Comparative performance of insulin resistance-related indices in predicting adverse cardiovascular events among individuals with NAFLD and MASLD: a multi-center cohort study

Background: Metabolic dysfunction-associated steatotic liver disease (MASLD) and non-alcoholic fatty liver disease (NAFLD) are closely linked to insulin resistance and elevated cardiovascular risk, triglyceride-glucose (TyG)-related indices, the atherogenic index of plasma (AIP), and the cardiometabolic index (CMI) have emerged as practical surrogate markers for cardiometabolic risk assessment in these populations. However, it remains unclear whether the associations of these indices with adverse cardiovascular events remain consistent when transitioning from the NAFLD to the newly defined MASLD framework, and a comprehensive comparison of these indices across both diagnostic criteria is lacking. Methods: This study included 4693/3266, 74,173/67,864, and 7823/10,576 individuals with MASLD/NAFLD from the National Health and Nutrition Examination Survey (NHANES), UK Biobank (UKB), and China Pudong cohort, respectively. Multivariate Cox proportional hazards models, restricted cubic spline analyses, and time-dependent receiver operating characteristic curves were employed to assess associations. Linear regression models evaluated relationships between TyG-related indices and cortical/subcortical structural volumes. Mediation analyses examined the role of oxidative stress, phenotypic aging, and inflammatory markers. Results: Most IR-related indices demonstrated nonlinear, predominantly J-shaped associations with cardiovascular disease (CVD) and related mortality, especially within the UKB. TyG-WC optimally predicted 3-year CVD mortality in MASLD/NAFLD across UKB and NHANES, as well as ischemic stroke (IS) in UKB-MASLD. TyG-WHTR was the strongest predictor for CVD mortality in the Pudong cohort and myocardial infarction (MI) in UKB-MASLD. Notably, elevated TyG-WHTR was consistently associated with heightened risks across all endpoints: CVD mortality (NAFLD: UKB: HR 1.60, 95%CI 1.38–1.84, NHANES: 1.90, 1.34–2.70, Pudong: 1.24, 1.10–1.40; MASLD: UKB: 1.59, 1.38–1.82, NHANES: 1.86, 1.36–2.54, Pudong: 1.70, 1.43–2.03), MI (NAFLD: 1.20, 1.09–1.33; MASLD: 1.22, 1.11–1.34), and IS (NAFLD: 1.41, 1.24–1.61; MASLD: 1.39, 1.23–1.57). Structural neuroimaging analyses revealed significant negative correlations between TyG-WC/TyG-WHTR and subcortical volumes (P < 1 × 10⁻⁴). Mediation analyses indicated that oxidative stress, phenotypic aging, and inflammatory markers collectively accounted for 1.4–27% of the observed associations. Conclusions: Insulin Resistance-Related Indices demonstrate robust clinical utility in predicting CVD and mortality in NAFLD/MASLD across three distinct cohorts, with oxidative stress, inflammatory activation, and accelerated aging serving as potential mechanistic pathways. © The Author(s) 2026.