ABSTRACT
AIM: When using leg-to-leg bioelectrical impedance analysis (LBIA) to examine body composition, a computer-programmed mode (i.e., standard STD or athletic ATH) must be selected prior to assessment. This study examined the effect of LBIA mode selection on the estimated %BF of collegiate wrestlers. METHODS: Forty hydrated (Usg <1.02) wrestlers had %BF estimated using the ATH mode, which was then compared to the STD mode and hydrostatic weighing (HW), used as the reference method. Mean difference from HW (MD), standard error of estimate (SEE), and pure error (PE) values were calculated for the entire sample and three data subsets according to body mass index (BMI, kg/m(2)): <25 (n=16), 25-29.9 (n=18), and >30 (n=6). RESULTS: The %BF (mean+/-SD) was underestimated by the ATH (12.1+/-4.7) and overestimated by the STD (17.1+/-5.2) mode when compared to HW (14.1+/-6.3) for the entire sample (P<0.05). When examined relative to BMI, the ATH mode accurately estimated %BF in the BMI <25 group (MD=-1.2%, SEE=2.7%, PE=2.8%) and the STD mode accurately estimated %BF in the BMI >30 group (MD=1.5%, SEE=2.4%, PE=2.8%). Both modes inaccurately assessed %BF in the BMI 25-29.9 group with predictive errors >3.5%BF. CONCLUSIONS: The ATH mode is not appropriate for all individuals meeting the definition of athletic. However, the predictive accuracy of LBIA may be improved by selecting the ATH mode when BMI <25 and the STD mode when BMI >30, even when testing athletes.
