RILEY, DONNA J.1; WINGARD, DEBORAH2; MORTON, DEBORAH3; NICHOLS, JEANNE F.4; JI, MING1; SHAFFER, RICHARD1; MACERA, CAROLINE A.1
1Graduate School of Public Health, San Diego State University, Naval Health Research Center, San Diego, CA; 2Department of Family and Preventive Medicine and 3Department of Family and Preventive Medicine, University of California San Diego, San Diego, CA; and 4Department of Exercise and Nutritional Sciences, San Diego State University, San Diego, CA
Address for correspondence: Dr. Donna Riley, Cuyamaca College, Department of Exercise Science and Health Education, 900 Rancho San Diego Pkwy., El Cajon, CA 92019; E-mail: email@example.com.
Submitted for publication July 2004.
Accepted for publication January 2005.
The views expressed in this manuscript are those of the authors and do not reflect the official policy or position of the Department of the Navy, the Department of Defense, or the U.S. Government.
This research has been conducted in compliance with all applicable federal regulations governing the protection of human subjects in research.
This research was supported by the Naval Health Research Center under work unit number 60213, funded by the U.S. Navy Bureau of Medicine and Surgery under document number NC 2275 N0001802WR20006.
Various branches of the military have reported low levels of fitness within incoming recruit populations who are expected to begin a rigorous training regimen immediately after arrival at boot camp (4). The inability to pass basic military physical fitness tests because of low fitness levels may lead to injury and contribute to overall attrition in the armed forces (7). This is a particular problem among Marine Corps recruits. Previously identified risk factors for military attrition include lack of previous exercise, low levels of fitness, previous injury, lack of participation in previous exercise of sufficient intensity, higher body mass index (BMI), and other lifestyle factors (4–6,12). However, many of these potential risk factors are difficult to identify before training, and some cannot be modified during a training program. Low fitness is an important predictor of lack of military success at the training level, and understanding how to evaluate this parameter before entry may be useful in assuring that incoming recruits have the best chance of success.
Among the ways to identify low entry fitness is to objectively assess fitness at entry. However, at this point the recruit is already at the site and beginning to train. A simple pen-and-paper test that flags men who may not be successful after the first month of training could save much time and effort. One way to do this is to identify self-assessed fitness and exercise parameters that are associated with objective fitness; previous studies show that this has been done with varying levels of success in other settings (1,6,13). The goal of the present study is to determine the usefulness of self-assessed fitness and exercise parameters in predicting objectively measured physical fitness in U.S. Marine Corps recruits.
The study population consisted of male Marine Corps recruits entering the Marine Corps Recruit Depot (MCRD), San Diego, CA, between September 19 and November 4, 2002. Recruits were asked to complete a questionnaire administered within 3 d of their arrival at MCRD. After approximately 1 month, all recruits participated in a physical fitness test (PFT-1). The final study population consisted of the 1583 recruits who completed a questionnaire and participated in the PFT-1. The study protocol was reviewed and approved by the institutional review board of the Naval Health Research Center, and participants read and signed informed consent documents before completing the questionnaire.
Our sample size calculations used a difference of 13.8 points as our effect size of interest (11). Assuming an alpha level of 0.05 and 80% power, 1530 subjects would be required. A total of 2040 recruits were surveyed to allow for the high dropout rate during the first few weeks of boot camp (about 22.4% in this case), leaving a final sample size of 1583.
Objective fitness, the major outcome variable, was measured by the PFT-1 score. This test consists of abdominal crunches, pull-ups, and a 3-mile run, and is scored as follows: number of abdominal crunches completed in 2 min (1 point each) to a maximum of 100 points, plus number of pull-ups (5 points each for a maximum of 20 pull-ups) to a maximum of 100 points, plus a sliding scale of points associated with time on a 3-mile (4.8-km) run in 18:00 min or less, worth 100 points. The maximum PFT-1 score is 300 points; to pass the fitness test, the recruit must meet a minimum standard for each component and earn a total score of at least 135 points.
Items on the questionnaire included age, race/ethnicity, general demographics, self-assessed fitness (excellent, very good, good, fair, or poor), and physical activity habits before boot camp, such as amount of competitive activity, frequency of sweating during physical activity (indicative of exercise intensity), and predicted number of pull-ups (see Appendix). The subjects filled out the questionnaire during their routine physical exams at the on-base MCRD Branch Medical Clinic. Project staff explained the consent form and answered any questions the recruits had before administration of the items. Official Marine Corps personnel not associated with the study left the area during the administration of the survey to ensure confidentiality. Each item was read aloud to the recruits, and study personnel circulated around the group to answer any questions that the subjects might have about the items or procedure. The Marine Corps administrative records provided additional demographic information, including height (inches) and weight (lb) measured at the medical clinic the first week of arrival at camp. These values were converted to meters and kilograms, respectively, and BMI (kg·m−2) was calculated.
For analysis, four exposure variables were created: predicted number of pull-ups (used as a continuous variable), self-assessed fitness (dichotomized as “poor/fair” vs “good/very good/excellent”), competition (dichotomized as “no exercise/no competitive exercise” vs “competitive exercise”), and sweat frequency (dichotomized as “never/occasionally/fairly often” vs “quite a lot/most or all the time”). Age and BMI were used as continuous variables, and race/ethnicity was dichotomized as white versus other.
All data were double entered and validated visually by the principal investigator. All analyses were performed using Epi Info 2000 (Centers for Disease Control and Prevention, Atlanta, GA), SAS Version 8 (SAS Institute, Inc., Cary, NC), and SPSS Version 9 (SPSS, Inc., Chicago, IL) programming packages. Descriptive analyses included frequencies and distributions of demographic and questionnaire variables. In addition, questionnaire items and demographic variables were compared using a chi-square test for categorical variables, and a t-test for comparing means for study participants and the 454 recruits who did not complete their required fitness test because of illness or attrition.
Univariate analyses were performed to determine potential associations between questionnaire items (exposure variables) and PFT-1 score (outcome variable). Linear regression analysis was used to determine the strength of association between the score on the PFT-1 and potentially associated variables, controlling for age, BMI, and race/ethnicity. Statistical significance levels were based on a P value less than 0.05.
Of the original 2040 recruits, three men declined to fill out the questionnaire, and 454 men who filled out the questionnaire were dropped from the study because they did not take the PFT-1. Of the 454 men with missing fitness data, 331 did not take the PFT-1 because of illness or injury at the time of the test, and 123 men left the military sometime during the first 4 wk for various reasons. The 454 recruits who completed questionnaires but were missing fitness scores differed from study subjects only on whether they participated in competitive activity (62% of the study subjects participated in competitive activity versus only 58% of men who were missing PFT-1 scores, P < 0.05).
The mean PFT-1 score for the study recruits was 226.8 (range, 72–300). The average age of the 1583 recruits in this study was 19.2 ± 1.9 yr (range, 17–32 yr). Univariate comparisons showed no association between age and PFT-1 scores. The average BMI was 23.9 ± 3.4 and was inversely associated with PFT-1 scores. The population was predominantly white (63%), followed by Hispanic (25%) and African American (4%); about 8% were classified as “other or decline to state.” There were statistically significant differences in PFT-1 scores among the race/ethnicity groups, with the African American group having the highest score (238.2) and the white group having the lowest score (225.4). When used as a dichotomized variable, the white group had lower PFT-1 scores than the combined group (225.4 vs 229.2; P < 0.01) (Table 1).
Table 2 shows several self-assessed fitness and exercise parameters that were statistically significantly associated with PFT-1 scores. The mean number of predicted pull-ups was 10, and was associated with PFT-1 scores (P < 0.01). Predicted and actual number of pull-ups were correlated (r = 0.72; P < 0.01). Univariate associations indicated that self-assessed fitness, competitiveness, and frequency of sweating during exercise were all positively associated with PFT-1 scores (Table 2).
In the multivariate linear regression analysis, self-assessed fitness and exercise parameters remained statistically significant predictors of objective fitness (PFT-1) even after controlling for each other and age, BMI, and race/ethnicity (Table 3). Overall, the model explained 47% of the variance. Although all variables in the final model were statistically significant at P < 0.01, predicted number of pull-ups was the strongest variable, with a standardized beta of 0.62.
This study identified self-assessed fitness and exercise parameters that were associated with objectively measured fitness. Subjects were asked to rate themselves in terms of physical fitness, competitiveness, and exertion level (sweat frequency), and to estimate how many pull-ups they could currently do, all of which were significantly associated with PFT-1 score. Previous research demonstrates that individuals who are reasonably active can accurately assess their fitness and exercise intensity levels (6,13); this seems to have held true for the subjects included in this study.
The fact that recruits who believed before boot camp that they could do more pull-ups were more successful on the PFT-1 suggests that strength training might be beneficial for incoming recruits. This finding also corroborates previous research indicating that upper-body strength measures were more closely related with self-assessments than measures of lower-body strength (8).
Questionnaires in general have proven very useful for physical activity assessment in a variety of populations (5,12). Questionnaires are useful because they are nonreactive, practical, applicable to specific populations, and reasonably accurate. Indeed, the predictive strength of the direct self-assessment questions in regard to the recruits’ current fitness status and exercise parameters may be more accurate than variables for which recruits need to recall actual exercise habits. Remembering details of prior participation in physical activity may be very difficult, and requires great effort on the part of the respondent (3). Other studies have suggested that self-assessment may be biased according to social desirability of the behavior (1), although that does not seem to be operating here given the moderately high correlation between predicted and actual pull-ups (r = 0.72).
In the current study, categorical variables were dichotomized for use in a linear regression model, as the original categories did not meet the criteria for inclusion in a linear regression model. This dichotomization method has been used in other studies of health and fitness, and the dichotomized variables behaved very similarly to the original categorical variables from which they were derived in regard to size and significance of association (9).
These data could be useful to military personnel by identifying those individuals at risk for low fitness, and tailoring physical training before the PFT-1 to improve their scores. The questionnaire items identified can also be used to fine-tune preenlistment evaluation of incoming recruits.
The purpose of this research was to understand whether there was any predictive value in the questions about self-assessed fitness and exercise before the recruits had any objective data regarding their ability. Also, many men drop out during the first few weeks of boot camp, and our concern was to look at fitness predictors for those who made it through that critical time. Those who do not make it to the PFT-1 mainly consist of men who are injured/ill, or who have left the Corps for a number of reasons.
The specific limitations of using a military population involve generalizability issues because military volunteers might differ from other people in the same age range. Also, use of volunteers (in this case, individuals who volunteered to be Marines) raises the problem of selection bias within any population. Even within branches of the U.S. armed forces there are differences in health indicators. For example, Marine Corps men and women exhibit a lower prevalence of overweight than their Navy counterparts (2). However, these results may be very relevant to Marine recruits, and are representative of the population that entered during the study recruitment period (only three men who were selected did not complete the questionnaire).
The benefits of using a military population include the excellent quality of military health databases and performance records. The standardized fitness testing conducted regularly during Marine Corps training provided a strong measure of objective fitness for this study. The relatively homogeneous nature of the recruits enabled the men to be queried under standard conditions at the same point in their training program. For these reasons, it is appropriate and convenient to utilize military data to determine the usefulness of self-assessed fitness data, as long as the limitations of using military populations for this type of research are also recognized.
The present study demonstrates that self-assessed fitness and exercise parameters can predict objective physical fitness 1 month later. These measures are easily ascertainable by questionnaire. Including these types of questions when assessing beginning exercisers or those increasing the frequency, intensity, and time of their physical exercise (such as those beginning a military training program) might improve exercise prescription, which in turn might help prevent overuse injuries. Injuries resulting from inappropriate training levels are so common that preventive strategies and activities are justified on medical as well as economic grounds (10). Convenient, noninvasive methods for identification of lifestyle factors that predict low fitness levels or slower fitness progress could enable recruiters to make recommendations for fitness preparation to incoming recruits.
1. Eriksson, I., A. L. Unden, and S. Elofsson. Self-rated health: comparisons between three different measures: results from a population study. Int. J. Epidemiol
. 30:326–333, 2001.
2. Graham, W. F., L. L. Hourani, D. Sorenson, and H. Yuan. Demographic differences in body composition of Navy and Marine Corps personnel: findings from the Perception of Wellness and Readiness Assessment. Mil. Med
. 165:60–69, 2000.
3. Haskell, W. L., and M. Kiernan. Methodologic issues in measuring physical activity and physical fitness when evaluating the role of dietary supplements for physically active people. Am. J. Clin. Nutr
. 72(Suppl.):541S–550S, 2000.
4. Jones, B. H., and J. J. Knapik. Physical training and exercise-related injuries. Surveillance, research and injury prevention in military populations. Sports Med
. 27:111–125, 1999.
5. Kaufman, K. R., S. Brodine, and R. Shaffer. Military training-related injuries: surveillance, research, and prevention. Am. J. Prev. Med
. 18:54–63, 2000.
6. Knapik, J. J., B. H. Jones, K. L. Reynolds, and J. S. Staab. Validity of self-assessed physical fitness. Am. J. Prev. Med
. 8:367–372, 1992.
7. Knapik, J. J., M. A. Sharp, M. Canham-Chervak, K. Hauret, J. F. Patton, and B. H. Jones. Risk factors for training-related injuries among men and women in basic combat training. Med. Sci. Sports Exerc
. 33:946–954, 2001.
8. Kraemer, W. J., N. A. Ratamess, and D. N. French. Resistance training for health and performance. Curr. Sports Med. Rep
. 1:165–171, 2002.
9. Manor, O., S. Matthews, and C. Power. Dichotomous or categorical response? Analysing self-rated health and lifetime social class. Int. J. Epidemiol
. 29:149–157, 2000.
10. Parkkari, J., U. M. Kujala, and P. Kannus. Is it possible to prevent sports injuries? Review of controlled clinical trials and recommendations for future work. Sports Med
. 31:985–995, 2001.
11. Rupinski, T. E. Physical Fitness of Marine Corps Recruits
. Alexandria, VA: Center for Naval Analysis, Publication CRM 88-190, 1989, pp. 1–21.
12. Shaffer, R. A., S. K. Brodine, S. A. Almeida, K. M. Williams, and S. Ronaghy. Use of simple measures of physical activity to predict stress fractures in young men undergoing a rigorous physical training program. Am. J. Epidemiol
. 149:236–242, 1999.
13. Washburn, R. A., S. R. Goldfield, K. W. Smith, and J. B. McKinlay. The validity of self-reported exercise-induced sweating as a measure of physical activity. Am. J. Epidemiol
. 132:107–113, 1990.
Questions used to measure self-assessed fitness and exercise.
How many pull-ups can you do right now? _________
How would you rate your current physical fitness compared with others your age? (circle one answer)
Is your exercise activity oriented toward competition (for example, marathons, triathlons, power lifting, body building)?
0—Does not apply. I do not exercise
In your exercise or leisure activities how often do you “work up a good sweat”? (circle one answer)
3—Quite a lot
4—Most or All the time Cited Here...