VALIDITY OF ANTHROPOMETRIC TECHNIQUES FOR ESTIMATING PERCENTAGE BODY-FAT IN OBESE FEMALES BEFORE AND AFTER SIZABLE WEIGHT-LOSS

The validity of selected generalized (n = 7) and obese population-spec ific (n = 2) skinfold and girth based anthropometric equations (using hydrostatic weighing as the criterion measure) was evaluated in 19 obe se women before and after a mean body weight loss of 19.4 kg (percent body fat was reduced from 45.3% to 35.4%). For the present study, eval uation criteria of mean differences in percent body fat (%BF) of < 3%, r values > 0.70, and total error (T.E.) values of < 5% were selected. Only the equations that utilized girths as part of the prediction tec hnique (Barrows and Snook, 1987; Jackson et al., 1980; Tran and Weltma n, 1989; Weltman et al., 1988) resulted in mean differences in %BF of less than 3%, r values of > 0.70 and T.E. values of less than 5% in ob ese women before weight loss. After 19.4 kg of weight loss, all but th ree of the equations resulted in mean differences in %BF of less than 3%, r values of > 0. 70 and T.E. values of less than 5%. The obese pop ulation-specific equations of Barrows and Snook (1987) and Weltman et al. (1988) were not accurate predictors of body composition after weig ht loss as they resulted in mean differences of 4.5 - 7.0 %BF. Althoug h some of the equations examined were valid predictors of %BF either b efore or after weight loss, only the Jackson et al. (1980) equations w hich utilized both skinfolds and hip girth, and the Tran and Weltman ( 1988) equation which utilized trunk girths, resulted in mean differenc es of < 3 %BF, r values > r = 0.70, and T.E. scores of < +/- 5 %BF bot h before and after weight loss. It was concluded that the prediction t echniques of Jackson et al. (1980) and Tran and Weltman (1989), which utilize girth measures as part of the prediction technique, may have c linical utility for estimating changes in %BF in obese females before and after sizable weight loss. (C) 1993 Wiley-Liss, Inc.