Cause-specific hospital admissions associated with short-term PM_2·5 exposure: a multicountry time-series analysis

Fine particulate matter (PM_2·5) pollution constitutes a leading global health burden. However, multicountry evidence about the impact of PM_2·5 on cause-specific hospitalisations, which are key markers of clinically substantial morbidity, is scarce. This study aimed to examine the association of short-term exposure to ambient PM_2·5 pollution with hospital admissions for specific causes across multiple countries.

We collected 346·5 million hospitalisation records from 949 communities across six countries and territories (Australia, Brazil, Canada, Chile, New Zealand, and Taiwan) between 2000 and 2019. At each location, we fitted quasi-Poisson generalised linear time-series models to estimate PM_2·5-associated hospitalisation risks. These location-specific estimates were then pooled in a meta-analysis. We conducted stratified analyses by country, demographic factors, income, and time period.

Short-term PM_2·5 exposure was significantly associated with increases in all-cause and cause-specific hospitalisations, with effects persisting for up to 21 days. For all-cause admissions, each 10 μg/m³ increase in PM_2·5 concentration was associated with an 8% higher risk of hospitalisation (relative risk [RR] 1·080; 95% CI 1·074-1·086), with substantial heterogeneity across countries and territories. Short-term PM_2·5 exposure was associated with increased hospitalisation risks across all disease categories, with the strongest associations observed for malignant neoplasms (RR 1·108, 95% CI 1·094-1·123), mental disorders (1·100, 1·085-1·114), other neoplasms (1·096, 1·078-1·114), and nervous system diseases (1·085, 1·069-1·102), all exceeding the RR for respiratory diseases (1·067, 1·060-1·074). Attenuated but significant risks were observed in 2010-19, compared with those in 2000-09, for all disease categories. Using a PM_2·5 concentration of 15 μg/m³ according to the WHO daily guideline as reference, attributable measures estimated from the observed PM_2·5-hospitalisation associations declined over time, with the attributable fraction decreasing from 8·9% in 2000-09 to 3·9% in 2010-19 and the cumulative attributable hospitalisation rate decreasing from 11·6 to 3·5 per 1000 population per year.

Our findings revealed that short-term PM_2·5 exposure was associated with persistent increased risks of hospitalisation across diverse settings. Reducing PM_2·5 exposure remains an important preventive measure to lower hospitalisation risk.

Australian Research Council and Australian National Health and Medical Research Council.