Body composition measured by bioelectrical impedance analysis is a viable alternative to magnetic resonance imaging in children with nonalcoholic fatty liver disease

Objective To determine the relationship between bioelectrical impedance analysis (BIA) and magnetic resonance imaging (MRI) obtained measures of body composition in children with nonalcoholic fatty liver disease (NAFLD). Methods Youth with obesity and NAFLD who had BIA and abdominal MRI testing were included. BIA measured skeletal muscle mass (SMM), appendicular lean mass (ALM), trunk muscle mass (TMM), and percent body fat. MRI measured total psoas muscle surface area (tPMSA) and fat compartments. Univariate analysis described the relationship between BIA- and MRI-derived measurements. Multivariable regression analyses built a model with body composition measured via MRI. Results 115 patients (82 (71%) male, 38 (33%) Hispanic, median age14 years) were included. There was a strong correlation between tPMSA and SMM, ALM, and TMM (correlation coefficients [CCs]: 0.701, 0.689, 0.708, respectively; all P < .001). Higher SMM, ALM, and TMM were associated with higher tPMSA. This association remained after controlling for age, sex, ethnicity, type 2 diabetes mellitus status, and body mass index z-score. Total fat mass by BIA and MRI-determined total, subcutaneous, and intraperitoneal fat area correlated significantly (CCs: 0.813, 0.808, 0.515, respectively; all P < .001). In univariate regression, higher total fat mass by BIA was associated with increased total fat area and increased fat in each of the four regions measured by MRI. After controlling for confounders, the association between total fat mass by BIA and total fat area by MRI persisted. Conclusions BIA measures of muscle and fat mass correlate strongly with MRI measures of tPMSA and fat areas in children with obesity and NAFLD.