PREDICTIVE ACCURACY OF BIOIMPEDANCE EQUATIONS IN ESTIMATING FAT-FREE MASS OF HISPANIC WOMEN

This study assessed the predictive accuracy of previously published bi oelectrical impedance analysis equations for estimating fat-free mass of young (20-39 yr) Hispanic women (N = 29). The reference method was hydrostatic weighing at residual lung volume. Body density was convert ed to percent body fat using the Siri equation. Resistance and reactan ce were measured with a Valhalla bioimpedance analyzer. The bioelectri cal impedance analysis equations of Lohman, Rising, Stolarczyk, Segal, Gray, and Van Loan were cross-validated. There were significant corre lations between criterion and predicted fat-free mass (r = 0.86-0.95) for all equations. The standard error of estimate for each equation wa s acceptable; however, the total error for the Stolarczyk (3.2 kg) and the Van Loan (4.6 kg) equations exceeded the recommended value (2.8 k g). For all equations, the difference between average criterion and pr edicted fat-free mass was significant (P < 0.05). However, the mean di fferences for the Lohman (0.8 kg), Segal (0.8 kg), and Gray (0.9 kg) e quations were small. In conclusion, the Segal, Lohman, and Gray equati ons may have potential for assessing the body composition of healthy, acculturated, Hispanic women.