Reassessing of the incremental predictive value of triglyceride-glucose index beyond traditional cardiovascular risk factors: a cohort study based on the China health and nutrition survey and meta-analysis

Why clinicians should know about it

• Picked for Family Practice (paper of the day, 9 August 2026).

Abstract

BACKGROUND: This study aims to explore whether incorporating the triglyceride-glucose (TyG) index into the Framingham Risk Score (FRS) can enhance discriminative performance using raw data from the China Health and Nutrition Survey (CHNS) cohort, and to systematically quantify its incremental predictive value through a meta-analysis. METHODS: Participants were adults aged 45 to 80 years without baseline cardiovascular disease (CVD). Incident CVD was defined as new diagnoses of myocardial infarction or stroke. Cox regression was used to examine the association between TyG and CVD, and the results were expressed as hazard ratios (HR) with 95% confidence intervals (CI). While the incremental predictive value of TyG added to FRS was assessed via C-index, integrated discrimination improvement (IDI) and net reclassification improvement (NRI). Restricted cubic splines in Poisson models explored non-linearity. For the meta-analysis, we included effect estimates for the incremental value of the TyG index. RESULTS: The highest TyG quartile was significantly associated with higher risks of CVD (HR = 2.073; 95% CI: 1.233 to 3.488) and stroke (HR = 2.078; 95% CI: 1.428 to 5.521), with a non-linear dose-response relationship (P-nonlinear < 0.001). However, at the level of incremental predictive value, adding continuous TyG to FRS yielded a minimal C-index increase (from 0.731 to 0.732; P = 0.559). Adding TyG quartiles improved some reclassification metrics (IDI and NRI) despite no significant C-index gain. Subgroup analyses showed no significant interactions. For the meta-analysis, Pooling five cohort studies (CVD incidence: 3.3% to 14.4%) demonstrated a minimal overall C-index gain of 0.0004 (95% CI: -0.0092 to 0.0100). Sensitivity analysis revealed that this result was heavily influenced by the Niloofar Barzegar 2020; its exclusion increased the pooled C-index gain to 0.0051 (95% CI: 0.0007 to 0.0094). According to the Grading of Recommendations Assessment, Development, and Evaluation framework, the overall certainty of evidence for the incremental C-index change was rated as very low. CONCLUSIONS: While the TyG index shows a significant statistical association with CVD risk, its incremental predictive value beyond established FRS factors is minimal at the cohort level and negligible at the meta-analytic level. These findings suggest that the TyG index, despite being a significant risk marker, may not materially improve the predictive performance of existing cardiovascular risk scores.

Abstract as published, via PubMed.