AHA Scientific Sessions 2026 Abstract · Nov 2, 2026
Metabolic and Hepatic Contributions to Cardiovascular Disease Risk in Metabolic Dysfunction-Associated Steatotic Liver Disease: A Meta-Analytic Structural Equation Modeling Study
By Sethi Y, Padda I, Singhal S et al.
Background: Metabolic dysfunction-associated steatotic liver disease (MASLD) is consistently associated with an increased risk of cardiovascular disease (CVD), the leading cause of death among affected individuals. However, whether this excess risk primarily reflects shared cardiometabolic abnormalities or represents an independent contribution of liver-specific pathobiology remains uncertain. Methods: We conducted a systematic review and meta-analytic structural equation modeling (MASEM) study to quantify the pathways linking metabolic dysfunction-associated steatotic liver disease (MASLD) to cardiovascular disease (CVD). PubMed, Embase, Web of Science, and Scopus were searched from inception through January 2026. Prespecified subgroup analyses evaluated effect modification according to fibrosis severity, geographic region, MASLD ascertainment method, and cardiovascular outcome type. Results: A total of 87 studies involving 5,214,683 participants were included. The pooled standardized total association between MASLD and CVD was β=0.182 (95% CI, 0.154-0.210). Metabolic pathways collectively mediated 74.7% of the observed association (indirect effect β=0.136; 95% CI, 0.112-0.161). Insulin resistance and type 2 diabetes represented the largest mediating component (β=0.056; 95% CI, 0.043-0.070), accounting for 30.8% of the total association, followed by central adiposity (β=0.029; 95% CI, 0.019-0.040), hypertension (β=0.021; 95% CI, 0.013-0.031), atherogenic dyslipidemia (β=0.018; 95% CI, 0.010-0.027), and systemic inflammation (β=0.012; 95% CI, 0.005-0.020). Despite extensive mediation through metabolic pathways, a significant residual direct effect persisted (β=0.046; 95% CI, 0.021-0.071), accounting for approximately one-quarter of the total association. This residual effect increased with fibrosis severity and was greater among South Asian populations than other populations (β difference=0.022; 95% CI, 0.006-0.038; P=0.008). Model fit was excellent (CFI=0.97, RMSEA=0.041, SRMR=0.036). Conclusions: Nearly three-quarters of the association between MASLD and cardiovascular disease appears to be explained by established metabolic risk pathways, particularly insulin resistance, dysglycemia, and excess adiposity. However, a persistent residual association remains after accounting for these intermediaries, suggesting that liver-specific processes contribute independently to cardiovascular risk. These findings support a dual-risk framework in MASLD, in which optimal cardiovascular prevention may require both aggressive metabolic risk-factor modification and interventions targeting progressive liver disease.