PREDICTIVE ACCURACY OF BIOIMPEDANCE METHOD IN ESTIMATING FAT-FREE MASS (FFM) OF AMERICAN INDIAN, HISPANIC AND WHITE MEN 1158

Abstract

Although some BIA equations (EQ) accurately predict the average FFM for a total group, clinicians need information about the accuracy of BIA EQs for individual clients. This study compared the predictive accuracy of previously published BIA EQs for estimating both the group and individual's FFM. Although the Segal (1988) fatness-specific (FS) EQs (<20%BF; ≥20% BF) have documented validity for men, the clinician must select the appropriate FS EQ for each client based on an “a priori” subjective rating of body fat (BF). This is difficult, especially for clients between 17-25% BF. To resolve this problem, we tested the hypothesis that using the average FFM estimated from each of the FS EQs would yield an accurate estimate of the client's BF. Reference FFMhw was obtained for 94 American Indian, 53 Hispanic and 83 White men, 18-62 yr, using hydrodensitometry at RV and Siri's formula to convert Db to%BF. R and Xc were measured with a Valhalla bioimpedance analyzer. The results for the total sample were:Table The Segal average EQ accurately estimated average FFM of this sample and yielded the smallest group (E=3.62kg) and individual prediction errors (33%). In 2 out of every 3 cases, the individual's%BF was accurately estimated within±3.5%. We recommend using Segal's FS equations with the average method to assess body composition of American Indian, Hispanic and White men.