Impact of Extreme Ambient Temperatures on Psychiatric Hospital Admissions in an Arid Urban Environment in Iran: A Time-Series Analysis

Background: Climate change is increasing the frequency of extreme temperature events, which may aggravate mental health disorders. While growing evidence links temperature extremes to psychiatric outcomes, limited research has investigated these associations in hot-arid regions, particularly in central Iran.Methods: We conducted a time-series analysis of 7,639 psychiatric hospital admissions from 2014 to 2020, utilizing data from the sole psychiatric facility in Yazd. We applied distributed lag non-linear models with quasi-Poisson regression to examine the associations between daily minimum temperature and hospital admissions over a 7-day lag period. Analyses were performed for all psychiatric diagnoses, with stratified models for major categories, including mood disorders and schizophrenia. Further stratification was conducted by sex, marital status, and age group. Sensitivity analyses incorporated alternative temperature metrics (mean, maximum, and humidex) and additional covariates including precipitation and daylight duration.Results: Extremely hot days (≥99th percentile of maximum temperature) were associated with an elevated risk of psychiatric admissions, with the highest effect observed at lag 1 (RR= 1.08; 95% CI: 0.98–1.20). Mood disorders demonstrated the strongest heat-related association, peaking at lag 0 (RR= 1.31; 95% CI: 1.07–1.59). Subgroup analysis revealed heightened vulnerability among men, married individuals, and younger adults. In contrast, extremely cold temperatures exhibited delayed effects in women and older adults. Results remained robust across alternative temperature metrics and model specifications.Conclusion: Both extreme heat and cold are associated with increased psychiatric hospitalizations in hot-arid climates. These findings underscore the necessity for climate-responsive mental health strategies, particularly targeting socio-demographically vulnerable populations.