A comparison of methods to predict minimal weight in high school wrestlers

The purpose of this study was to compare the accuracy of body fat determinations and subsequent calculation of minimal weight (MW) by dual energy x-ray absorptiometry (DEXA), bioelectrical impedance (BIA), near-infrared photospectometry (NIR), and anthropometry (LOHMAN). Necessitated by mandatory state minimal weight testing, the methods were cross-validated on 95 Wisconsin high school wrestlers (mean +/- SD; age: 15.1 +/- 1.2 yr, height: 170.4 +/- 7.1 cm, weight: 63.4 +/- 9.8 kg). MW, defined as fat-free body/0.93, determined by hydrostatic weighing (HW) and residual volume via O2 dilution, served as the criterion. The validity of the four selected MW predictions were evaluated against HW by examining mean differences (MD), standard deviation differences (SDD), correlations (r), standard error of estimate (SEE), and total errors (TE). Statistically significant differences were shown between the methods and the criterion by t-tests; however, these were clinically small in Lohman (0.6 kg) and BIA (0.9 kg). TE ranged from 2.25 kg (Lohman) to 6.03 kg (NIR). The results indicated that Lohman skinfold equation provided the most accurate prediction of MW, demonstrating the highest correlation (0.972), lowest MD (0.6 kg), lowest SEE (2.12 kg), and lowest TE (2.25 kg) of the methods evaluated.