Association between ambient temperature and all-cause hospitalization by functional status in Japan: A time-stratified case-crossover study.

BACKGROUND: Older adults, children, and low-income populations are recognized as vulnerable to climate-related health risks. However, evidence on morbidity risks by functional status remains limited. This study investigated the association between ambient temperature and all-cause hospitalization by functional dependency in Japan. METHODS: We obtained daily hospitalization data from the Diagnosis Procedure Combination database and meteorological data for nine Japanese prefectures between 2012 and 2019, comprising over 10 million admissions. Functional status at admission was assessed using the Barthel Index (0-100), with scores >60 classified as functionally independent and ≤60 as functionally dependent. A time-stratified case-crossover design with conditional quasi-Poisson regression was applied to estimate temperature-hospitalization associations. RESULTS: Among functionally independent individuals, relative risks (RRs) were 0.93 (95% confidence interval [CI]: 0.90, 0.96) at extreme cold (5th percentile) and 1.02 (95% CI: 1.01, 1.04) at extreme heat (95th percentile), with the temperature of peak risk (cold) (TPR [cold]) at the 35th percentile (TPR [cold] RR: 1.08 [1.05, 1.10]). Functionally dependent people showed higher risks at extreme cold (RR: 1.11; 95% CI: 1.07, 1.15), comparable risks at TPR (cold) (23rd percentile), and extreme heat (RR: 1.03; 95% CI: 1.02, 1.04). Compared with independent individuals, functionally dependent individuals had higher cold risks, with ratios of RR of 1.20 (95% CI: 1.14, 1.26) at extreme cold and 1.05 (95% CI: 1.01, 1.09) at TPR (cold). These disparities persisted across age and sex subgroups. CONCLUSIONS: Temperature-related hospitalization risks vary by functional status. Functionally dependent individuals are particularly vulnerable to cold exposure, highlighting the importance of targeted cold-weather preparedness strategies.