Association between obesity and insulin resistance-related indices and ischemic stroke: Analysis of data from the UK Biobank

Ischemic stroke is a leading cause of global mortality and disability. While obesity is a major risk factor, conventional metrics such as body mass index (BMI) may not adequately capture the risk associated with central adiposity and related metabolic dysfunction. This study aimed to evaluate the associations between a comprehensive panel of obesity and triglyceride-glucose (TyG)-related indicators and ischemic stroke risk. We analyzed data from 380,637 participants in the UK Biobank prospective cohort and assessed ten indicators, covering obesity [BMI, fat mass index (FMI), body fat percentage (BFP), trunk fat percentage (TFP), waist-to-hip ratio (WHR), waist-to-height ratio (WHtR)] and TyG-obesity composites (TyG-WHR, TyG-WHtR, TyG-FMI, TyG-BMI). Multivariable Cox proportional hazards regression and false discovery rate (FDR) correction were conducted to evaluate the relationships between obesity and TyG-related indicators with ischemic stroke risk. Potential nonlinear associations were explored using restricted cubic spline (RCS) analysis. Subgroup analyses were stratified by sex, hypertension, and diabetes status, with formal interaction testing conducted to evaluate interaction effects. After full adjustment and FDR correction, WHR [hazard ratio (HR) = 1.095, 95% confidence interval (CI): 1.058, 1.134], WHtR (HR = 1.059, 95% CI: 1.030, 1.088), TyG-BMI (HR = 1.034, 95% CI: 1.008, 1.062), TyG-WHR (HR = 1.106, 95% CI: 1.070, 1.144), TyG-WHtR (HR = 1.068, 95% CI: 1.039, 1.098), and TyG-FMI (HR = 1.033, 95% CI: 1.004, 1.064) were associated with increased risk of ischemic stroke. FMI, BFP, TFP, TyG-WHR, and TyG-FMI exhibited non-linear associations with ischemic stroke risk (all P for non-linearity < 0.05). The associations of WHtR, WHR, TyG-WHtR, and TyG-WHR with ischemic stroke risk varied significantly across sex, hypertension, and diabetes status, respectively. Central obesity indicators, particularly WHR and WHtR, and their TyG-based composites (TyG-WHR, TyG-WHtR) had significant predictive values for the risk of ischemic stroke, with their predictive effects significantly modulated by sex, hypertension, and diabetes status. These findings support the integration of central adiposity and insulin resistance metrics into clinical risk assessment, thereby enabling subgroup-specific risk stratification in ischemic stroke prevention.