Metabolic score for insulin resistance and the incidence of cardiovascular disease: a meta-analysis of cohort studies

TL;DR

This study finds that a metabolic score for insulin resistance is linked to higher risks of cardiovascular disease, heart disease, and stroke, with nonlinear risk thresholds.

Contribution

The study is the first to systematically evaluate the association between METS-IR and incident cardiovascular diseases using a meta-analysis of cohort studies.

Findings

Higher METS-IR scores are associated with increased risks of composite CVD, CAD, and stroke.

Nonlinear dose-response relationships were observed for CAD and stroke with inflection points at specific METS-IR values.

METS-IR independently predicts cardiovascular risk despite moderate-to-high heterogeneity across studies.

Abstract

Cardiovascular disease (CVD) remains the leading global cause of mortality, with insulin resistance as a pivotal metabolic risk factor that promotes endothelial dysfunction, inflammation, and atherosclerosis via mechanisms such as impaired nitric oxide signaling and enhanced oxidative stress. The metabolic score for insulin resistance (METS-IR), a non-insulin-based index derived from fasting blood glucose, triglycerides, high-density lipoprotein cholesterol, and body mass index, offers a practical surrogate for assessing insulin sensitivity. However, its association with incident CVD has not been systematically evaluated in a meta-analysis. This meta-analysis aimed to quantify the relationship between baseline METS-IR and the incidence of composite CVD, coronary artery disease (CAD), and stroke in adults without baseline CVD, including categorical, continuous, and dose-response analyses. We searched PubMed, EMBASE, Cochrane Library, and Web of Science up to August 2, 2025, for cohort studies. Hazard ratios (HRs) were pooled using random-effects models to account for heterogeneity for highest versus lowest METS-IR categories and per standard deviation (SD) increment. Nonlinear dose-response relationships were modeled with restricted cubic splines. Heterogeneity, sensitivity, and publication bias were assessed. Eight cohort studies involving 437,283 participants were included. Highest vs. lowest METS-IR was associated with increased risks (HR [95% CI]; I²): composite CVD (1.65 [1.36-2.02]; 85.6%), CAD (1.82 [1.50-2.20]; 59.7%), stroke (1.47 [1.19-1.83]; 76.3%). Per SD increment: composite CVD (1.16 [1.10-1.22]; 70.7%), CAD (1.18 [1.11-1.25]; 52.4%), stroke (1.13 [1.06-1.19]; 67.9%). Dose-response analyses revealed a nonlinear association for CAD (P for nonlinearity: 0.011), marginal nonlinearity for stroke (P: 0.072), and suggested nonlinearity for composite CVD (P: 0.145), with inflection points at METS-IR values of 40.56 (composite CVD), 38.24 (CAD), and 48.88 (stroke), beyond which risks appeared to accelerate. Elevated METS-IR independently predicts higher incidence of composite CVD, CAD, and stroke with nonlinear thresholds for CAD, marginal nonlinear thresholds for stroke, and potential nonlinear thresholds for composite CVD, despite moderate-to-high heterogeneity, supporting its integration into risk stratification and preventive strategies for metabolic health management.

https://www.crd.york.ac.uk/PROSPERO/display_record.php?ID=CRD420251104293, identifier CRD420251104293.

Figures