As climate change accelerates, extreme heat events are rising in frequency and intensity, prompting concerns about mental health impacts. However, evidence linking heat to psychiatric emergencies remains mixed, partly due to methodological heterogeneity and limited attention to contextual factors like school holidays. This study investigated the association between ambient temperature - including thermal comfort indices - and emergency visits for psychotic and mood disorders, while accounting for school holiday periods. We used a time-stratified case-crossover design over 11 years of data from a psychiatric emergency department in suburban France. High- and extreme-heat days (above the 90th and 99th percentiles) were compared to median-temperature days using four indicators: daily mean and nighttime temperatures, and two thermal comfort indices (Steadman and humidex). Poisson regression models estimated short- (0-7 days) and long-term (0-21 days) associations, adjusted for school holidays and corrected for multiple comparisons using the False Discovery Rate (FDR). Bipolar disorders visits significantly decreased (-8.6 % to -19.7 %) during high- and extreme-heat periods in both the short and long term. Before FDR correction, depressive disorders visits increased (+7.7 %) in the long term with humidex during extreme heat. Psychotic disorders visits rose in the short term (+14.1 % to +25.2 %) across all metrics except humidex. Adjusting for school holidays reversed the association for mood disorders and nullified that for psychotic disorders. These results underscore that psychiatric responses to heat vary by disorder type, exposure duration, and thermal metric, and highlight the importance of accounting for contextual factors, such as holidays.
Keywords: Bipolar disorder; Depression; High temperature; Hospital emergency department; Humidex; Psychotic disorders; Thermal comfort index.
Copyright © 2025. Published by Elsevier B.V.