Agreement of Single-Frequency Electrical Bioimpedance in the Evaluation of Fat Free Mass and Fat Mass in Peritoneal Dialysis Patients

Abstract

Background: Protein-energy wasting is related to impairment of quality of life and lower survival of end-stage kidney disease (ESKD) patients. The evaluation of body composition, especially fat free mass (FFM) and fat mass (FM), is important for the prediction of outcomes in these individuals. The aim of this study was to compare the FFM and FM measurements obtained by single-frequency bioimpedance (SF-BIA) and by a multiple frequency bioimpedance (MF-BIA) device, using dual energy X-ray absorptiometry (DXA) peritoneal dialysis (PD) patients. Methods: This was a cross-sectional study involving adult patients undergoing regular PD, in which we performed SF-BIA, MF-BIA, and DXA at the same visit. To compare the bioimpedance values with DXA, we used: Person correlation (r), intraclass correlation coefficient (ICC), and Bland-Altman concordance analysis. Results: The sample consisted of 50 patients in the PD, with mean age of 55.1 +/- 16.3 years. Both bioimpedance methods showed a strong correlation (r > 0.7) and excellent reproducibility (ICC > 0.75) compared to DXA. According to the Bland-Altman diagram, SF-BIA showed agreement in body compartment measurements, with no proportionality bias (p > 0.05), without systematic bias for FFM (-0.5 +/- 4.9, 95% CI -1.8 to 0.9, p = 0.506), and for FM (0.3 +/- 4.6, p = 0.543). MF-BIA did not present a proportionality bias for the FFM, but it underestimated this body compartment by 2.5 +/- 5.4 kg (p = 0.002). In addition, MF-BIA presented proportionality bias for FM. Conclusion: SF-BIA was a more accurate assessing method than MBIA for FFM and FM measurements in PD patients. Because it is a low-cost, non-evaluator-dependent measurement and has less systematic bias, it can also be recommended for fat mass and free-fat mass evaluation in PD patients.