High BMI drives diabetes burden and risk: evidence from global trends and multi-country cohorts

High body mass index (BMI) drives diabetes, but evidence linking population burden and individual risk across settings is limited. To characterize high-BMI-attributable diabetes burden (1990–2023) and BMI-associated incident diabetes risk across diverse cohorts. Using GBD 2023, we quantified high-BMI-attributable deaths and disability-adjusted life years (DALYs) from 1990 to 2023 in six countries. We analyzed 42,280 diabetes-free adults aged ≥ 45 years from six cohorts (CHARLS, HRS, ELSA, KLoSA, MHAS, CRELES; follow-up 3.5–12.0 years). Cox models estimated hazard ratios (HRs) per 1 kg/m² BMI increase with adjustment; restricted cubic splines assessed dose-response. Absolute DALY counts rose 60.6%-441.5%; age-standardized DALY rates declined in Korea (-5.1%) but increased in Costa Rica (+ 54.2%). Over follow-up, 5,987 incident diabetes cases occurred. Each 1 kg/m² higher BMI was associated with 7%-14% higher diabetes risk (all P < 0.001). Thresholds were lower in Asian cohorts (~ 23.0 kg/m²) than Western cohorts (25.4–27.2 kg/m²). Twenty-year cumulative incidence was 39.5% in the highest versus 13.9% in the lowest BMI tertile. High-BMI-attributable diabetes burden is increasing, and BMI predicts incident diabetes with population-specific thresholds, supporting tailored prevention strategies.