Comparison of nine insulin resistance surrogates for predicting cardiovascular disease: a cohort study

Insulin resistance (IR) is an established independent risk factor for cardiovascular disease (CVD). Although numerous simple surrogate indicators for IR have been proposed, their comparative predictive utility for CVD remains unclear. This study aimed to evaluate the associations between nine IR surrogate indicators and incident CVD and to comparatively assess their predictive capacities using nationally representative data from China.

7,662 participants without CVD from the China Health and Retirement Longitudinal Study (CHARLS) were included in the study. Nine IR surrogate measures including triglyceride-glucose (TyG) index, triglyceride to high-density lipoprotein ratio (TG/HDL), metabolic score for insulin resistance (METS-IR), Chinese visceral adiposity index (CVAI), lipid accumulation product (LAP), atherogenic index of plasma (AIP), triglyceride glucose-body mass index (TyG-BMI), triglyceride glucose-waist circumference (TyG-WC), and triglyceride glucose-waist-to-height ratio (TyG-WHtR) were calculated. Cox model and restricted cubic spline model were used to estimate the relationships between distinct IR surrogates and incident CVD. We also computed the time-dependent Harrell's concordance index (C-index) to compare the predictive performance of IR surrogates.

After a mean follow-up duration of 8.2 years, a total of 1,906 individuals developed CVD. The full adjusted cox model revealed that per SD increase in all IR indicators was significantly associated with elevated CVD risk, with the hazard ratio (95%CI) of 1.09 (1.04-1.14) for TyG; 1.05 (1.01-1.10) for TG/HDL-C; 1.09 (1.03-1.14) for METS-IR; 1.13 (1.08-1.19) for CVAI; 1.07 (1.03-1.12) for LAP; 1.08 (1.03-1.13) for AIP; 1.10 (1.05-1.15) for TyG-BMI; 1.11 (1.06-1.16) for TyG-WC; and 1.11 (1.05-1.16) for TyG-WHtR. Predictive performance analysis showed TyG had the highest C-index of 0.742 (95% CI, 0.737-0.747).

Among nine IR surrogates, the TyG index exhibited the highest predictive performance for incident CVD in Chinese middle-aged and older adults. Acknowledging limitations such as the observational design and self-reported outcomes, our findings support the TyG index as a simple, powerful, and clinically accessible tool for early CVD risk prediction.