Impedance-based measures of muscle mass can be used to predict severity of hepatic steatosis in pediatric nonalcoholic fatty liver disease

Abstract

ObjectivesThe objective of this study was to investigate the association between measures of body composition based on bioelectrical impedance analysis (BIA) and histologic severity of liver disease in a pediatric cohort with nonalcoholic fatty liver disease (NAFLD). MethodsThis was a cross-sectional study of patients < 20 y old with histologically confirmed NAFLD followed in our Steatohepatitis Center from 2017 to 2019. Contemporaneous body-composition data were obtained using a multifrequency octopolar BIA device (InBody 370, InBody, Seoul, South Korea). BIA data collected were skeletal muscle mass, appendicular muscle mass, and percentage body fat. Skeletal and appendicular muscle mass were corrected for height (dividing by the square of height), generating their respective indices. Univariate linear and logistic regression, followed by multivariable logistic regression analyses, were used. ResultsOf the 79 children included (27% female, 73% male; 38% Hispanic; median age, 13 y; median body mass index Z-score, 2.43), the median NAFLD Activity Score was 4 (interquartile range, 3–5). In multivariable regression analyses, the skeletal muscle mass index was negatively associated with hepatic steatosis after controlling for confounders (odds ratio, 0.76; 95% confidence interval, 0.62–0.93). Similarly, the appendicular muscle mass index was negatively associated with severity of hepatic steatosis severity (odds ratio, 0.69; 95% confidence interval, 0.53–0.90). In contrast, percentage body fat was not associated with hepatic steatosis. NAFLD Activity Score, lobular inflammation, ballooning scores, and fibrosis stage were not associated with measures of body composition. ConclusionsThere is an inverse association between BIA-based measures of muscle mass and severity of hepatic steatosis in children with NAFLD. BIA data could further inform clinical decision making in this context.