MILITARY MEDICINE, 177, 9:1090, 2012

Modifying the U.S. Air Force Fitness Test to Reflect Physical Combat Fitness: One Study’s Perspective Capt Thomas Worden, USAF; Edward D. White III, PhD ABSTRACT Within the past few years, the U.S. Army and U.S. Marine Corps have either incorporated or in the process of incorporating physical fitness tests that assess a soldier’s fitness capability in a combat environment. In this article, we investigate how the U.S. Air Force Physical Fitness Test (AFPFT) events compare to sister-services’ physical fitness test events with respect to their predictability of a proxy for combat fitness. Using linear regression via ordinary least squares, we demonstrate that the AFPFT correlates poorly with an airman’s combat fitness. AFPFT scores had minimal predictability (adjusted R2 0.20 – 0.23). Investigating all of the individual events from the three services tests, we develop a new Air Force fitness to consider and to adopt. The test incorporates a timed 1/2-mile run, 30-lb dumbbell lifts in 2 minutes, and maximum number of push-ups in 1 minute. The adjusted R2 for this model was 0.90, an almost 300% improvement over the current test, in terms of predicting combat fitness.

INTRODUCTION Based on FY04 decrement rates, only 17% of a typical group of new entrants attain 20 years of active duty service and become eligible for nondisability retirement from active duty.1 Castro and Adler2 projected an average military deployment rate of once every 18 months, which equates to several deployments before a first term enlisted member’s commitment or a new officer’s obligation time is completed. Consequently, an airman today is more likely to be deployed than to spend his or her entire career in the U.S. Air Force. Combat of today is evolving and the Air Force overseas mission is getting closer and closer to the Army and Marines role. At any given time, between 4,000 and 5,000 airmen have been or are deployed to Iraq and Afghanistan in jobs intended to be filled by soldiers and marines, and airmen are increasingly responsible for patrolling the areas around their expeditionary bases.3 All the armed services of the United States have at least one version of a distance run, timed push-ups, and timed sit-ups test.4 In a comprehensive review, Harman and Frykman5 concluded that push-up, sit-up, and 2-mile run scores are not potent determinants of physically demanding military task performance. Because of this, both the Army and the Marines have questioned that a fitness program simply measuring general fitness may not be sufficient at maintaining or measuring combat fitness in their personnel. These sister services then explored additional or alternative tests to better measure combat fitness, or the ability to handle the stresses, strains, and urgent demands required in combat situations. Traditional physical training programs have focused on aerobic endurance because this capacity is a required element of military physical fitness tests.6 The combat tasks anticipated in distributed operations suggest the need for increased emphasis on other capabilities, including anaerobic power, balance, and flexibility.7 In August 2008, the Marines

Department of Mathematics and Statistics, Air Force Institute of Technology, AFIT/ENC, 2950 Hobson Way, Wright-Patterson AFB, OH 45433.

1090

approved a second fitness test called the Combat Fitness Test (CFT), which is more focused on broader, real-life combat necessary skills that mainly look at burst speed and anaerobic ability.8 Unlike the traditional physical fitness test that measures endurance and strength with pull-ups, crunches, and a distance run, the CFT is designed to measure abilities demanded of Marines in a war zone.9 Like the U.S. Marines, the U.S. Army recognizes that their fitness tests need to change to ascertain combat fitness in their forces. Since 2002, the Army has been considering changing their current 3-part Army Physical Fitness Test (APFT) to a 6-part Army Physical Readiness Test (APRT), which includes more events to better encompass combat ability in addition to general fitness, per their Draft Field Manual 3-25.20.10 Because of this shift to a more combat-oriented fitness test, an overhauled physical fitness test program that links a soldier’s fitness and exercise to the demands of combat and long deployments is on its way to the operational Army.11 The Air Force appears to also recognize that training for combat needs to evolve. December 2008 marked the beginning of the newly incorporated “Beast” 4-day expeditionary training site within the Air Force’s newly expanded 8.5 week Basic Training. “Beast” is perhaps the most visible sign yet that the Air Force is adapting its training to the realities of fighting two counterinsurgencies that increasingly put airmen in harm’s way.3 However, unlike the Army and Marines, the Air Force has not changed or altered their current physical fitness test to assess how airmen may fare in combat by assessing their combat fitness. Over the years, the Air Force Physical Fitness Test (AFPFT) has changed from mainly a 1.5-mile run, to the bike (ergometry) test, and then back to the run test. Throughout this time frame, the Air Force fitness test aim is to promote physical health and overall fitness. The AFPFT involves a timed 1.5-mile run, an abdominal circumference (AC) measurement, the maximum number of push-ups in 1 minute, and the maximum number of sit-ups in 1 minute. Total possible AFPFT score consists of 100 points. At the time of this study, MILITARY MEDICINE, Vol. 177, September 2012

Downloaded from publications.amsus.org: AMSUS - Association of Military Surgeons of the U.S. IP: 037.044.207.053 on Jan 27, 2017. Copyright (c) Association of Military Surgeons of the U.S. All rights reserved.

Modifying the U.S. Air Force Fitness Test to Reflect Physical Combat Fitness

the 100 points were proportioned as 50/30/10/10 for the run, AC, push-ups, and sit-ups accordingly. The current AFPFT has the same events, but the scaling has changed slightly to 60/20/10/10. Studies have shown that the three events of the run, push-up, and sit-up bias against heavier, not fatter, personnel when predicting fitness for load-carrying tasks12 and correlate poorly with the ability to perform physically demanding occupational tasks.13,14 The importance of the ability to carry external loads beyond body weight has been well established for service members15; yet heavier personnel are penalized in the scoring and are not related to body fatness but are based on biological scaling and have a physiological basis.16 Similarly, being smaller and lighter serves as advantages for the push-up, sit-up, and distance run tests of the military.17 With these studies in mind in conjunction with anecdotal accounts that the AFPFT promotes development of lean runners in lieu of bulkier airmen with combat strength, we studied if the AFPFT correlated well with respect to combat fitness as assessed by the Army and Marines. We tested the hypothesis that achieving a high score on the AFPFT translated to achieving a high score on a test that assessed combat fitness. Our results suggest the Air Force fitness test poorly predicted combat fitness. Because of this result, we developed and submit for consideration a simple three-event fitness test that is straightforward to administer, takes less time than the current fitness test, and has strong correlation to an airman’s combat fitness as assessed by the Army and the Marines. METHODS For our study, we collected data on the performance measures of each event on the AFPFT, as well as each event on the sister-service CFTs (the Marine CFT and the proposed APRT) for each subject in the study. We had a total of 86 volunteers, 81 men and 5 women, from Wright-Patterson Air Force Base (WPAFB) who took part in this study, which was approved by the WPAFB Institutional Review Board for human subjects testing. We also collected demographic data such as height, gender, age, and weight. Neither smoking history nor medical history was recorded of the subjects. None of the volunteers were on profile, and each subject completed all aspects of testing. Our research interest is based on combat fitness, which is evaluated differently for the U.S. Marines and U.S. Army. Equally weighted stratifications on the Army PRT and the Marine CFT were used to simulate the dependent variable called “combat fitness.” The Army test had six equally weighted events that had their own mean times/distances/ repetitions and their own standards of deviation. The Marine test had three equally weighted events that had their own mean times/repetitions and their own standards of deviation. We gave a percentile rank for each volunteer of the study for each event. Percentile ranks were then converted to a point score between 0 and 100, where any volunteer falling into the bottom 10 percentile received no points and those in

the top 10 percentile received all 100 points. Those percentile ranks between the bottom 10% and the top 10% of each event received a score between 0 and 100 based on a normal distribution of the middle 80% of volunteers and a mean score of 50. For the Army PRT, there were six events; so the total “Army Composite” had a range of up to 600 points. The six events consist of the standing long jump, power squat, shuttle run, push-ups, 1-mile run, and heel hooks. For the Marine test, there were three events so the total “Marine Composite” had a range of up to 300 points. The three events consist of a halfmile run, 30-lb lift (study used a 30-pound weight in lieu of the 30-pound ammunition crate that the Marines typically use), and the Maneuver-Under-Fire drill (a timed 300-yard obstacle course). The Combat Fitness Composite was 50% Army Composite and 50% Marine Composite. Marine Composite points were doubled to end with a total point range between 0 and 1200 points. Combat Fitness Composite scores from all volunteers were used to calculate a mean and a standard deviation, and standardized scores were created and then utilized to form a percentile rank. This Combat Fitness Composite is a proxy for the stratified measure of combat fitness among this group of peers. This same process was also done utilizing Air Force fitness test results for each of the volunteers, since volunteers were required to submit their Air Force Portal printouts of their latest Air Force fitness test results. Each event performance measure on the Air Force test (waist circumference if body mass index [BMI] over 25, 1.5-mile run time, 1-minute pushups repetitions, and 1-minute sit-ups repetitions) was recorded as both raw data and again as the Air Force points rewarded per the appropriate charts that determine event point totals based on sex and age. Since the Air Force currently uses these events and these charts within their Fit-To-Fight program, these point totals served as the baseline for this study. After data collection, we performed several ordinary least squares (OLS) regressions to ascertain possible correlations between the Combat Fitness Composite Score and the various elements of the AFPFT to include the entire score itself. If we noted a lack of correlation, we further investigated how one could improve predicting combat fitness from a smaller subset of events to include any elements from the Air Force, Army, or Marine fitness tests. We document these findings next. RESULTS Using OLS regressions, we first addressed whether or not there was a correlation between a volunteer’s percentile ranking on the Air Force fitness test and his or her percentile ranking on the Combat Fitness Composite. Whether we looked at the percentile versus percentile comparison or the actual AFPFT composite score versus composite fitness score, we could detect only a weak association between these measurements at best. Figures 1 and 2 illustrate these findings. The adjusted R2 of these figures are 0.23 and 0.20, respectively. (Note: we used adjusted R2 in lieu of R2 since R2 can be artificially

MILITARY MEDICINE, Vol. 177, September 2012 Downloaded from publications.amsus.org: AMSUS - Association of Military Surgeons of the U.S. IP: 037.044.207.053 on Jan 27, 2017. Copyright (c) Association of Military Surgeons of the U.S. All rights reserved.

1091

Modifying the U.S. Air Force Fitness Test to Reflect Physical Combat Fitness

FIGURE 1. Scatterplot of an airman’s Air Force fitness test percentile (stratification) score versus his/her combat composite percentile score.

increased).18 The results show poor correlation between a volunteer’s combat fitness as assessed by the Army and Marines and his or her score on their Air Force fitness test. We conducted c 2 goodness of fit tests to test whether or not there was a dependency between being heavier and improvement in peer stratification. Improvement in peer stratifications from the AFPFT to the Combat Fitness Composite has a dependency on BMI >25 (p = 3.19E-7), even with AC excluded from the AFPFT scoring (p = 0.00586). Both statistical tests show a volunteer’s percentile ranking among their peers will improve on the CFT compared to their ranking on the AFPFT if their BMI is greater than 25, whether AC is included or not in the AFPFT. We also investigated whether being heavier increases one’s chance of successfully completing the fireman’s carry of the Marines Maneuver-Under-Fire (MANUF) component. Completion of the fireman’s carry has a dependency on both

FIGURE 2. Scatterplot of an airman’s Air Force fitness test score versus his/her combat composite score.

1092

BMI >25 ( p = 0.00152) or a waist >32.5² ( p = 0.00521). Heavier personnel had a higher likelihood of success of conducting a fireman’s carry in comparison to their thinner peers. Our study supports previous research that shows heavier, not fatter, military personnel perform better on load-carrying tasks than their lighter peers4 and having biased scoring on fitness tests but not on combat-oriented fitness tests.16 Since these results imply the AFPFT poorly predicts an airman’s combat fitness as assessed by the Army and Marines, we aimed to build a new Air Force test that would correlate better with an airman’s combat fitness with the available study data. While doing so, we were cognizant that the Air Force scoring method causes a bias in the raw data for both the pushups and the sit-ups because of the charts having pre-existing maximum scores. In some cases, a subject may stop at the maximum even though they have the ability to continue since they already received the maximum score. It is clear that the maximum scores per the age/sex charts are too easily achieved when approximately 85% of the volunteer subjects got to or exceeded the maximum for sit-ups and 77% got to or exceeded the maximum for push-ups. This likely causes a limited range in the number of repetitions for subjects and a lower level of predictability in the variable itself. If sit-ups and/or push-ups appeared in the new proposed Air Force CFT, there would be no maximum number (which the Army test currently adopts). During the course of volunteer testing, subjects attempted nine new events that they were not familiar with (six from the APRT and three from the Marine CFT). Some of these events may be very predictive when included in our Air Force test or used to replace a less predictive event on the Air Force test. The first step to designing an ideal testing model based on these possible events is to check each variable for individual predictability when plotted against the Combat Fitness Composite. We did this through individual regression analysis as demonstrated previously in Figures 1 and 2. If the adjusted R2 of the model with just the single variable is significantly high and the p value is under the 0.05 threshold (our designated level of significance for the study), then that variable is considered significantly predictive. After conducting the regression analysis, we present our new Air Force fitness test in Figure 3 and Table I. This threeevent model consists of the 1/2-mile timed run, 30-lb overhead dumbbell lifts (total number in 2 minutes), and the Army push-ups (number of push-ups one can do in 1 minute). The adjusted R2 of this model is 0.903, which is a significant improvement over the current Air Force fitness test. We also investigated what model would be reached if we did not double the Marine Composite score of 300 and added that directly to the Army’s composite score of 600 for a total of 900 instead of 1200 as we did. We reached similar conclusions as what is presented. The main limitation that we see for these results stems from the fact we only had 86 volunteers for the study. It is possible that this small group may not accurately reflect the entire MILITARY MEDICINE, Vol. 177, September 2012

Downloaded from publications.amsus.org: AMSUS - Association of Military Surgeons of the U.S. IP: 037.044.207.053 on Jan 27, 2017. Copyright (c) Association of Military Surgeons of the U.S. All rights reserved.

Modifying the U.S. Air Force Fitness Test to Reflect Physical Combat Fitness

FIGURE 3. Scatterplot of an airman’s predicted combat composite score versus the airman’s actual composite score.

range of airmen performance on both the Air Force fitness test and on the proxy for combat fitness as administered by this study. Yet, we mitigate this limitation with the results that we saw. That is, our limited study’s findings precisely mirrored what troops coming back from combat areas anecdotally relayed: leaner, runner-style airmen who excelled in the Air Force fitness test generally underperformed their huskier counterparts when it came to the combat fitness proxy. Our results also are supported by studies that show heavier, not fatter, personnel better perform in load-carrying activities, of which carrying equipment, weapons, food, etc., easily constitutes in combat areas. DISCUSSION Through their incorporation of the CFT in addition to their general fitness test, the Marines have taken an excellent step in raising the standards of combat fitness in its members to minimize combat risks. This is a very recent step and should be critically evaluated on its success by tracking scores, scoring criteria, as well as mishaps that could have been avoided because of increased physical capabilities on combat necessary tasks. If these trends mimic what has been seen in units incorporating other forms of functional fitness such as Cross Fit, it is likely that both combat and noncombat mishaps will decrease for the Marines as this new test and training becomes TABLE I. Explanatory Variables, Parameter Estimates, t-Ratios, and Associated p-Values for the Proposed New Air Force Fitness Test Term Intercept 30-lb Lifts 1/2-Mile Run Army Push-ups

Estimate 1339.1334 5.8866984 −7.59029 3.6253556

t-Ratio

p > |t|

7.79 10.65 −10.65 3.68