Validity of bioimpedance for assessment of fat-free mass in women with Rheumatoid Arthritis compared to non-rheumatic controls

BACKGROUND & AIMS: The aim was to assess the validity of bioimpedance in the assessment of fat free mass (FFM) among women and to study if the validity differs between women with and without Rheumatoid Arthritis (RA). METHODS: 38 women with RA and 24 non-rheumatic controls were included. FFM was measured in the non-fasting state using DXA (Lunar Prodigy), multi-frequency bioelectrical impedance analysis (BIA) (MF-BIA [Tanita MC-180 MA]), single-frequency BIA (SF-BIA) and bioelectrical impedance spectroscopy (BIS) (both Impedimed SFB7). BIS raw data were also used to calculate FFM from equations by Matthie, Jaffrin and Moissl. Results were compared using correlation, Bland Altman analysis, Mann Whitney U test and Wilcoxon Signed Ranks Test. RESULTS: Women with RA had different body water distribution compared to women without RA (p < 0.05). Median bias in FFM assessed by bioimpedance was 0.62-7.87 kg with wide limits of agreement for all methods. Median FFM differed significantly from DXA by all bioimpedance methods except for BIS by Jaffrin. Women with RA had significantly smaller biases compared to non-rheumatic controls using BIS equations by Matthie (p = 0.012) and Moissl (p = 0.025). Correlations between FFM measured by DXA and bioimpedance (r = 0.73-0.85, all p < 0.001) did not differ between groups. The sensitivity of bioimpedance to detect low fat free mass index (FFMI) was 0-47%. CONCLUSION: The results of this study show that bioimpedance has similar validity in women with RA compared to non-rheumatic controls, despite differences in body water distribution. Agreement with DXA improved when applying specific equations, but the clinical utility of bioimpedance is questionable as all methods failed to identify low FFMI with acceptable precision. TRIAL REGISTER: Clinicaltrials.org, NCT04247009.