A COMPARISON OF FOUR METHODS IN ASSESSING BODY FAT IN COLLEGIATE WRESTLERS

Abstract

Dangerous weight-cutting practices in the sport of wrestling have prompted the NCAA to implement a minimum weight certification program. Currently, the NCAA requires percent body fat (%BF) be determined by either hydrostatic weighing (HW) with measured residual volume or a three-site skinfold (SF) procedure. However, additional body composition methods such as air-displacement plethysmography (BodPod) and leg-to-leg bioelectrical impedance (BIA), which are easier to administer and reduce technician error, may provide alternatives of estimating %BF in wrestlers. PURPOSE To compare four different methods of estimating body fat and to determine the prediction error of SF, BIA, and BodPod when compared to HW in a group of collegiate wrestlers. METHODS A Division III collegiate wrestling team (N=25) volunteered to participate in this study. A urine specific gravity < 1.020, measured by refractometry, was required prior to all testing. SF, BIA, BodPod, and HW were used to estimate %BF. A three-site (subscapular, triceps, abdominal) SF procedure and prediction equation by Lohman was used to determine body density. The Brozek equation was used to convert the body density to %BF. BIA measurements (athletic mode) were determined using a Tanita body fat analyzer (model TBF-300A). HW, corrected for residual lung volume, was used as the criterion measurement. RESULTS The %BF data (mean ±D) for the methods were BIA (12.3 ± 4.6), SF (14.2 ± 5.3), BodPod (13.8 ± 6.3), and HW (14.5 ± 6.0). BIA was significantly lower (p < 0.01) than HW and SF. There were no significant differences between SF, BodPod, and HW. All methods showed significant correlations (r=0.80–0.96, p < 0.01) with HW. The standard errors of estimate (SEE) for %BF were relatively small including 1.7% BodPod, 1.9% SF, and 3.6% BIA. CONCLUSION These findings support the use of SF and BodPod as alternative methods to HW in estimating %BF in collegiate wrestlers. BIA, which consistently underestimated %BF, is not supported by these data as a valid alternative in this sport.