Exploring BMI's mediating influence on cardiovascular risk correlations with the triglyceride-glucose index: using NHANES and CHARLS cohorts

Objective This investigation employed population-based datasets to elucidate the pathophysiological interplay between triglyceride-glucose index (TyG) and incident cardiovascular disease (CVD), and quantify the extent to which body mass index (BMI) operates as a biological mediator within this association, utilizing a dual-cohort analytical framework. Methods In this study, 17,976 Americans from the NHANES (1999–2020) and 6,218 Chinese from the CHARLS (2011–2020) were included. To investigate the intricate link between the TyG index, BMI and CVD, researchers employed weighted multiple logistic regression, linear regression, restricted cubic spline (RCS) analysis, mediation analysis, and subgroup analysis. Results Among the study population, 1,895 Americans and 1,798 Chinese were diagnosed with cardiovascular disease (CVD). The regression analysis indicated that individuals in the higher quartile of the TyG index had a significantly greater risk of developing CVD (NHANES: P < 0.01, 95% CI: 1.11–1.78; CHARLS: P < 0.001, 95% CI: 1.37–1.89). In both surveys, participants with elevated TyG indices and BMI levels exhibited the highest incidence of CVD. The TyG index significantly affected CVD in both the NHANES and CHARLS cohorts. The total effect in the NHANES cohort was 1.438 × 10 −3 ( P < 0.001), and in the CHARLS cohort, it was 0.007 ( P < 0.001). Conclusions In this study, two independent cross-sectional cohort studies demonstrated significant positive correlations among TyG, BMI, and CVD. Multivariate analyses identified BMI as a partial mediator in the TyG-CVD pathway, with robust effect magnitudes remaining stable after controlling for age, sex, and other confounders.