Body Roundness Index and All‐Cause Mortality in Metabolic Dysfunction‐Associated Steatotic Liver Disease: A Dual‐Cohort Study

ABSTRACT Background Whether associations between the body roundness index (BRI) and mortality differ across population settings in metabolic dysfunction‐associated steatotic liver disease (MASLD) remains unclear. Methods We analysed two nationally representative cohorts: NHANES (1999–2018; n = 7723; US population) and CHARLS (2011–2020; n = 6553; Chinese population). Cox proportional hazards regression with restricted cubic splines was used to assess dose–response relationships. Between‐cohort heterogeneity was evaluated using I 2 statistics. Results In NHANES (mean BMI, 31.99 kg/m 2 ; 924 deaths over median follow‐up of 8.9 years), BRI demonstrated a significant linear association with all‐cause mortality (hazard ratio [HR], 1.18 per SD; 95% confidence interval [CI], 1.049–1.327; p = 0.006). In CHARLS (mean BMI, 25.73 kg/m 2 ; 110 deaths over median follow‐up of 7.3 years), no significant association was observed (HR, 0.91; 95% CI, 0.749–1.101; p = 0.327). Between‐cohort heterogeneity was substantial ( I 2 = 80.7%; p = 0.023). In NHANES, the triglyceride‐glucose (TyG) index was associated with 33.8% (95% CI, 18.2%–52.4%) statistical attenuation of the BRI‐mortality association. The null finding in CHARLS likely reflects insufficient power (14% to detect HR = 1.18) rather than evidence against an association, highlighting the need for adequately powered studies. Conclusions BRI demonstrates a significant linear association with mortality in the US MASLD population. In the Chinese elderly MASLD population, statistical power was insufficient to draw definitive conclusions about BRI‐mortality associations. The observed between‐cohort heterogeneity likely reflects multiple factors, including differences in age structure, obesity phenotype, mortality ascertainment methods, and potentially genuine population‐specific characteristics. These findings underscore the importance of population‐specific validation before extrapolating Western‐derived BRI thresholds to Asian populations.