The validity of anthropometric regression equations in predicting percent body fat in collegiate wrestlers

Background This study was designed to assess the validity of previously-developed anthropometric regression equations in determination of percent body fat in collegiate wrestlers (N=22). Methods. Means (M) standard deviations (SD), correlation coefficients(R), standard errors of estimate (SEE) and total errors (TE) were compared to the criterion test of hydrostatic weighing (HW), (M=12.49%, SD=4.67%). Results. Repeated measures analysis of variance revealed the equations of Lohman (M=12.9%, SD=3.59%, R=0.743, SEE=3.13%, TE=3.58%), Katch and McArdle (M=12.41%, SD=3.41%, R=0.750, SEE=3.10%, TE=3.18%), and Tipton and Oppliger (M=11.89%, SD=3.01%, R=0.763, SEE=3.03%, TE=3.63%) to be valid. The equations of Thorland et al. (M=14.11%, SD=5.12%, R=0.711, SEE=3.27%, TE=4.89%), Durning and Wommersley (M=14.075, SD=4.45%, R=0.741, SEE=3.13%, TE=4.71%), Jackson and Pollock (M=9.76%, SD=3.13%, R=0.737, SEE=3.17%, TE=5.90%), and Sinning (M=10.84%, SD=4.19%, R=0.741, SEE=3.13%, TE=4.78%) were significantly different from HW. Conclusions. It is concluded that the above-validated formulas have acceptably low SEE's and TE's to be of practical value in determination of appropriate weight classes for collegiate wrestlers.