence can often account for 72–95% of the total injuries that occur.[63] These injuries, despite not directly affecting the player’s participation in matches, are important because they have a direct and indirect economic ª 2010 Adis Data Information BV. All rights reserved.

impact[14] through areas such as lost employment and associated rehabilitation costs. The following definition of injury severity in rugby league was eventually agreed to by all but two of the six researchers and is therefore suggested for use in forthcoming studies: ‘‘Transient (no matches/training weeks missed), Minor (one missed match/training week), Moderate (two to four missed matches/training weeks), or Major (five or more missed matches/ training weeks).’’[52] For the purpose of studies in rugby league, a transient injury is defined as ‘‘any injury that causes a player to seek medical or first aid treatment during or after a rugby league activity but does not lead to loss of further participation or non-selection for matches’’.[52] These include injuries that are ongoing but not of sufficient severity to prevent the player from participating in match activities or being selected for match participation. 3.3 Classification of Injury Site and Injury Type for Rugby League

Ideally, the classification of injuries should not only include type and location of the injury but also differentiate between trauma-induced injuries and those that occur from overuse or overexposure to a causative agent.[65] For example, Van Mechelen et al.[45] suggested that an Sports Med 2010; 40 (2)

Match and Training Injuries in Rugby League

acute type injury is caused by a single event that causes macro-trauma. If the injury is caused as a consequence of exposure to repetitive microtraumas then it should be classified as an overuse injury. Literature specific to rugby league has categorized injuries according to both site using anatomical location[12,33,37,40,51,69,75,76] and type (table VI),[1,12,27,31,33,37,40,51,62,69,76-78] but whether the injuries are acute or overuse has not been explicitly stated. Suggested injury type definitions for rugby league are provided in table VI. 4. Match versus Training Rugby League Injuries 4.1 Injury Rates for Rugby League Matches and Training

The majority of injuries occur in the match environment, with rates typically increasing as the playing level increases.[12] Junior rugby league injury rates ranged from 1[37] to 197[39] per 1000 playing hours and increased proportionately with the participation level.[37] Amateur match injury rates were slightly higher, ranging from 134 to 701 per 1000 playing hours,[1,2,12,15-18,21,27,35,75] while semiprofessional participation injury rates were higher still, ranging from 115[26,79] to 825[8,12] Table VI. Injury type definitions for rugby league Concussion: A traumatic injury to the brain as a result of a violent blow, shaking or spinning resulting in a transient neurological dysfunction Dislocation: Complete dislocation of a joint (also termed a luxation). A partial dislocation is a subluxation. Dislocations result from trauma Sprain: An injury to a ligament that results from overuse or trauma. Sprains occur when there is a stretch or tear in one or more ligaments (slightly elastic bands of tissue that keep the bones in place while permitting movement at a joint) Bruise/haematoma or ‘contusion’: A traumatic injury of the soft tissues resulting in breakage of local capillaries and leakage of red blood cells. It can be seen as a reddish-purple discoloration that does not blanch when pressed upon Fracture: A break in bone or cartilage, usually the result of trauma. Fractures are classified according to their character and location. Stress fractures result from overuse Laceration: A cut or any break in the skin integrity caused by trauma Strain: An injury to a tendon or muscle resulting from overuse, trauma or overexertion Other: All other medical conditions not incorporated with any of the definitions previously identified

ª 2010 Adis Data Information BV. All rights reserved.

171

per 1000 playing hours. However, professional level injury rates were less than for semiprofessional, ranging from 58[80] to 211[57] per 1000 playing hours. Only a few studies have reported training injuries in rugby league,[1,8,12,13,16,27,28] where injury rates (12–89 per 1000 training hours) were reported to be less than match injuries.[1,12] No studies have been undertaken on junior rugby league training injuries.[12] Amateur training injury rates have been reported to be 22 per 1000 training hours,[36] while semiprofessional/sub-elite training injury rates have been reported to occur at between 27 and 89 per 1000 training hours.[1,8,27] Professional training injury rates are reported to be less, at 12 per 1000 training hours.[13] Some studies indicated a reduction in training injury rates with an alteration in training methods,[1,28,31] and the majority of training injuries were reported as transient (44 per 1000 training hours).[62] 4.2 Incidence of Severe Injuries in Rugby League Matches and Training

Severe injuries occurring in rugby league matches have been identified as those causing the player to miss five or more subsequent matches[1,3,12,13,27,31,33,37,51,60,62,69] and have been associated with players’ biological maturity, playing intensity and skills.[37,75] Approximately 16–30% of all rugby league injuries have been reported as severe, resulting in players missing five or more matches; the injury rate is therefore high.[1,2,8,13,60,69] These high rates of severe injury place demands upon other team members and, if the player returns to playing too early, places them at an increased risk of further injuries.[60] Amateur match severe injuries have been reported to occur at an injury rate of 27 per 1000 playing hours.[69] Recognizing that a rugby league match is 35 playing hours in duration (13NP · 1.33GD · 2NT · 1NG, where NT is number of teams), a severe match injury could occur approximately once in every match. Semiprofessional severe match injury rates have been recorded at 68 per 1000 playing hours,[1,8] while in professional rugby league, severe match injury rates have ranged from 34 to 52 per 1000 playing Sports Med 2010; 40 (2)

King et al.

172

hours,[1-3,13,16,22,57] indicating there is the risk of a severe match injury every game. Differences in the reporting of match severe injury rates between the different participation levels may be due to the unavailability of specialized medical staff at the event or through the club facilities.[1,8] Amateur teams may not have direct access to medical services,[1,69] whereas semiprofessional teams may have a physiotherapist and/or team physician present when they play.[8,12] At the professional level of participation, all teams have direct and full access to specialized rehabilitation services through their club.[69] Another reason for the lower injury rates in professional teams could be the under-reporting of injuries due to the pressure to return to play,[8,12] especially if players risk losing their place in the team or if a financial reward occurs as a result of their participation.[8] A further aspect of match severe injuries is the socioeconomic impact.[1,12,69,81] Significant longterm career limitations, medical costs and loss of income have been associated with major match injuries.[1,12,69,81] Gabbett[69] identified the mean respective costs associated with severe match in-

juries as $A75 ( year value 2001) [medical expenses] and $A205 (wages lost) per playing injury. Meir et al.[81] also reported long-term job limitations, loss of income and medical costs associated with professional rugby league-related match injuries, but did not specify a figure. Severe injuries in training are uncommon,[12] with amateur training severe injury rates reported to be 17 injuries per 1000 training hours while semiprofessional training rates were one injury per 1000 training hours over 2 years.[1,8,27] Professional training was similar, with one injury per 1000 training hours reported.[13] However, there were no reported training-related severe injuries in junior or amateur rugby league.[1,12,60] 4.3 Site and Type of Rugby League Injuries during Matches and Training

Early research in rugby league identified that ligament and joint injuries (54%) were the common injuries[2] and occurred primarily to the knee (24%).[2] More recently, studies have shown a change in anatomical sites being injured at all levels of participation (table VII), where the

Table VII. Site, type and cause of common rugby league match injuries Study

Playing level

Injury site

Injury type

Alexander et al.[82]

Professional

Head and neck

Haematomas and strains

Alexander et al.[83]

Professional

Head and neck

Contusions

Gibbs[2]

Professional

Knee

Ligament and joint

Gissane et al.[3]

Professional

Lower limb

Gissane et al.[16]

Professional

Head and neck

Haematomas and strains

Being tackled

Gissane et al.[17]

Professional

Head and neck

Haematomas and strains

Being tackled

Gissane et al.[18]

Professional

Head, neck, knee and shoulder

Joint sprains

Being tackled

Hodgson-Phillips et al.[13]

Professional

Knee

Haematomas and strains

Seward et al.[22]

Professional

Head and neck

Lacerations and contusions

Stephenson et al.[15]

Professional

Head and neck

Haematomas and strains

Being tackled

King et al.[26]

Semiprofessional

Shoulder

Haematomas and strains

Being tackled

Gabbett[8]

Semiprofessional

Thigh and calf

Haematomas and strains

Being tackled

King et al.[35]

Amateur

Thigh and lower leg

Haematomas and strains

Being tackled

King et al.[32]

Amateur women

Knee and lower leg

Haematomas and strains

Physical collisions

Gabbett[33]

Amateur

Head and neck

Haematomas and strains

Gabbett[69]

Amateur

Arm and hand

Joint sprains

While tackling

Gabbett[38]

Junior

Shoulder

Sprains

Being tackled

King[39]

Junior

Knee and shoulder

Sprains and strains

Being tackled

Raftery et al.[37]

Junior

Knee

Fractures

Being tackled

ª 2010 Adis Data Information BV. All rights reserved.

Injury cause

Sports Med 2010; 40 (2)

Match and Training Injuries in Rugby League

common injury sites were to the head and neck.[1,8,12,13,15,17,22,28,33,40,62,68,82,83] The changes in injury site were thought to be due to changes in the match rules in recent years (i.e. defensive line back to 10 m, ball stripping in the tackle).[12,33] Recently published research has identified that the shoulder is now the common reported injury site.[26,38,79] One study on semiprofessional players[8] reported that haematomas (271 per 1000 playing hours) and injuries to the calf and thigh (168 per 1000 playing hours) were the common type and site of injuries.[1,8] This is in conflict with another semiprofessional study where strains (28 per 1000 playing hours) and injuries to the shoulder (16 per 1000 playing hours) were the common injury type and site of injuries reported.[26,79] Injuries to other anatomical sites were higher for the Australian than the New Zealand studies (face 115[1,8] vs 9[26,79] per 1000 playing hours; arm and hand 115[1,8] vs 13[26,79] per 1000 playing hours; knee 109[1,8] vs 15[26,79] per 1000 playing hours; and head and neck 104[1,8] vs 14[26,79] per 1000 playing hours). Junior rugby league injuries differed in injury site.[1,37] Although the lower limb (knee 14% and ankle 13%)[1,37] was the frequent injury region reported previously, the shoulder (28%) has now been reported to be the common injury site.[38] Injuries to the head and neck occurred in 11% of all junior amateur injuries,[1,37] and fractures have become common, especially in players aged 6–17 years.[1,37] Changes in defensive strategies, levels of participation and anthropometric aspects of players have been attributed to the differences in injury sites.[12,33] Lower limbs incur the majority of injuries in training for semiprofessional players,[1,8,13,27] with the calf and thigh the commonly recorded injury site.[1,8,12,13,27] Lower limb strain injuries have been reported to be directly related to an increased need for players to be able to rapidly accelerate, decelerate and change direction.[1,12,84] The type of training injuries recorded may reflect the emphasis on game-specific skills and increased playing intensity required.[1,8,12,28,31,60] The majority of semiprofessional training injuries (91 per 1000 training hours) occurred in traditional conditioning activities such as running and ª 2010 Adis Data Information BV. All rights reserved.

173

sprints,[27] while skill-based conditioning games (ball handling, inter-team games) incurred less injuries (9 per 1000 training hours).[1,27] The common training injury type for semiprofessional participants was muscular strains (137 per 1000 training hours).[8,27] Other injury causes were overuse (20 per 1000 training hours), fall or stumble (14 per 1000 training hours) and collision with another player (14 per 1000 training hours).[28] Gissane et al.[16] reported similar findings for professional rugby league players in the UK, but there has been no published research on training injuries in the National Rugby League (Australia and New Zealand) to enable comparison. Professional injuries were similar to semiprofessional injuries,[1,2,12] with the most frequent injuries (38 per 1000 playing hours) being to the head and neck.[15] Other anatomical areas (thigh and calf 20 per 1000 playing hours; knee 12 per 1000 playing hours; thorax and abdomen 10 per 1000 playing hours) occurred less frequently.[2,12,15] Muscular injuries were the most frequent injury type (34 per 1000 playing hours),[15] while joint sprains (27 per 1000 playing hours),[15] lacerations (20 per 1000 playing hours)[15,22] and muscle strains (18 per 1000 playing hours)[15] occurred less frequently. 4.4 Cause of Rugby League Injuries during Matches and Training

Studies have identified that the tackle is the major cause of injuries in rugby league matches.[1,8,15-18,37,51,60,69,80] Between 46% and 90% of all injuries that occurred in matches were tackle related,[1,8,15-18,37,51,69,80] with all concussions recorded resulting from the tackle.[1,12] However, high rates of concussion could be expected given each player was involved in an average of 41 physical collisions per match.[16,17,19,20,51] Research into amateur rugby league also identified the tackle as the most frequent cause of injuries,[33,35,40,69] with injury rates from the tackle reported to be as high as 538 per 1000 playing hours.[35] Injuries were more prominent to the ball carrier (405 per 1000 playing hours) than the tackler (133 per 1000 playing hours).[35] Junior Sports Med 2010; 40 (2)

King et al.

174

rugby league had a lower injury rate but showed similar prominence as in amateur rugby league.[37,38] Research into semiprofessional matches also demonstrated that injuries occurred frequently during the tackle (95[26,79] to 382[1,8] per 1000 playing hours). The tackler sustained more injuries than the ball carrier (27%[1,8] to 44%[26,79] vs 20%[1,8] to 39%[26,79], respectively), which is the reverse of the trend seen in amateur[8] and professional rugby league[15,17] where the ball carrier sustained more injuries (46 per 1000 playing hours) than the tackler (21 per 1000 playing hours).[15,17] Injuries resulting from the tackle in professional matches increased when the competition format was changed from a winter to a summer competition,[13,21] where the rate of tackled player injuries increased from 16 to 30 per 1000 playing hours.[13,21] The injury rate for tackling players also increased from 6 to 14 per 1000 playing hours.[13,21] 4.5 Type of Player Injured during Matches and Training: Forwards versus Backs

Early research on rugby league injuries identified that there were no differences between injury rates for forwards and backs during matches.[2,17] However, more recent research identified forwards have a higher injury rate than backs,[1,4,12,15,60,81,82,85] which possibly reflects the greater involvement of forwards in physical collisions.[1,4,15,60,81,82,85] Forwards recorded 55–139 injuries per 1000 playing hours for tacklerelated injuries,[15,17,18,20] while backs had a slightly lower tackle-related injury rate (29–93 per 1000 playing hours).[15,17,20] Injury rates have also been studied in defensive and attacking roles in a professional competition.[20] In attack, the injury rates of forwards were higher than backs (16 per 1000 playing hours vs 13 per 1000 playing hours)[20] and their injury rate increased substantially when in a defensive role (forwards 39 per 1000 playing hours vs backs 16 per 1000 playing hours).[20] With seasonal changes, positional injury rates changed for both forwards (winter 35 vs summer 68 per 1000 playing hours) and backs (winter 26 vs summer 54 per 1000 playing hours).[1,2] ª 2010 Adis Data Information BV. All rights reserved.

Participation in other levels of rugby league (semiprofessional and amateur) has also shown higher injury rates in forwards than backs.[1,8,12,33,40,69] Forwards also had higher rates of head, neck, face and knee injuries than backs in all competition levels,[12,33] whereas backs had a higher injury rate in the ankles and ‘other’ site categories than forwards.[12,33] Studies of semiprofessional players identified that forwards had a higher incidence of training injuries compared with backs,[1,8,12] with an injury rate of 53 per 1000 training hours compared with 38 per 1000 training hours for backs.[1,8,12] Forwards also had a higher injury rate to the head and neck, shoulder, thigh and calf, ankle and foot compared with backs.[1,8] During the early part of the season, the forward injury rate was 70 per 1000 training hours, while backs recorded an injury rate of 48 per 1000 training hours.[1,8] Injury rates decreased for all players by late season, with forwards recording 35 injuries per 1000 training hours, while the backs recorded 28 injuries per 1000 training hours.[1,8] The common training injury types were muscular strains and overuse injury.[1,8] Forwards recorded more muscular strains (24 per 1000 training hours) than backs (15 per 1000 training hours)[1,8] and more overuse injuries (12 per 1000 training hours) than backs (6 per 1000 training hours).[1,8] However, there was no position-specific training injury research for amateur, junior or professional rugby league.[1,8] 4.6 Time of Injury during Matches and Training

There were limited studies of the timing of injuries during matches.[12] Gabbett[33] reported significantly more injuries in the second (71%) than the first half (29%) in amateur competition,[1,33] which was reversed in semiprofessional games,[1,8,12] with more injuries in the first than the second half of matches (1014 per 1000 playing hours [62%] vs 636 per 1000 playing hours [38.5%]).[1,12] Professional competition was reportedly similar to semiprofessional competition in that most,[1,22] but not all,[26] studies reported more first half injuries (57%). There have been no Sports Med 2010; 40 (2)

Match and Training Injuries in Rugby League

junior rugby league studies evaluating injury occurrence in relation to time of injury.[1,12] The only study on the time of injury during training was for semiprofessional rugby league.[1,8] Teams often integrate skill and conditioning sessions in the later stages of the training session to simulate game-related conditions, promoting skill development under fatigued conditions.[12,62] Injury occurrence has been reported to be higher in the later stages of the training session,[1,8,12,28] a finding that may be related to fatigue.[1,12,62] 4.7 Seasonal Variations for Match and Training Injuries

Traditionally, the rugby league season runs from late summer through to early spring.[13,18,21] In the southern hemisphere this is from April to September,[69] while in the northern hemisphere this is from August to April.[13,17,21] Early research undertaken in Australia[2,22,57] reported a higher injury rate than that reported in England,[15,16,18] and this was thought to be due to harder surfaces and higher temperatures.[15,21] In 1996 the northern hemisphere competition season changed, requiring players to have a shortened winter season before a compressed summer season.[13,18,21] This change exposed players to higher temperatures and harder grounds in conditions similar to those experienced in the southern hemisphere.[13,18,21] As a result, the injury rate doubled from 367 (winter) to 617 (summer) per 1000 playing hours.[13] The type of injures remained the same but the risk to the tackler had increased 2-fold.[13,21] Traditional winter rugby league injury rates have been reported to fluctuate throughout the season,[1,2,12,15-18,21,27,33,75] with amateur match injuries reportedly occurring more in the latter half of the season.[1,12,33] In one study[33] the injury rate was 134 per 1000 playing hours at the start of the season (March), declined in April but then increased progressively from May to September, culminating at 196 per 1000 playing hours[33] at the end of the season. The increasing injury rate was attributed to player fatigue and accumulative microtrauma,[1,12,33] while other factors reported to have an influence on injury ª 2010 Adis Data Information BV. All rights reserved.

175

rates were environmental changes in weather conditions and changes in ground hardness near the end of the season.[1,12,86] Studies of semiprofessionals have identified a similar trend in injury rates, increasing from 561 to 1339 per 1000 playing hours, that was consistent for all playing positions.[8] The increasing injury rate was attributed to playing intensity as the final series approached[1,8] – a concept supported by Gabbett[28] who reported a significant correlation (r = 0.74) between match injury rates and match intensity.[1,28] Studies of professional rugby league players have also identified similar trends[12,13] of increased injury rates as the season progressed.[13,83] Injury severity also increased when the northern hemisphere season changed to a shortened winter season before a compressed summer season. Other studies have identified a slight variation with more injuries at the beginning of the season,[16,22] with a decrease midseason before rising towards the final series.[16] During the training pre-season period (December to April) in Australia, the injury rate was 157 per 1000 training hours in 2001,[1,31] which reduced to 78 per 1000 training hours in 2003 as a result of changing training methods from traditional conditioning activities to game-specific skill-based conditioning games.[1,31] Injury rates were noted to increase from December to March[1,31] with the highest injury rate recorded yearly in February, with an average of 142 per 1000 training hours.[1,31] This rate decreased in March, prior to the season starting, and was identified as the result of an alteration in the training intensity.[1,31] Training injury rates through the southern hemisphere season (April to September) have been recorded to be highest early in the season (116 per 1000 training hours[1,8,12]), which is reportedly three times higher than the season average rate of 45 per 1000 training hours.[1,8,12] The training injury rate decreased as the training intensity decreased throughout the season.[1,8] The high injury rate at the beginning of the season was significantly correlated (r = 0.86)[12] to the increased training loads,[1,8,12] which suggests that more injuries occurred when players trained harder.[1,8,12,28] Professional training injuries were recorded at 12 per 1000 training hours[13] Sports Med 2010; 40 (2)

King et al.

176

and a similar injury rate for training injuries over a season (1990–1) was reported for a professional team in England.[16] However, studies of professional rugby league[13,16] participation have not identified details of the type of training injuries. 5. Conclusions This systematic review of rugby league injuries reported the extent of injuries that occur in all levels of participation for matches and training. The majority of injuries occurred in the match environment with injury rates increasing as playing level increased. Early research in rugby league identified that ligament and joint injuries were common for match injuries, head and neck injuries were common during senior rugby league, and knee injuries were common at junior levels of participation. Recent research reported the shoulder as the common injury site in matches, with lower limb muscular strains common for training injuries. Muscular injuries were common injury types for senior rugby league players while fractures were common for junior levels of participation. Being tackled as the ball carrier was the common cause for injury at all levels of participation in matches. Forwards were more likely to be injured in the tackle and had higher rates of head, neck, face and knee injuries than backs in all match and training competition levels. More injuries occurred in the second than the first half of amateur matches but this was reversed in semiprofessional matches. Fatigue was reported to influence the injury incidence in both match and training environments, suggesting that changes in playing and training intensity may have contributed to injury. Although comparisons between different participation levels have been reported in the literature, there is a paucity of studies looking at the different divisions in these participation levels (i.e. Division 1 and Division 2 amateur participation) and the injury incidence. Studies utilizing single clubs or teams over a short period provide a snap shot of the injury incidence but do not address the breadth of participation; nor do they reflect the different conditions seen in the various countries where rugby league is played. Also not ª 2010 Adis Data Information BV. All rights reserved.

addressed in the literature is how much fitness is a protective factor when comparing injury incidence at the different levels of participation. Ongoing review of the different levels of participation enables an epidemiological perspective to be established of the injuries in rugby league. The change in the injury site and type reported in the research literature reflects the changing match and training styles seen in the modern game of rugby league. Changes in injury site and type could be used to prompt further research and development of injury reduction programmes to readdress the issue of injuries that occur as a result of participation in rugby league activities. Further research is warranted at all participation levels of rugby league in both the match and training environments to confirm the strongest risk factors for injury. In particular, specific research is warranted on the tackle, which is the leading cause of injuries reported in rugby league in both the match and training environments. Sports medicine practitioners should be cognizant of the extent and nature of rugby league injuries and work with coaches, athletes and trainers to prevent these injuries occurring. Acknowledgements No sources of funding were used to assist in the preparation of this review. The authors have no conflicts of interest that are directly relevant to the content of this review.

References 1. Gabbett TJ. Science of rugby league football: a review. J Sports Sci 2005; 23 (9): 961-76 2. Gibbs N. Injuries in professional rugby league: a threeyear prospective study of the South Sydney Professional Rugby League Football Club. Am J Sports Med 1993; 21 (5): 696-700 3. Gissane C, Jennings D, Kerr K, et al. A pooled data analysis of injury incidence in rugby league football. Sports Med 2002; 32 (3): 211-6 4. Meir R, Arthur D, Forrest M. Time and motion analysis of professional rugby league: a case study. Strength Cond Coach 1993; 1 (3): 24-9 5. Coutts A, Reaburn P, Abt G. Heart rate, blood lactate concentration and estimated energy expenditure in a semiprofessional rugby league team during a match: a case study. J Sports Sci 2003; 21: 97-103 6. Brewer J, Davis J. Applied physiology of rugby league. Sports Med 1995; 20 (3): 129-35

Sports Med 2010; 40 (2)

Match and Training Injuries in Rugby League

7. Gabbett TJ, Jenkins D. Applied physiology of rugby league. Sports Med 2008; 38 (2): 119-38 8. Gabbett TJ. Incidence of injury in semi professional rugby league players. Br J Sports Med 2003; 37 (1): 36-44 9. Gabbett TJ. Influence of playing position on the site, nature and cause of rugby league injuries. J Strength Cond Res 2005; 19 (4): 749-55 10. Meir R, Newton R, Curtis E, et al. Physical fitness qualities of professional rugby league football players: determination of positional differences. J Strength Cond Res 2001; 15 (4): 450-8 11. Clark L. A Comparison of the speed characteristics of elite rugby league players by grade and position. Strength Cond Coach 2002; 10: 2-12 12. Gabbett TJ. Incidence of injury in junior and senior rugby league players. Sports Med 2004; 34 (12): 849-59 13. Hodgson Phillips L, Standen PJ, et al. Effects of seasonal change in rugby league on the incidence of injury. Br J Sports Med 1998; 32 (2): 144-8 14. Hodgson L, Standen PJ, Batt ME. An analysis of injury rates after seasonal change in rugby league. Clin J Sports Med 2006; 16 (4): 305-10 15. Stephenson S, Gissane C, Jennings D. Injury in rugby league: a four year prospective survey. Br J Sports Med 1996; 30 (4): 331-4 16. Gissane C, Jennings DC, Standing P. Incidence of injury in rugby league football. Physio 1993; 79: 305-10 17. Gissane C, Jennings DC, Cumine AJ, et al. Differences in the incidence of injury between rugby league forwards and backs. Aust J Sci Med Sport 1997; 29 (4): 91-4 18. Gissane C, Jennings D, White J, et al. Injury in summer rugby league football: the experiences of one club. Br J Sports Med 1998; 32 (2): 149-52 19. Gissane C, White J, Kerr K, et al. Physical collisions in professional rugby league: the demands on different player positions. Clev Med J 2001; 4: 137-46 20. Gissane C, Jennings D, Jennings S, et al. Physical collisions and injury rates in professional super league rugby: the demands of different player positions. Clev Med J 2001; 4: 147-55 21. Gissane C, Jennings D, Kerr K, et al. Injury rates in rugby league football: impact of change in playing season. Am J Sports Med 2003; 31 (6): 954-8 22. Seward H, Orchard J, Hazard H, et al. Football injuries in Australia at the elite level. Med J Aust 1993; 159 (5): 298-301 23. Orchard J, Seward H. Epidemiology of injuries in the Australian football league: seasons 1997-2000. Br J Sports Med 2002; 36 (1): 39-45 24. Orchard JW, Hoskins W. Rugby league injuries at State of Origin level [online]. Available from URL: http://www. injuryupdate.com.au/images/research/Origininjuries20002006. pdf [Accessed 2009 Dec 1] 25. Gabbett TJ. Influence of physiological characteristics on selection in a semi-professional first grade rugby league team: a case study. J Sports Sci 2002; 20 (5): 399-406 26. King DA, Gabbett TJ. Injuries in the New Zealand semiprofessional rugby league competition. NZ J Sports Med 2009; 36 (1): 6-15

ª 2010 Adis Data Information BV. All rights reserved.

177

27. Gabbett TJ. Training injuries in rugby league: an evaluation of skill-based conditioning games. J Strength Cond Res 2002; 16 (2): 236-41 28. Gabbett TJ. Influence of training and match intensity on injuries in rugby league. J Sports Sci 2004; 22: 409-17 29. Gabbett TJ. Influence of the limited interchange rule on injury rates in sub-elite rugby league players. J Sci Med Sport 2005; 8 (1): 111-5 30. Gabbett TJ, Domrow N. Relationships between training load, injury, and fitness in sub-elite collision sport athletes. J Sports Sci 2007; 25 (13): 1507-19 31. Gabbett TJ. Reductions in pre-season training loads reduce training injury rates in rugby league players. Br J Sports Med 2004; 38 (6): 743-9 32. King DA, Gabbett TJ. Injuries in a national women’s rugby league tournament: an initial investigation. NZ J Sports Med 2007; 34 (2): 18-22 33. Gabbett TJ. Incidence, site, and nature of injuries in amateur rugby league over three consecutive seasons. Br J Sports Med 2000; 34 (2): 98-103 34. Lythe MA, Norton RN. Rugby league injuries in New Zealand. NZ J Sports Med 1992; 20: 6-7 35. King DA, Gabbett TJ. Amateur rugby league match injuries in New Zealand. NZ J Sports Med 2009; 36 (1): 16-21 36. King DA, Gabbett TJ. Training injuries in New Zealand amateur rugby league players. J Sci Med Sport 2008; 11 (6): 562-5 37. Raftery M, Parker R, Stacey E, et al. Incidence of injury in junior rugby league in the Penrith and district rugby league area. Westmead: Children’s Hospital Institute of Sports Medicine, Research and Development Office, The New Children’s Hospital, 1999: 1-22 38. Gabbett TJ. Incidence of injury in junior rugby league players over four competitive seasons. J Sci Med Sport 2007; 11 (3): 323-8 39. King DA. Incidence of injuries in the 2005 New Zealand national junior rugby league competition. NZ J Sports Med 2006; 34 (1): 21-7 40. Gabbett TJ. Incidence of injury in amateur rugby league sevens. Br J Sports Med 2002; 36 (1): 23-7 41. King DA, Gabbett TJ, Dreyer C, et al. Incidence of injuries in the New Zealand national rugby league sevens tournament. J Sci Med Sport 2006; 9 (1-2): 110-8 42. Norton R, Wilson M. Rugby league injuries and patterns. NZ J Sports Med 1995; 22: 37-8 43. Hoskins W, Pollard H, Hough K, et al. Injury in rugby league. J Sci Med Sport 2006; 9 (1-2): 46-56 44. Brooks JHM, Kemp SPT. Recent trends in rugby union injuries. Clin J Sports Med 2008; 27 (1): 51-73 45. van Mechelen W, Hlobil H, Kemper HCG. Incidence, severity, aetiology and prevention of sports injuries: a review of concepts. Sports Med 1992; 14 (2): 82-99 46. Ekstrand J, Karlsson J. The risk for injury in football: there is a need for a consensus about definition of the injury and the design of studies. Scand J Med Sci Sports 2003: 13 (3): 147-9 47. Orchard JW, Newman D, Stretch R, et al. Methods for injury surveillance in international cricket. Br J Sports Med 2005; 39 (4): e22

Sports Med 2010; 40 (2)

178

48. Finch C. A new framework for research leading to sports injury prevention. J Sci Med Sport 2006; 9 (1): 3-9 49. Brooks JHM, Fuller CW. The influence of methodological issues on the results and conclusions from epidemiological studies of sports injuries: illustrative examples. Sports Med 2006; 36 (6): 459-72 50. Finch CF. An overview of some definitional issues for sports injury surveillance. Sports Med 1997; 24 (3): 157-98 51. Gissane C, White J, Kerr K, et al. Health and safety implications of injury in professional rugby league football. Occ Med 2003; 53: 512-7 52. King DA, Gabbett TJ, Gissane C, et al. Epidemiological studies of injuries in rugby league: suggestions for definitions, data collection and reporting methods. J Sci Med Sport 2009; 12 (1): 12-9 53. Hodgson L, Gissane C, Gabbett TJ, et al. For debate: consensus injury definitions in team sports should focus on encompassing all injuries. Clin J Sports Med 2007; 17 (3): 188-91 54. Orchard J, Hoskins W. For debate: consensus injury definitions in team sports should focus on missed match playing time. Clin J Sports Med 2007; 17 (3): 192-6 55. Fuller CW, Molloy MG, Bagate C, et al. Consensus statement on injury definitions and data collection procedures for studies of injuries in rugby union. Clin J Sports Med 2007; 17 (3): 177-81 56. Fuller CW, Ekstrand J, Junge A, et al. Consensus statement on injury definitions and data collection procedures in studies of football (soccer) injuries. Br J Sports Med 2006; 40 (3): 193-201 57. Estell J, Shenstone B, Barnsley L. Frequency of injuries in different age-groups in an elite rugby league club. Aust J Sci Med Sport 1995; 27 (4): 95-7 58. Orchard J. Missed time through injury and injury management at an NRL club. Sports Health 2004; 22 (1): 11-9 59. O’Connor D. Groin injuries in professional rugby league players: a prospective study. J Sports Sci 2004; 22: 629-36 60. Gabbett TJ, Domrow N. Risk factors for injury in subelite rugby league players. Am J Sports Med 2005; 33 (3): 428-34 61. Pringle RG, McNair P, Stanley S. Incidence of sporting injury in New Zealand youths aged 6-15. Br J Sports Med 1998; 32 (1): 49-52 62. Gabbett TJ. Influence of injuries on team playing performance in rugby league. J Sci Med Sport 2004; 7 (3): 340-6 63. Powell JW, Dompier TP. Analysis of injury rates and treatment patterns for time-loss and non-time-loss injuries among collegiate student athletes. J Ath Train 2004; 39 (1): 56-70 64. Timpka T, Ekstrand J, Svanstro¨m L. From sports injury prevention to safety promotion in sports. Sports Med 2006; 36 (9): 733-45 65. Junge A, Dvorak J. Influence of definition and data collection on the incidence of injuries in football. Am J Sports Med 2000; 28 (5 Suppl.): S40-6 66. Schootman M, Powell JW, Torner JC. Study designs and potential biases in sports injury research: the case-control study. Sports Med 1994; 18 (1): 22-37

ª 2010 Adis Data Information BV. All rights reserved.

King et al.

67. Chambers RB. Orthopedic injuries in athletics (ages 9 to 17): comparison of injuries occurring in six sports. Am J Sports Med 1979; 7: 195-7 68. Hodgson Phillips L. Sports injury incidence. Br J Sports Med 2000; 34 (2): 133-6 69. Gabbett TJ. Severity and cost of injuries in amateur rugby league: a case study. J Sports Sci 2001; 19 (5): 341-7 70. Hodgson Phillips L. Methodology in research. In: MacAuley D, Best T, editors. Evidence-based sports medicine. Cornwall: BMJ Publishing Group, 2002: 12-28 71. Garraway WM, McLeod D. Epidemiology of rugby football injuries. Lancet 1995; 345: 1485-7 72. Stevenson MR, Hamer P, Finch CF, et al. Sport, age, and sex specific incidence of sports injuries in Western Australia. Br J Sports Med 2000; 34 (3): 188-94 73. Sandelin J, Santavirta S, La¨ttila¨ R, et al. Sports injuries in a large urban population: occurrence and epidemiological aspects. Int J Sports Med 1987; 9 (1): 61-6 74. Knowles SB, Marshall SW, Guskiewicz KM. Issues in estimating risks and rates in sports injury research. J Ath Train 2006; 41 (2): 207-15 75. Gabbett TJ. Physiological characteristics of junior and senior rugby league players. Br J Sports Med 2002; 36 (5): 334-9 76. Takarada Y. Evaluation of muscle damage after a rugby match with special reference to tackle plays. Br J Sports Med 2003; 37 (5): 416-9 77. Gibbs N. Common rugby league injuries: recommendations for treatment and preventative measures. Sports Med 1994; 18 (6): 438-50 78. Addley K, Farren J. Irish rugby injury survey: Dungannon football club. Br J Sports Med 1988; 22 (1): 22-4 79. King DA. Injuries in the New Zealand Bartercard cup competition. Dunedin: University of Otago, 2007 80. Walker RD. Sports injuries: rugby league may be less dangerous than union. Practitioner 1985 Mar; 229 (1401): 205-6 81. Meir R, McDonald KN, Russell R. Injury consequences from participation in professional rugby league: a preliminary investigation. Br J Sports Med 1997; 31 (2): 132-4 82. Alexander D, Kennedy M, Kennedy J. Injuries in rugby league football. Med J Aust 1979; 2: 341-2 83. Alexander D, Kennedy M, Kennedy J. Rugby league football injuries over two competitive seasons. Med J Aust 1980; 2: 334-45 84. Meir R. Evaluating players’ fitness in professional rugby league: reducing subjectivity. Strength Cond Coach 1993; 1 (4): 11-7 85. Larder P. The rugby league coaching manual. 2nd ed. London: Kingswood Press, 1992 86. Orchard J. Is there a relationship between ground and climatic conditions and injuries in football? Sports Med 2002; 32 (7): 419-32

Correspondence: Mr Doug King, Emergency Department, Hutt Valley District Health Board, Private Bag 31-907, Lower Hutt, New Zealand. E-mail: [email protected]

Sports Med 2010; 40 (2)

CORRESPONDENCE

Sports Med 2010; 40 (2): 179-182 0112-1642/10/0002-0179/$49.95/0

ª 2010 Adis Data Information BV. All rights reserved.

Maximal Lactate Steady-State Prediction The evaluation of blood lactate response to exercise tests provides several physiological- and performance-related indexes, such as lactate turn-points (two turning points may be observed during incremental exercise tests) and the maximal lactate steady-state (MLSS) concept. These indexes have been the subject of extensive research. Several exercise physiologists have addressed questions related to their physiological bases as well as to the effects of exercise test protocols on the measurement of such variables. In addition, these lactate-based indexes have also been employed in numerous human and animal studies as a tool for assessing exercise performance and/or training effects. Most initial questions regarding the biological bases of these indexes and their utilization as an evaluation tool in exercise science were addressed by German scientists. Approximately 30 years ago, the German groups led the research in this field and the results of their ongoing work have contributed notably to generations of exercise scientists to date. Kindermann’s group is certainly one of the pioneering groups in this research area. Despite the amount of work done and the reviews published on this topic (lactate-based indexes), the recent review by Faude et al.[1] gathers important data and draws a wide overview of the several lactate threshold (LT) concepts that have emerged in the literature since the 1970s. In their review, Faude et al.[1] evaluated the validity of LT concepts with regard to assessing aerobic endurance capacity or prescribing training intensity (the latter as the agreement between MLSS work-rate [MLSSWR] and the elected LTs [LTAn]). The authors accurately indicated that the blood lactate concentration (bLa) at MLSSWR varies considerably among individuals (values from 2 up to 10 mM). Based on these data, they suggest that LTAn determination should be performed by individualized approaches rather than using fixed bLa threshold concepts. Later in their review, the authors state that fixed bLa threshold

concepts do not take into account interindividual differences and that LT4 may underestimate (particularly in anaerobically trained athletes) or overestimate (in aerobically trained athletes) the MLSSWR. Alternatively, the authors present individualized concepts for LTAn determination (some of these fixed and individualized concepts have been evaluated in the literature and are cited in their review). Their criticisms of fixed bLa thresholds based on interindividual variations in bLa at MLSSWR seem simplistic and may not take into account all of the data in the literature.[2,3] The classical study by Heck et al.[2] shows, in table II, individual values of MLSSWR and their bLa, as well as the bLa corresponding to MLSSWR as identified during incremental exercise. If a correlation test is performed between MLSS bLa and the incremental exercise bLa corresponding to MLSSWR, no relationship is observed between these two variables. This lack of relationship indicates instead that interindividual variations in MLSS bLa may not preclude the validity of fixed bLa thresholds, at least under exercise test conditions similar to those of Heck et al.[2] Nevertheless, the above relationship may indeed exist under slightly different test conditions (lower increment rate) or exercise mode (see Figueira et al.[3] for an overview). Some referenced individualized LTAN approaches[4-8] were previously found to be invalid or their study designs were inappropriate for evaluation of the validity of LTAN. However, this issue is addressed further in another section of their review. The belief that aerobic training status influences the validity of bLa fixed thresholds has been stated since the 1980s,[4,9,10] but without direct experimental evidence. On the contrary, the only available paper[11] on this matter (i.e. aerobic fitness level vs fixed bLa threshold validity) shows that aerobic fitness level does not affect the validity of fixed threshold (onset of blood lactate accumulation [OBLA] 3.5 mM) to predict MLSSWR. It is also known that aerobic training status is not related to bLa at MLSSWR,[11,12] which places two of the authors’ rationales in conflict with one another. The authors stated that interindividual variations in bLa at MLSSWR may preclude the validity of fixed thresholds; later it is stated that training

180

Letter to the Editor

status would also preclude the validity of fixed thresholds (if the first statement is true, the latter would be in disagreement, since bLa at MLSSWR is not related to training status). It is recommended, and in fact it is recognized by the authors, that the validity of a given method should be analysed by comparing it with the gold standard measurement (in this case, the direct and individual determination of MLSSWR). Faude et al.[1] also appropriately indicated the importance of the Bland and Altman[13] statistical approach in order to assess the validity of a candidate concept. However, in the authors’ initial discussion regarding MLSSWR prediction by LTAN, they included many studies whose experimental design or statistical approaches are questionable. Actually, these studies differ from those selected to be in their table V and could have been classified as such (i.e. questionable experimental design). The authors meant that table V includes all of the 11 studies that used the ‘recommended procedure’ to evaluate the validity of LTAN. Unfortunately, they overlooked some recent studies[3,11,14-17] that would fit the so-called ‘recommended procedure’ and support the aim of their review. However, two of these studies[16,17] might have appeared in PubMed after the preparation of their review. Tiago R. Figueira,1 Herbert G. Simo˜es2 and Benedito S. Denadai3 1 Faculty of Medical Sciences, State University of Campinas, Campinas, Sao Paulo, Brazil 2 Catholic University of Brasilia, Taguatinga Distrito Federal, Brazil 3 Human Performance Laboratory, Sa˜o Paulo State University, Rio Claro, Sao Paulo, Brazil

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

Acknowledgements 17.

The authors have no conflict of interest that are directly relevant to the content of this letter.

References 1. Faude O, Kindermann W, Meyer T. Lactate threshold concepts: how valid are they? Sports Med 2009; 39 (6): 469-90 2. Heck H, Mader A, Hess G, et al. Justification of the 4 mmol/L lactate threshold. Int J Sports Med 1985; 6 (3): 117-30 3. Figueira TR, Caputo F, Pelarigo JC, et al. Influence of exercise mode and maximal lactate-steady-state concentration on

ª 2010 Adis Data Information BV. All rights reserved.

the validity of OBLA to predict maximal lactate-steady-state in active individuals. J Sci Med Sport 2008; 11 (3): 280-6 Stegmann H, Kindermann W. Comparison of prolonged exercise tests at the individual anaerobic threshold and the fixed anaerobic threshold of 4 mmol/L lactate. Int J Sports Med 1982; 3 (2): 105-10 Beneke R. Anaerobic threshold, individual anaerobic threshold, and maximal lactate steady state in rowing. Med Sci Sports Exerc 1995; 27 (6): 863-7 Jones AM, Doust JH. The validity of the lactate minimum test for determination of the maximal lactate steady state. Med Sci Sports Exerc 1998; 30 (8): 1304-13 Tegtbur U, Busse MW, Braumann KM. Estimation of an individual equilibrium between lactate production and catabolism during exercise. Med Sci Sports Exerc 1993; 25 (5): 620-7 Simo˜es HG, Campbell CS, Kushnick MR, et al. Blood glucose threshold and the metabolic responses to incremental exercise tests with and without prior lactic acidosis induction. Eur J Appl Physiol 2003; 89 (6): 603-11 Keul J, Simon G, Berg A, et al. Bestimmung der individuellen anaeroben Schwelle zur Leistungsberwertung und Trainingsgestaltung. Dtsch Z Sportmed 1979; 30: 212-8 Simon G, Berg A, Dickhuth H-H, et al. Bestimmung der anaeroben Schwelle in Abhangibkeit von Alter un von der Leistungsfanhigkeit. Dtsch Z Sportmed 1981; 32: 7-14 Denadai BS, Figueira TR, Favaro OP, et al. Effect of the aerobic capacity on the validity of the anaerobic threshold for determination of the maximal lactate steady state in cycling. Braz J Med Biol Res 2004; 37 (10): 1551-6 Beneke R, Hutler M, Leithauser RM. Maximal lactatesteady-state independent of performance. Med Sci Sports Exerc 2000; 32 (6): 1135-9 Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1: 307-10 Denadai BS, Gomide EBG, Greco CC. The relationship between onset of blood lactate accumulation, critical velocity, and maximal lactate steady state in soccer players. J Strength Cond Res 2005; 19 (2): 364-8 Pardono E, Sotero RC, Hiyane W, et al. Maximal lactate steady-state prediction through quadratic modeling of selected stages of the lactate minimum test. J Strength Cond Res 2008; 22 (4): 1073-80 Sotero RC, Padorno E, Campbell CS, et al. Indirect assessment of lactate minimum and maximal blood lactate steady-state intensity for physically active individuals. J Strength Cond Res 2009; 23 (3): 847-53 Sotero RC, Padorno E, Landwehr R, et al. Blood glucose minimum predicts maximal lactate steady state on running. Int J Sports Med 2009; 30 (9): 643-6

The Authors’ Reply We appreciate the interest of Figueira and colleagues in our recent review[1] of lactate threshold (LT) concepts. A lot of research has Sports Med 2010; 40 (2)

180

Letter to the Editor

status would also preclude the validity of fixed thresholds (if the first statement is true, the latter would be in disagreement, since bLa at MLSSWR is not related to training status). It is recommended, and in fact it is recognized by the authors, that the validity of a given method should be analysed by comparing it with the gold standard measurement (in this case, the direct and individual determination of MLSSWR). Faude et al.[1] also appropriately indicated the importance of the Bland and Altman[13] statistical approach in order to assess the validity of a candidate concept. However, in the authors’ initial discussion regarding MLSSWR prediction by LTAN, they included many studies whose experimental design or statistical approaches are questionable. Actually, these studies differ from those selected to be in their table V and could have been classified as such (i.e. questionable experimental design). The authors meant that table V includes all of the 11 studies that used the ‘recommended procedure’ to evaluate the validity of LTAN. Unfortunately, they overlooked some recent studies[3,11,14-17] that would fit the so-called ‘recommended procedure’ and support the aim of their review. However, two of these studies[16,17] might have appeared in PubMed after the preparation of their review. Tiago R. Figueira,1 Herbert G. Simo˜es2 and Benedito S. Denadai3 1 Faculty of Medical Sciences, State University of Campinas, Campinas, Sao Paulo, Brazil 2 Catholic University of Brasilia, Taguatinga Distrito Federal, Brazil 3 Human Performance Laboratory, Sa˜o Paulo State University, Rio Claro, Sao Paulo, Brazil

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

Acknowledgements 17.

The authors have no conflict of interest that are directly relevant to the content of this letter.

References 1. Faude O, Kindermann W, Meyer T. Lactate threshold concepts: how valid are they? Sports Med 2009; 39 (6): 469-90 2. Heck H, Mader A, Hess G, et al. Justification of the 4 mmol/L lactate threshold. Int J Sports Med 1985; 6 (3): 117-30 3. Figueira TR, Caputo F, Pelarigo JC, et al. Influence of exercise mode and maximal lactate-steady-state concentration on

ª 2010 Adis Data Information BV. All rights reserved.

the validity of OBLA to predict maximal lactate-steady-state in active individuals. J Sci Med Sport 2008; 11 (3): 280-6 Stegmann H, Kindermann W. Comparison of prolonged exercise tests at the individual anaerobic threshold and the fixed anaerobic threshold of 4 mmol/L lactate. Int J Sports Med 1982; 3 (2): 105-10 Beneke R. Anaerobic threshold, individual anaerobic threshold, and maximal lactate steady state in rowing. Med Sci Sports Exerc 1995; 27 (6): 863-7 Jones AM, Doust JH. The validity of the lactate minimum test for determination of the maximal lactate steady state. Med Sci Sports Exerc 1998; 30 (8): 1304-13 Tegtbur U, Busse MW, Braumann KM. Estimation of an individual equilibrium between lactate production and catabolism during exercise. Med Sci Sports Exerc 1993; 25 (5): 620-7 Simo˜es HG, Campbell CS, Kushnick MR, et al. Blood glucose threshold and the metabolic responses to incremental exercise tests with and without prior lactic acidosis induction. Eur J Appl Physiol 2003; 89 (6): 603-11 Keul J, Simon G, Berg A, et al. Bestimmung der individuellen anaeroben Schwelle zur Leistungsberwertung und Trainingsgestaltung. Dtsch Z Sportmed 1979; 30: 212-8 Simon G, Berg A, Dickhuth H-H, et al. Bestimmung der anaeroben Schwelle in Abhangibkeit von Alter un von der Leistungsfanhigkeit. Dtsch Z Sportmed 1981; 32: 7-14 Denadai BS, Figueira TR, Favaro OP, et al. Effect of the aerobic capacity on the validity of the anaerobic threshold for determination of the maximal lactate steady state in cycling. Braz J Med Biol Res 2004; 37 (10): 1551-6 Beneke R, Hutler M, Leithauser RM. Maximal lactatesteady-state independent of performance. Med Sci Sports Exerc 2000; 32 (6): 1135-9 Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1: 307-10 Denadai BS, Gomide EBG, Greco CC. The relationship between onset of blood lactate accumulation, critical velocity, and maximal lactate steady state in soccer players. J Strength Cond Res 2005; 19 (2): 364-8 Pardono E, Sotero RC, Hiyane W, et al. Maximal lactate steady-state prediction through quadratic modeling of selected stages of the lactate minimum test. J Strength Cond Res 2008; 22 (4): 1073-80 Sotero RC, Padorno E, Campbell CS, et al. Indirect assessment of lactate minimum and maximal blood lactate steady-state intensity for physically active individuals. J Strength Cond Res 2009; 23 (3): 847-53 Sotero RC, Padorno E, Landwehr R, et al. Blood glucose minimum predicts maximal lactate steady state on running. Int J Sports Med 2009; 30 (9): 643-6

The Authors’ Reply We appreciate the interest of Figueira and colleagues in our recent review[1] of lactate threshold (LT) concepts. A lot of research has Sports Med 2010; 40 (2)

Letter to the Editor

been conducted in this area throughout the last 30 years and a large scientific debate has emerged. Therefore, we were not surprised that our paper has provoked some comments. The article aimed at evaluating LTs with regard to their linear relationship with (simulated) endurance performance and their similarity to the maximal lactate steady state (MLSS). Figueira et al.[2] raised several concerns, in particular with regard to blood lactate concentrations [bLa] at the maximal lactate steady state (MLSS) versus at the anaerobic threshold (LTAn) and with regard to the selection of articles. We would like to comment on their major points and hopefully clarify these issues. First, it is stated that we have overlooked some recent studies on this topic. This is true (and we regret that) in the case of the paper by Denadai et al.[3] who evaluated the effect of aerobic capacity on the maximal lactate steady state. However, in contrast to the opinion of Figueira et al.,[2] the study of Jones and Doust[4] has been cited and discussed as one of the 11 papers evaluating the validity of LTAn. A further relevant study[5] was published at the time that we finished the review. Unfortunately, to date (3 November 2009) this paper has not appeared in PubMed under the search terms used for the review article (‘lactate threshold’ or ‘anaerobic threshold’ in combination with ‘maximal lactate steady state’). This example clearly demonstrates the importance of selecting reasonable titles and key words to increase the probability that a publication can be detected. However, both studies are well within the presented data and support the conclusion that scientific data regarding the validity of fixed lactate threshold markers are very heterogeneous. Further mentioned and probably relevant studies were not included in our review because they appeared after the final preparation and publication of the manuscript. Another point made by Figueira et al.[2] is that the ‘‘criticism of fixed bLa thresholds based on interindividual variations in bLa at MLSSWR seem simplistic and may not take into account all of the data in the literature.’’[5,6] As stated above, Figueira et al.[5] did not appear in PubMed under the used search terms. Notably, this study conª 2010 Adis Data Information BV. All rights reserved.

181

cluded that the validity of a fixed LT to predict MLSS is dependent on exercise mode (cycling vs running) and that the differences observed for cycling and running exercise are related to individual variations in MLSS. Thus, our conclusion is supported by those data. The reference Heck et al.[6] has been cited several times and its conclusion has been taken into account (together with several other relevant papers). We agree that [bLa] during constant exercise at the MLSS may vary considerably between subjects and scientific evidence does not exist that they differ between trained and untrained individuals. Therefore, it does not seem possible to predict [bLa] at MLSS in different individuals. That is, when referring to a fixed [bLa] reached during constant load or incremental exercise, for instance 4 mmol/L, it is not possible to definitely say if the subject is below, at or above the MLSS. In addition, many other factors, for example preceding exercise, nutrition, the testing protocol or methodological influences on absolute lactate levels (as has been pointed out in our manuscript) have to be considered when interpreting absolute lactate concentrations in performance diagnosis or training control. Thus, the criticism of fixed LTs is based on a variety of points and is far from being simplistic. It has also been claimed that two of our rationales were in conflict with each other, namely the statements ‘‘that interindividual variations in bLa at MLSSWR may preclude the validity of fixed thresholds’’ and ‘‘that training status would also preclude the validity of fixed thresholds.’’[2] We are convinced that these comments are mainly based on confusion regarding the [bLa] response to steady state versus incremental exercise. In daily practice, it is important to estimate the MLSS from one single incremental exercise test (GXT [graded incremental exercise test]). The use of a fixed [bLa] as a basis for this estimation is questionable because of the above outlined reasons. In addition, validation studies obtained very heterogeneous results for fixed LTs.[1,3,5] In some individuals the use of the 4 mmol/L marker (obtained during a GXT) may underestimate and in others overestimate real MLSS. In our experience, the latter is predominant in aerobically Sports Med 2010; 40 (2)

182

trained subjects, whereas in anaerobically trained subjects the real MLSS tends to occur at higher [bLa] during GXT. This view is shared by other working groups (as Figueira et al.[2] have correctly stated) and is substantiated by some indirect evidence. The individual anaerobic threshold (IAT) according to Stegmann et al.[7] in several studies has been shown to estimate the maximal lactate steady state well (studies conducted independently by different national and international institutions).[1,8] This is true at least for cycling and running on a group level (as well as in most individuals). A common finding in our laboratory (and in others) is that the [bLa] at the IAT (measured during GXT) is lower in highly endurance trained subjects compared with more anaerobically or untrained subjects. For instance, such data are reported in the original work of Stegmann et al.[7] with average [bLa] at IAT of 2.1 mmol/L (ranging from 1.4 to 3.0 mmol/L in long-distance runners), 3.9 mmol/L (range 2.0–5.7 mmol/L, handball players), and 4.6 mmol/L (range 3.0–7.5 mmol/L, sports students). IAT velocity was clearly highest in the long-distance runners. Similar data have been published several times on national German conferences in the 1980s. However, no conclusive data have been published about that issue yet. At first sight, this might sound contradictory to the above-mentioned observations from steadystate MLSS exercise. However, there is evidence that endurance training affects lactate production as well as lactate removal rates during exercise.[9,10] Thus, lactate kinetics during incremental exercise might be influenced in a way that lactate increase will be slower after training (less production and/or quicker removal). This may have an influence on [bLa] after 3 or 5 minutes’ (step duration) but not necessarily on overall [bLa] during longer steady-state exercise. Although it remains speculative, there might be a good explanation for this observed phenomenon and, hence, our rationales are not in conflict with each other.

ª 2010 Adis Data Information BV. All rights reserved.

Letter to the Editor

However, we agree that there is definitely a need for further research and hope that such research will contribute to an improved understanding of the maximal lactate steady state, lactate thresholds and lactate kinetics during exercise. Oliver Faude, Wilfried Kindermann and Tim Meyer Institute of Sports and Preventive Medicine, University of Saarland, Saarbru¨cken, Germany

Acknowledgements The authors have no conflicts of interest that are relevant to the content of this manuscript.

References 1. Faude O, Kindermann W, Meyer T. Lactate threshold concepts: how valid are they? Sports Med 2009; 39: 469-90 2. Figueira TR, Simoes HG, Denadai BS. Maximal lactate steady-state prediction. Sports Med 2010; 40: 179-80 3. Denadai BS, Figueira TR, Favaro OR, et al. Effect of the aerobic capacity on the validity of the anaerobic threshold for determination of the maximal lactate steady state in cycling. Braz J Med Biol Res 2004; 37: 1551-6 4. Jones AM, Doust JH. The validity of the lactate minimum test for determination of the maximal lactate steady state. Med Sci Sports Exerc 1998; 30: 1304-13 5. Figueira TR, Caputo F, Pelarigo JG, et al. Influence of exercise mode and maximal lactate-steady-state concentration on the validity of OBLA to predict maximal lactate-steady-state in active individuals. J Sci Med Sport 2008; 11: 280-6 6. Heck H, Mader A, Hess G, et al. Justification of the 4-mmol/L lactate threshold. Int J Sports Med 1985; 6: 117-30 7. Stegmann H, Kindermann W, Schnabel A. Lactate kinetics and individual anaerobic threshold. Int J Sports Med 1981; 2: 160-5 8. Urhausen A, Coen B, Weiler B, et al. Individual anaerobic threshold and maximum lactate steady state. Int J Sports Med 1993; 14: 134-9 9. Messonnier L, Freund H, Denis C, et al. Effects of training on lactate kinetics parameters and their influence on short high-intensity exercise performance. Int J Sports Med 2006; 27: 60-6 10. Bergman BC, Wolfel EE, Butterfield GE, et al. Active muscle and whole body lactate kinetics after endurance training in men. J Appl Physiol 1999; 87: 1684-96

Sports Med 2010; 40 (2)