Extreme Temperature and Incident Diabetes Risk Among Middle-Aged and Older Adults in China: A National Longitudinal Cohort Study

Background and Aims: The metabolic consequences of extreme temperature exposure in nondiabetic populations remain poorly understood. This study aimed to examine associations between heatwave and coldwave exposure and incident diabetes mellitus (DM) and impaired glucose tolerance (IGT) in middle-aged and older Chinese adults. Methods: A total of 1803 China Health and Retirement Longitudinal Study participants aged ≥45 years with normoglycemia at baseline were followed from Wave 1 (2011) to Wave 3 (2015). Eighteen extreme-temperature indicators were derived from city-level fifth-generation European Centre for Medium-Range Weather Forecasts atmospheric reanalysis data. Outcomes were classified according to American Diabetes Association criteria. Generalized linear mixed-effects models (GLMMs) were pooled across five imputed datasets, with Bonferroni correction for multiple comparisons (α = 0.0028) and sensitivity analysis adjusting for individual follow-up duration. Results: All nine heatwave indicators showed odds ratios (ORs) < 1.0 for DM and IGT. HT9 (≥97.5th percentile, ≥4 consecutive days) was the sole Bonferroni-significant result: OR = 0.845 (95% confidence interval [CI]: 0.802–0.890). Coldwave indicators showed no consistent associations. Age significantly modified the HT9 effect ( P -interaction = 0.004): adults aged 65–84 showed a stronger inverse association (OR = 0.616) than those aged < 65 (OR = 0.921). Conclusion: Prolonged heatwave exposure was consistently associated with reduced diabetes risk, with pronounced age heterogeneity. Replication in larger prospective studies is warranted.