A Comparison of the Predictive Value of 12 Body Composition Markers for Metabolic Dysfunction-Associated Steatotic Liver Disease, At-Risk Metabolic Dysfunction-Associated Steatohepatitis, and Increased Liver Stiffness in a General Population Setting.

Abstract

Adipose tissue is a key mediator of metabolic dysfunction-associated steatotic liver disease (MASLD) development and progression into metabolic dysfunction-associated steatohepatitis (MASH) and fibrosis. Since direct comparisons of body composition parameters are lacking, we here investigate 12 different body composition parameters. Adult participants from National Health and Nutrition Examination Survey 2017-2023 with liver health data were included. Exclusion criteria were age older than 80 years, excessive alcohol (>60 g/d), viral hepatitis, and missing anthropometrics. MASLD was defined as controlled attenuation parameter ≥275 dB/m with metabolic dysfunction, MASH as FibroScan-aspartate aminotransferase ≥0.35, and increased liver stiffness measurement (LSM) as ≥8 kPa. Predictive performance of 12 body composition parameters was assessed using area under the curve analysis. Predicted probabilities of outcomes were visualized for standardized parameters, and nonlinearity was assessed through restricted cubic splines. Among 11,579 participants (age 51 [35-63], 47% male), 41% had MASLD, 6.5% at-risk MASH, and 9.9% increased LSM. Waist circumference (WC) and not BMI or waist-to-height ratio obtained the highest area under the curve for MASLD (0.82), at-risk MASH (0.73), and increased LSM (0.75) outperforming or equaling all other indices across subgroups. Associations between WC and MASLD were nonlinear, with slight risk saturation beyond 100 cm; at-risk MASH was linearly associated across the entire spectrum; and increased LSM risk rose only after WC >100 cm. In the general population, MASLD and MASH risk increased even when WC < 100 cm, while increased LSM risk was increasing only >100 cm. Although relatively minor differences, WC consistently demonstrated the highest predictive value for MASLD, at-risk MASH, and increased LSM and therefore most suited for MASLD diagnosis, management, and risk stratification.