Validity of equations for estimating fat-free mass in people living with the human immunodeficiency virus

Abstract

Background & aimsEquations estimating fat-free mass (FFM) in people living with the human immunodeficiency virus (HIV) show differences in the validation process. The current study aimed to verify the validity of FFM estimation equations derived from bioelectrical impedance (BIA) in people living with HIV aged 40 years and older. MethodsA cross-sectional study was conducted with 68 participants evaluated using dual-energy X-ray absorptiometry (DXA) and by two BIA devices (Analyzer and Biodynamics). The study aimed to determine the validity of six different FFM equations from four different studies by Lukaski and Bolonchuk (1987), Kotler et al. (1996), Beraldo et al. (2015) and Hegelund et al. (2017). Comparisons were made using the t-test or Wilcoxon test. To verify the validity between DXA and two BIA devices, the following statistical analyses were performed: Lin's concordance correlation coefficient, intraclass correlation coefficient, coefficient of determination, standard error of the estimate, differences in the limits of agreement by Bland and Altman analysis, correlation between the average and the differences of the methods by Pearson or Spearman correlation. ResultsOnly equation 2 of Kotler et al. (1996) for males by Biodynamics BIA showed no difference in FFM. The Lin's concordance correlation coefficient was excellent (0.96), irrespective of sex, for Equation 2 of Kotler et al. (1996) by BIA Analyzer. All equations were reproducible (>0.85). The coefficient of determination ranged from 68% to 92%, and the standard error of the estimates ranged from 1.8 kg to 5.0 kg. The differences between the limits of agreement ranged from 7.2 kg to 14.9 kg, and the correlations between the average and the differences of the methods showed differences in FFM for three equations (p<0.01). ConclusionThe choice of equations must consider the equipment used and the sex of the sample investigated. Only Equation 2 of Kotler et al. (1996) was considered valid, irrespective of sex, to estimate the FFM by BIA Analyzer.