TyHGB and CVD high-risk stratification: nonlinear association and discrimination in the ChinaHEART Luohe study.

Introduction: TyHGB is a composite metabolic index integrating triglyceride-to-high-density lipoprotein cholesterol ratio, fasting plasma glucose, and body mass index, but its association with World Health Organization (WHO)-defined cardiovascular disease (CVD) high-risk status in community settings remains unclear. Methods: We analyzed 6,755 adults from a community-based ChinaHEART Luohe screening cohort (2021-2022). CVD high-risk was defined as an estimated 10-year WHO CVD risk ≥20%. TyHGB was examined per 1-unit increase and as high versus low using a receiver operating characteristic (ROC)-derived cutoff (Youden index; approximately 7.6). Logistic regression and restricted cubic splines were used to assess associations and potential nonlinearity. Discrimination was evaluated using ROC curves with DeLong's tests, and incremental value beyond conventional risk factors was assessed using integrated discrimination improvement (IDI) and net reclassification improvement (NRI). Results: The prevalence of CVD high-risk was 22%. TyHGB was independently associated with CVD high-risk (fully adjusted OR 1.131 per 1-unit increase), and high TyHGB (≥7.6) was associated with higher odds (OR 1.642). In unadjusted analyses, TyHGB showed higher discrimination than TyG (AUC 0.624 vs 0.585). Adding TyHGB to the base model modestly improved reclassification (IDI 0.003; NRI 0.216). TyHGB was also retained in exploratory machine-learning analyses. Discussion: TyHGB was positively associated with WHO-defined CVD high-risk and provided modest incremental discrimination and reclassification for cross-sectional high-risk classification beyond conventional risk factors, supported by sensitivity analyses. These findings relate to estimated risk status rather than adjudicated incident CVD events.