Triglyceride–glucose body roundness index and risks of incident cardiovascular disease and cardiometabolic multimorbidity: evidence from prospective cohorts in China and England

Cardiovascular disease (CVD) and cardiometabolic multimorbidity (CMM) impose substantial health burdens. The triglyceride–glucose body roundness index (TyG-BRI) integrates metabolic dysfunction reflected by TyG with body-shape information reflected by BRI, but its associations with incident CVD and CMM and its comparative predictive performance across populations remain uncertain. We included 8,747 participants free of CVD or CMM at baseline, comprising 6,496 from CHARLS and 2,251 from ELSA. TyG-BRI was evaluated continuously and by quartiles. Among 5,721 participants with two measurements, K-means clustering identified low, intermediate, and high two-wave level groups. Cohort-stratified Cox models and restricted cubic splines assessed associations. Four-year Cox models developed in CHARLS were externally evaluated in ELSA and compared with models incorporating TyG, BRI, TyG-BMI, or TyG-WC. Competing-risk and sensitivity analyses assessed robustness. During follow-up, 1,401 participants developed CVD and 1,151 developed CMM. Compared with the lowest TyG-BRI quartile, the highest quartile was associated with higher risks of CVD (HR, 1.666; 95% CI, 1.424–1.949) and CMM (HR, 3.958; 95% CI, 3.260–4.805). Corresponding HRs comparing the high with the low two-wave level group were 1.678 (95% CI, 1.400–2.011) and 3.354 (95% CI, 2.765–4.070). The association was approximately linear for CVD but nonlinear for CMM and was stronger for CMM among participants aged < 60 years after multiple-testing correction. Findings were generally consistent in competing-risk and sensitivity analyses. In ELSA, TyG-BRI had modest discrimination for CVD (AUC, 0.598) and better discrimination for CMM (AUC, 0.701). For CMM, it improved discrimination over the Base model and BRI, although TyG-WC had the highest AUC. Updating the baseline hazard substantially improved calibration. Higher TyG-BRI was associated with incident CVD and, more strongly, CMM across baseline and repeated assessments. TyG-BRI may provide complementary information on cardiometabolic risk but remains a candidate marker requiring further population-specific validation and recalibration.