The Validity of Self-Reported Physical Fitness Test Scores

Transcription

1 VOLUME 172 FEBRUARY 2007 NUMBER 2 ORIGINAL ARTICLES Authors alone are responsible for opinions expressed in the contribution and for its clearance through their federal health agency, if required. MILITARY MEDICINE, 172, 2:115, 2007 The Validity of Self-Reported Physical Fitness Test Scores Guarantor: Joseph J. Knapik, ScD Contributors: Sarah B. Jones, MPH ; Joseph J. Knapik, ScD*; Marilyn A. Sharp, MS ; Salima Darakjy, MPH ; Bruce H. Jones, MD* Epidemiological studies often have to rely on a participant s self-reporting of information. The validity of the self-report instrument is an important consideration in any study. The purpose of this investigation was to determine the validity of self-reported Army Physical Fitness Test (APFT) scores. The APFT is administered to all soldiers in the U.S. Army twice a year and consists of the maximum number of push-ups completed in 2 minutes, the maximum number of sit-ups completed in 2 minutes, and a 2-mile run for time. Army mechanics responded to a questionnaire in March and June 2004 asking them to report the exact scores of each event on their most recent APFT. Actual APFT scores were obtained from the soldier s military unit. The mean standard deviation (SD) of actual and self-reported numbers of push-ups was and 65 13, respectively. The mean SD of actual and selfreported numbers of sit-ups were and 68 10, respectively. The mean SD of actual and self-reported run times (minutes) were and , respectively. Correlations between actual and self-reported push-ups, sit-ups, and run were 0.83, 0.71, and 0.85, respectively. On average, soldiers tended to slightly over-report performance on all APFT events and individual self-reported scores could vary widely from actual scores based on Bland-Altman plots. Despite this, the close correlations between the actual and selfreported scores suggest that self-reported values are adequate for most epidemiological military studies involving larger sample sizes. *U.S. Army Center for Health Promotion and Preventive Medicine, 5158 Blackhawk Road, Aberdeen Proving Ground, MD Oak Ridge Institute for Science and Education, P.O. Box 53, Aberdeen Proving Ground, MD U.S. Army Research Institute of Environmental Medicine, Building 42, Kansas Street, Natick, MA Poster presentation at the American College of Sports Medicine 52nd Annual Meeting, June 1 4, 2005, Nashville, TN. The views expressed herein are the views of the author(s) and do not reflect the official policy of the Department of the Army, the Department of Defense, or the U.S. government. This manuscript was received for review in January The revised manuscript was accepted for publication in July Introduction hysical fitness has been defined in many different ways,1 5 P but perhaps the most succinct definition is a set of attributes that relates to one s ability to perform physical activity. 6 The attributes or components of physical fitness include muscular strength, muscular endurance, cardiorespiratory endurance, flexibility, coordination, balance, and body composition. 1,2 There are many laboratory tests that can be used to accurately measure the components of fitness. 6 However, many of these are difficult to apply in large epidemiological studies because they are technically complex, time-consuming, and require maximal, or near maximal, exertion on the part of the subject. Field tests 2 are less complex but still require time to set-up and administer and maximal or near maximal effort on the part of the subject. If valid information on physical fitness could be obtained by questionnaire, this method of data collection might be more practical for epidemiological studies involving large numbers of participants. Valid physical fitness information obtained could also be useful for health care providers in determining fitness levels to assist in assessing and prescribing care for their patients. Several previous studies have examined the validity of obtaining fitness and physical activity information using questionnaire techniques Four studies measured components of physical fitness and used them to validate physical activity questionnaires. 7,8,12,13 However, this form of validation can be criticized because physical activity and physical fitness are distinct (but related) concepts. 1 Two studies did validate responses to questions on specific physical fitness components against actual measures of these components and found modest to high relationships. 9,10 The purpose of the study reported here was to examine the validity of self-reports of physical fitness test scores. Army soldiers were asked by questionnaire for their exact Army Physical Fitness Test (APFT) scores and their responses were compared to their actual scores. The degree to which soldiers are able to 115

2 116 Self-Reported Physical Fitness Test Score Validity accurately report scores can determine the usefulness of selfreporting in future epidemiological military investigations. Methods Subjects Subjects were 88 soldiers (83 men and 5 women) who were serving in the U.S. Army as wheel vehicle mechanics at Fort Bragg, North Carolina. They were involved in a larger prospective study, the purpose of which was to determine injury rates and injury risk factors among the mechanics in March and June Individuals were recruited via a verbal briefing in which they were informed of the procedures and possible risks of participation. Volunteers signed an informed consent statement approved by an institutional review board. Procedures Soldiers completed a questionnaire that asked them to report the date and scores of their most recent APFT. Soldiers were asked: (a) When did you take your last APFT? and (b) What were your most recent raw scores for push-ups, sit-ups, and the 2-mile run? Actual APFT raw scores and the date of the test were obtained from the soldier s unit via a telephone call. The actual APFT raw scores obtained from the soldiers units served as the criterion fitness measures. The APFT consisted of three events: a 2-minute maximal effort push-up event, a 2-minute maximal effort sit-up event, and a 2-mile run for time. For the push-up, the soldier lowered his or her body in a generally straight line to a point where his or her upper arm was parallel to the ground and then returned to the starting point with elbows fully extended. For the sit-up, the soldier s knees were bent at a 90 angle, fingers were interlocked behind the head, and a second person held the soldier s ankles, keeping his or her heels firmly on the ground. The soldier raised his upper body to a vertical position so that the base of the neck was anterior to the base of the spine and then returned to the TABLE I PHYSICAL CHARACTERISTICS OF THE SOLDIERS Variable Men (n 83) Mean SD a Women (n 5) Combined (N 88) Age (years) Height (cm) Weight (kg) Body mass index (kg/m 2 ) starting position. Raw scores were the number of push-ups and sit-ups that were successfully completed within the separate 2-minute time periods. The performance measure for the run was the time taken to complete the 2-mile distance. Time between APFT events was no less than 10 minutes. 14 Soldiers weights and heights were obtained by direct measurement using a digital scale and anthropometer on the day of testing. Age was obtained from the informed consent statement. Body mass index was calculated as weight/height Data Analysis Validity was determined by comparing actual scores to the self-reported scores using Pearson product moment correlation coefficients. To graphically display the relationships, scatter plots were constructed. Paired sample t tests were used to test the hypothesis of no difference between the two sets of scores (self-report and actual). Bland-Altman plots were also constructed. 16 Statistical analyses, scatter plots, and Bland-Altman plots were performed using the Statistical Package for the Social Sciences (SPSS) version Bland-Altman variables and limits of agreement were calculated manually. 16 Time from the actual APFT to the self-report was calculated as the number of days from the last APFT reported by the unit to the date the questionnaire was completed. TABLE II ACTUAL AND SELF-REPORTED APFT SCORES Results A small proportion of the self-reported APFT data was missing. There were 3 missing push-up scores, 1 missing sit-up score, and 13 missing run times. The self-reported APFT scores were missing because soldiers left the question blank on the questionnaire. There was 1 missing actual push-up score, 1 missing actual sit-up score, and 3 missing actual run times. Actual scores were missing because soldiers had medical conditions that excused them from taking a particular APFT event on the date of the test. Table I shows the physical characteristics of the 88 soldiers. The mean SD) time between the actual APFT and the selfreport was days. The actual and self-reported APFT scores are shown in Table II. On average, soldiers reported slightly higher raw scores than they actually achieved. The average difference (official self-reported) between actual and selfreported push-ups, sit-ups, and run were four repetitions, two repetitions, and 0.3 minutes, respectively (p 0.01 for all measures). Note that a higher value for run time denotes a slower and therefore, a lower performance level. Correlations between actual and self-reported APFT scores ranged from 0.71 to Scatter plots with regression lines for push-ups, sit-ups, and the run are in Figures 1, 2, and 3, respectively. Mean SD APFT Event n Actual Raw Scores a Self-Reported Raw Scores a r b Push-ups Sit-ups Mile Run a Units for Push-ups and sit-ups are repetitions, units for the 2-mile run are minutes. b Pearson product moment correlation (all correlations significant at the p 0.01 level).

3 Self-Reported Physical Fitness Test Score Validity 117 Fig. 1. Scatter plot of self-report and actual push-ups. Fig. 2. Scatter plot of self-report and actual sit-ups. Bland-Altman plots for push-ups, sit-ups, and the run are shown in Figures 4, 5, and 6, respectively. These plots offer a graphical presentation of the amount of agreement between the actual and self-reported scores. Perfect agreement only occurs when all points lie along the line of equality, 16 or0onthex-axis. The 95% Bland-Altman limits of agreement were 20 to 12

4 118 Self-Reported Physical Fitness Test Score Validity Fig. 3. Scatter plot of self-report and actual 2-mile run. Fig. 4. Bland-Altman plots for push-ups. Shows difference against mean for push-up data (0 no difference on vertical axis). repetitions for push-ups, 18 to 13 repetitions for sit-ups, and 1.3 to 1.8 minutes for the run. Discussion The correlations between actual and self-reported scores were relatively high, indicating that subjects could provide an acceptably accurate average estimate of their fitness scores. Despite the close relationship, subjects tended to self-report slightly higher performances than they actually accomplished, but the average overestimation was not large. The proportion of soldiers reporting higher than actual performance was 54%, 52%, and 42% for push-ups, sit-ups, and the 2-mile run, respectively. Two previous studies have looked at the associations between physical activity and physical fitness. Siconolfi et al. 7 found a correlation of 0.29 between reported physical activity and maximal oxygen uptake (the latter an indicator of cardiorespiratory endurance). Lamb 8 reported correlations of 0.48 and 0.51 between measures of self-perceived fitness and physical work capacity and correlations of 0.47 and 0.35 between measures of self-perceived fitness and intense leisure time physical activity. These correlations were considerably lower than those found in

5 Self-Reported Physical Fitness Test Score Validity 119 Fig. 5. Bland-Altman plots for sit-ups. Shows difference against mean for sit-up data (0 no difference on vertical axis). Fig. 6. Bland-Altman plots for 2-mile run. Shows difference against mean for run data (0 no difference on vertical axis). the present study. Siconolfi et al. 7 and Lamb 8 were examining relationships between two different constructs, physical fitness and physical activity. Fitness is a set of components related to the ability to perform physical activity, as described in the Introduction. 1 Physical activity is any bodily movement produced by skeletal muscles that results in energy expenditure. 1 Physical activity of the proper mode, frequency, intensity, and duration can lead to or maintain fitness, but fitness and activity are distinct concepts. 17 The higher correlations in the present study may be partly attributed to the fact that physical fitness measures were directly compared. Knapik et al. 9 used objective measures of physical fitness to compare self-ratings of components of physical fitness. However, that study differed from the present one in that subjects were asked to rate their fitness on a 5-point scale for each of the four fitness components. Correlations between questionnaire responses and objective fitness measures (strength tests, maximal oxygen uptake, etc.) ranged from 0.29 to 0.53 (self-ratings of endurance and three measures of aerobic capacity), 0.10 to 0.17 (self-ratings of sprint speed and measures of anaerobic capacity), 0.28 to 0.53 (self-ratings of strength and measures of maximal strength), and 0.30 to 0.48 (self-ratings of flexibility and measures of hip/low back flexibility). Riley et al. 10 examined self-assessed fitness ratings on a 5-point scale and compared them to physical fitness tests among male Marine Corps recruits. A correlation of 0.72 was found between predicted and actual number of pull-ups. We found a somewhat closer relation between self-reported and actual push-ups, another measure of upper body muscular endurance like the pull-ups. Our soldiers had been in the military for a longer period of time and had likely taken more fitness tests than the Marine recruits tested by Riley et al. 10 Our findings are generally in consonance with those of Riley et al. 10 and suggest that both soldiers and Marine recruits can be fairly accurate in recalling, and in the case of the Riley et al. 10 study, predicting, their fitness test scores. The relatively high correlations between actual and self-reported physical fitness scores can be at least partly explained by the importance and repetitiveness of the APFT in the military. APFT scores are often included on efficiency reports and are a

6 120 Self-Reported Physical Fitness Test Score Validity factor in promotions. Soldiers take the APFT twice per year, reinforcing their memory of the scores. Physical training is a job requirement in the military and most soldiers train to maintain a relatively high fitness level, yet do not necessarily train to improve their fitness. Thus, a soldier s APFT score may not change much over testing periods, making it easier to remember. Some soldiers do make an effort to further advance their fitness. These soldiers may better remember their APFT scores because those scores serve as the target for improvement and the scores are periodically reviewed mentally and/or visually. For all of these reasons, the conclusions from this study may be more generalizable to groups that take regular fitness tests. Despite relatively high correlations, there are some other factors to consider. It is possible that a soldier remembered a previous APFT raw score. In this case, the reported score may have been close to the actual score on the last test for reasons mentioned above. However, some soldiers also reported scores that differed widely from the actual score (Figs. 4 6), thus reducing group accuracy. Another factor was the tendency of soldiers to slightly over-report their APFT performances. This could be due to the desire of some individuals to cast themselves in a more complimentary light. Several studies have suggested that individuals will provide questionnaire responses to make themselves look more favorable or less unfavorable However, the average over-reporting was not large, amounting to 4 and 2 repetitions for push-ups and sit-ups (respectively) and 0.3 minutes on the run. Another limitation in this study is sample size. Due to the small sample of women (n 5), we were unable to distinguish differences between the sexes. A final factor is the missing self-reported 2-mile run times. It is not clear why 15% of soldiers did not provide this data but it may suggest some bias on this measure, especially if those that did not answer the question were those that had the most difficulty remembering their score. In epidemiological studies, it is important to have easy to administer, valid means of obtaining information from a population. Despite the limitations mentioned above, this study suggests that among soldiers, a self-report questionnaire can provide a fairly accurate group measure of APFT scores. Individual scores may vary widely from the actual score; however, the close correlations between self-reported and actual scores suggest that, at least in this population, self-reported fitness values may be adequate for most military epidemiological studies involving larger sample sizes. Acknowledgments We appreciate the assistance of the U.S. Army Research Institute of Environmental Medicine (USARIEM) team at Fort Bragg, North Carolina: Major Lolita Burrell; Mr. Richard Carr, Anteon Corp.; Ms. Crystal Ross, Anteon Corp.; and Ms. Sandra Kelley. References 1. Caspersen CJ, Powell KE, Christenson GM: Physical activity, exercise, and physical fitness: definitions and distinctions for health-related research. Public Health Rep 1985; 100: Fleishman EA: The Structure and Measurement of Physical Fitness. Englewood Cliffs, NJ, Prentice Hall, Pate RR: A new definition of youth fitness. Phys Sportsmed 1983; 11: Bouchard C, Shephard RJ, Stephens T, et al: Exercise, fitness and health: the consensus statement. In: Exercise, Fitness, and Health: A Consensus of Current Knowledge. Edited by Bouchard C, Shephard RJ, Stephens T, Sutton JR, McPherson BD. Champaign, IL, Human Kinetics, Gutin B: A model of physical fitness and dynamic health. J Phys Ed Rec 1980; 51: McArdle WD, Katch FI, Katch VL: Exercise Physiology: Energy, Nutrition, and Human Performance. Philadelphia, PA, Lea and Febiger, Siconolfi SF, Lasater TM, Snow RCK, Carleton RA: Self-reported physical activity compared with maximal oxygen uptake. Am J Epidemiol 1985; 122: Lamb KL: Correlates of self-perceived fitness. Percept Mot Skills 1992; 74: Knapik JJ, Jones BH, Reynolds KL, Staab JS: Validity of self-assessed physical fitness. Am J Prev Med 1992; 8: Riley DJ, Wingard D, Morton D, et al: Use of self-assessed fitness and exercise parameters to predict objective fitness. Med Sci Sports Exerc 2005; 37: Washburn RA, Goldfield SRW, Smith KW, et al: The validity of self-reported exercise-induced sweating as a measure of physical activity. Am J Epidemiol 1990; 132: Fogelholm M, Malmberg J, Suni J, et al: International physical activity questionnaire: validity against fitness. Med Sci Sports Exerc 2006; 38: Friedenreich CM, Courneya KS, Neilson HK, et al: Reliability and validity of the past year total physical activity questionnaire. Am J Epidemiol 2006; 163: Physical Fitness Training: Army Field Manual (FM) Washington, DC, Headquarters, Department of the Army, Knapik JJ, Burse RL, Vogel JA: Height, weight, percent body fat, and indices of adiposity for young men and women entering the U.S. Army. Aviat Space Environ Med 1983; 54: Bland JM, Altman DG: Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 8476: Knapik JJ, Zoltick J, Rottner HC, et al: Relationships between self-reported physical activity and physical fitness in active men. Am J Prev Med 1993; 9: Wang PS, Benner JS, Glynn RJ, et al: How well do patients report noncompliance with antihypertensive medications: a comparison of self-report versus filled prescriptions. Pharmacoepidemiol Drug Saf 2004; 13: Chan D, Caprara D, Blanchette P, et al: Recent developments in meconium and hair testing methods for the confirmation of gestational exposures to alcohol and tobacco smoke. Clin Biochem 2004; 37: Palta M, Prineas RJ, Berman R, et al: Comparison of self-reported and measured height and weight. Am J Epidemiol 1982; 115: Niedhammer I, Bugel I, Bonenfant S, et al: Validity of self-reported weight and height in the French GAZEL cohort. Int J Obes Relat Metab Disord 2000; 24: Spencer EA, Appleby PN, Davey GK, et al: Validity of self-reported height and weight in 4808 EPIC-Oxford participants. Public Health Nutr 2001; 5: Eriksson I, Undén AL, Elofsson S: Self-rated health: comparisons between three different measures: results from a population study. Int J Epidemiol 2001; 30: