Hourly association between ambient PM2.5 and emergency ambulance dispatches in 11 cities in Japan

Abstract

Numerous epidemiological studies have demonstrated that short-term exposure to ambient PM2.5 increases mortality and morbidity. Investigating the association using hourly ambient PM2.5 exposure may provide important insights, as current evidence is limited mostly to daily lag term. This study aimed to investigate the hourly association between ambient PM2.5 concentrations and all-cause emergency ambulance dispatches (EAD) in 11 cities in Japan. We used a time-stratified case-crossover design and examined the hourly lags of ambient PM2.5 up to 24 h (unconditional distributed lags and moving average lags) using a conditional Poisson regression model. A significant increase in all-cause EAD was observed at lag 0 h [relative risk (RR): 1.0037 (95% CI: 1.0000, 1.0074)] and all moving average lags. The highest RR was observed within the first 6 h (at lag 0–5 h) [RR: 1.0091 (95% CI: 1.0068, 1.0114)], with a slight ascending pattern. This was followed by a descending pattern at lags 0–11, 0–17, and 0–23 h, but significant positive RR was observed even at lag 0–23 h, when the lowest RR was observed [RR: 1.0072 (95% CI: 1.0044, 1.0100)]. Though similar pattern was observed among the elderly, a different pattern was observed among the children (gradually ascending pattern). We conclude that all-cause EAD could be triggered by ambient PM2.5 exposure with very short lags.