A validation study of body composition by bioelectrical impedance analysis in human immunodeficiency virus (HIV)-positive and HIV-negative Hispanic men and women.

Bioelectrical impedance analysis (BIA) is a potentially useful tool for measuring body composition in people with human immunodeficiency virus (HIV). However, it is not clear that equations derived in healthy non-Hispanic whites can be applied to people who are of other races or ethnicities and who are infected with HIV. Body composition measures done by BIA, using the equations of Lukaski, were compared to measures of body composition derived from dual-energy x-ray absorptiometry (DXA) in Hispanic men and women of Caribbean origin (predominantly Puerto Rican) with and without HIV infection. In cross-sectional analyses, body composition was measured by BIA and DXA in four groups of Hispanics: 97 HIV-positive men, 70 HIV-negative men, 38 HIV-positive women, and 14 HIV-negative women. The method of Bland and Altman was used to evaluate the validity of BIA compared to DXA. Compared to DXA, BIA provided accurate measures of fat-free mass in HIV-positive and HIV-negative Hispanic men. Fat-free mass by BIA compared to DXA was overestimated by 2.7 kg (standard deviation=2.5; P<0.0001) in the HIV-positive Hispanic women and by 3.4 kg (standard deviation=2.6; P<0.01) in the HIV-negative women. The magnitude of the bias in fat-free mass was dependent on fat mass in both the men and the women. BIA, using the equations of Lukaski, appears to be useful in this Hispanic population of Caribbean origin with and without HIV, for whom it provided reasonable estimates of body composition. Fat mass affects the accuracy of estimates.