Associations of combined triglyceride-glucose and sarcopenia index with incident cardiovascular disease in middle-aged and older Chinese adults

Abstract Cardiovascular disease (CVD) is the leading global cause of mortality, with metabolic dysfunction and sarcopenia synergistically increasing risk in aging populations. We investigated whether cumulative exposure and dynamic trajectories of a novel combined triglyceride-glucose and sarcopenia index (TyG-SI) predict incident CVD in middle-aged and older Chinese adults.We analyzed data from 4,466 participants (≥45 years) from the China Health and Retirement Longitudinal Study (CHARLS) who were free of CVD at baseline (2011–2012). TyG-SI was calculated as TyG index (ln[triglycerides (mg/dL) × glucose (mg/dL)/2]) divided by the sarcopenia index [serum creatinine(mg/L)/serum cystatin C(mg/dL) × 100]. Cumulative TyG-SI (cuTyG-SI) was computed as the product of mean TyG-SI values from 2012 and 2015 and the time interval. K-means cluster analysis identified three distinct TyG-SI trajectory patterns (low-decreasing, moderate-stable, high-decreasing). Cox proportional hazards models and restricted cubic splines (RCS) analysis were used to assessed the associations between TyG-SI and CVD risk. Time-dependent C-statistics, integrated discrimination improvement (IDI), and net reclassification improvement (NRI) assessed predictive performance. Over a median follow-up of 8.22 years, 1,201 incident CVD cases occurred. In fully adjusted models, per 1 SD increase in cuTyG-SI was associated with a 1.10-fold increased CVD risk (hazard ratio [HR] 1.10, 95% CI 1.03–1.17). Participants in the highest cuTyG-SI tertile had 1.32-fold higher risk compared with the lowest tertile (HR1.32, 95% CI 1.12–1.55). Compared with the low-decreasing cluster, both moderate-stable (HR 1.28, 95% CI 1.11–1.48) and high-decreasing clusters (HR 1.30, 95% CI 1.09–1.56) showed significantly elevated CVD risk. Restricted cubic spline analysis revealed an approximately linear positive dose-response correlation (P for nonlinearity = 0.082). The cuTyG-SI demonstrated superior discriminative ability (C-index = 0.613) compared with baseline TyG-SI (0.608), SI (0.606), and TyG (0.603), with significant IDI (0.50–0.70%, all P < 0.05) and NRI (8.6–11.9%, all P < 0.05) improvements. Notably, age and hypertension status significantly modified these associations (P for interaction = 0.008 and 0.004, respectively), with stronger effects observed in participants aged <70 years and those without hypertension. Sensitivity analyses using competitive risk model, logistic regression, excluding dyslipidemia medication users, and omitting 2020 data (COVID-19 pandemic) yielded consistent results. Cumulative exposure and persistent high-decreasing cluster of TyG-SI were independently associated with increased CVD risk. The cuTyG-SI showed superior predictive performance compared with single index, suggesting that integrated assessment of metabolic-muscular dysfunction enhances CVD risk stratification in aging populations.