Abstract des Autors

The purpose of this study was to validate single-frequency foot-to-foot bioelectrical impedance analysis (FF-BIA) against a 3-compartment (3C) model in NCAA Division I male athletes. Thirty-three athletes (football = 19, baseball = 8, basketball = 3, and cheerleading = 3) had body fat percentages (BF%) and fat-free mass (FFM) estimated using a 3C model and FF-BIA. The criterion 3C model was derived from body mass, body volume (air displacement plethysmography), and total body water (bioimpedance spectroscopy). The mean BF% and FFM values for FF-BIA were not statistically significant when compared with the 3C model (p = 0.14 and 0.28, respectively). The standard error of estimate (2.79% and 2.64 kg), total error (2.95% and 2.64 kg), and 95% limits of agreement (±5.67% and ±5.15 kg) were considered acceptable for BF% and FFM, respectively. However, there was a significant trend in the regression line of the Bland-Altman plot, which indicated proportional bias for BF% (r = −0.50; p < 0.01). No proportional bias was present for FFM (r = 0.26; p > 0.05). Foot-to-foot BIA seems to be valid for estimating group and individual athlete’s FFM. Similarly, FF-BIA can be used for estimating group BF%. However, proportional bias indicates that FF-BIA is not valid for individual estimates of BF%.