RESUMO
INTRODUCTION: Bariatric surgery has a significant influence on body composition (BC), which should be monitored. However, there is a need to recommend low-cost practical methods, with good estimation of BC for class III obese and/or bariatric patients. OBJECTIVE: The aim of this study was to determine accuracy and agreement between BC assessed by direct segmental multifrequency bioelectrical impedance analysis (DSM-BIA) and doubly labeled water (DLW) as reference method. MATERIAL AND METHODS: Twenty class III obese women (age 29.3 ± 5.1 years; body mass index 44.8 ± 2.4 kg/m2) underwent Roux-en-Y gastric bypass surgery. BC (fat mass [FM], fat-free mass [FFM], and total body water [TBW]) was assessed by InBody 230 and DLW in the following periods: before and 6 and 12 months after surgery. Accuracy between the methods was evaluated by the bias and root mean square error. Pearson's correlation, concordance correlation coefficient (CCC), and Bland-Altman method were used to evaluate agreement between the methods. RESULTS: Correlations were significant (p < 0.001) and CCC was good/excellent between both methods for the evaluation of FM (r = 0.84-0.92, CCC = 0.84-0.95), FFM (r = 0.73-0.90, CCC = 0.68-0.80), and TBW (r = 0.76-0.91, CCC = 0.72-0.81) before and after bariatric surgery. In addition, no significant bias was observed between DSM-BIA and DLW for FM (mean error [ME] = - 1.40 to 0.06 kg), FFM (ME = 0.91-1.86 kg), and TBW (ME = 0.71-1.24 kg) measurements. CONCLUSION: The DSM-BIA was able to estimate the BC of class III obese women submitted to bariatric surgery with values consistent with those of the DLW method.