Climate hazards and mental health hospitalizations in China: socioeconomic disparities, demographic drivers, and adaptation – a nationwide population-based study

Background: Mental and behavioural disorders account for a large and growing share of the global disease burden, yet evidence on hazard-specific mental health risks and adaptation remains limited. We aimed to systematically evaluate disease-specific hospitalization risks and attributable burdens following multiple climate hazards in China, and examined whether socioeconomic conditions, demographic change, and historical hazard experience modified these risks. Methods: We conducted a retrospective observational study in 955 counties in China (2016–2023) using 392,924 daily disease-specific hospitalizations for mental and behavioural disorders from the Chinese Multiple County (CMC) Hospital Network. Storm, flood, and tropical cyclone events were linked to hospitalizations, and propensity score matching aided difference-in-differences (PSM-DID) models were used to estimate post–hazard excess relative risks across eight diagnostic categories. We then estimated hazard-attributable mental health hospitalizations and used counterfactual decomposition analysis to quantify the contribution of population aging and growth to hazard-attributable burden. Findings: Schizophrenia and delusion, mood disorders, and stress-related disorders showed stronger associations with climate hazards. Floods were associated with broader and longer-lasting increases in mental health hospitalizations, with elevated admissions persisting for 6–8 weeks after events. Averaged across 2016–2023, storms, floods, and tropical cyclones collectively generated approximately 117,717 (93,667, 147,158) additional mental and behavioural hospitalizations per year in China, 41.05% of which occurred in children, adolescents, and older adults. Population aging and growth accounted for 1.20–5.81% of hazard-attributable hospitalizations, depending on hazard type. Socioeconomically developed counties (quartile 4) experienced 16.18–24.96% lower hazard-attributable hospitalization risks than less developed counties (quartiles 1–3). Counties with the highest historical hazard frequency (quartile 4) showed 11.30–14.29% lower impacts from storms and tropical cyclones than counties with lower historical frequency (quartiles 1–3), whereas no such adaptive advantage was observed for floods. This adaptation observation was more pronounced in socioeconomically developed settings. Interpretation: Storms, floods, and tropical cyclones imposed substantial and inequitably distributed mental health burdens. Climate adaptation policies should prioritize vulnerable populations and less developed regions, leveraging both socioeconomic development and hazard preparedness to promote mental health equity and sustainability under escalating climate threats. Funding: National Natural Science Foundation of China, National Key Research & Development Program of Ministry of Science and Technology of China, Beijing Municipal Ecology and Environment Bureau. © 2026 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license. http://creativecommons.org/licenses/by-nc-nd/4.0/