Short‐Term PM2.5 Exposure and Daily Cancer Mortality Across 47 Japanese Prefectures: A Nationwide Time‐Stratified Case‐Crossover Study, 2013–2022

ABSTRACT Short‐term exposure to fine particulate matter (PM2.5) is an established risk factor for cardiopulmonary mortality, but its acute association with cancer mortality remains poorly characterized. We examined associations between short‐term PM2.5 fluctuations and daily all‐cancer and site‐specific mortality across all 47 Japanese prefectures from January 1, 2013, to December 31, 2022. Using a nationwide time‐stratified case‐crossover design, we fitted prefecture‐specific conditional quasi‐Poisson regression models and pooled the estimates by random‐effects meta‐analysis, expressing associations as the excess relative risk (ERR) per 10 μg/m 3 increase in PM2.5 (lag 0–2 days). Sensitivity analyses varied lag structures, meteorological spline specifications, and co‐pollutant adjustment. A 10 μg/m 3 increase in short‐term PM2.5 was associated with a 0.88% increase in all‐cancer mortality (95% CI: 0.57–1.18; p < 0.001). The association was consistent across sex, age, and primary site (all interaction p > 0.40), with significant estimates for breast (ERR 2.06%), colorectal (1.06%), stomach (0.81%), and lung (0.77%) cancers. Associations attenuated with longer moving‐average windows and were no longer statistically significant from lag 0–5 onward. A measurable fraction of cancer deaths was attributable to PM2.5 above the WHO 24‐h guideline (15 μg/m 3 ), with the largest fractions tracking prefecture‐specific coefficients rather than ambient concentration. These findings indicate that short‐term PM2.5 exposure is associated with higher daily mortality across diverse cancer types in Japan, even at relatively low ambient levels. Because the individual‐level effect size is small and attenuated to non‐significance after adjustment for either NO 2 or SO 2 , these findings are best interpreted as supporting population‐level air‐quality measures rather than individual clinical intervention.