Objective: The relationship between residential altitude and diabetes risk remains unclear, with limited prospective data currently available. This study aims to investigate whether residential altitude is associated with diabetes incidence risk across the full altitude gradient. Design and method: We conducted a population-based study using annual community health examination records (baseline 2017; follow-up through 2024). Participants were adults aged 18 years and older with no prior diagnosis of diabetes and a valid residential address. Elevation was determined using Space Shuttle Radar Topography Mission data (90-meter resolution) and categorized into four groups: below 500 meters, 500–1499 meters (reference), 1500–2499 meters, and 2500 meters and above. Diabetes onset was identified by fasting blood glucose >=7.0 mmol/L, physician diagnosis, or use of antidiabetic medications. We employed Cox proportional hazards regression models, sequentially adjusting for demographic characteristics, lifestyle factors, anthropometric measures, comorbidities, laboratory values, and baseline blood glucose. Overlap weights balanced baseline covariates. Restricted cubic splines explored nonlinear relationships. Results: Among 5,649,250 participants with a median follow-up of 6.62 years (32 million patient-years), 889,830 developed diabetes (incidence rate: 27.77 per 1,000 person-years). After full adjustment, compared with the reference group (500-1499), the hazard ratio was 0.973 (95% CI 0.967–0.979) at altitudes below 500 meters, 0.797 (0.791–0.803) at 1500–2499 meters, and 0.834 (0.810–0.858) at 2500 meters and above. These estimates remained stable after overlapping weighting. The spline model revealed a nonlinear pattern: diabetes risk declined sharply with increasing altitude up to approximately 1500 meters, after which the curve flattened. This negative correlation persisted consistently across subgroups defined by sex, age, body mass index, and abdominal obesity. Conclusions: Among adult residents at elevations ranging from 86 to 3,647 meters, higher altitude was independently associated with a reduced incidence of diabetes. The protective gradient was steepest below 1,500 meters, suggesting a threshold effect requiring further mechanistic investigation.
Nuerbolati et al. (2026) studied this question.