Evaluation of temperature modifying effect of PM_2.5 on respiratory mortality in high latitude monsoon area

Particulate matter (PM) has been demonstrated to be a significant risk for the onset of respiratory diseases worldwide. We used a generalized additive model (GAM) to evaluate the effect of PM_2.5 on respiratory mortality at both single and multiple lag days based on the daily data of meteorological factors and air pollution levels in Shenyang, northeast China, from 2014 to 2016. We stratified the daily mean temperature into low (≤ P25), lower (P25–P50), higher (P50–P75), and high (≥ P75) to estimate the modifying effect of PM_2.5 on respiratory mortality. The results indicated that for every inter-quartile range (IQR) increased concentration in PM_2.5, respiratory death for the total, males and females increased by 0.35% (95% CI: 0.00%, 0.71%), 0.46% (95% CI: 0.00%, 0.93%), and 0.58% (95% CI: 0.05%, 1.11%) at lag 7, lag 7, and lag 1, respectively. Under the multi-day lags, the largest cumulative relative risks (RRs) were 1.016 (95% CI: 0.990, 1.042) and 1.038 (95% CI: 1.000, 1.077) at lag 0–2 days for the total and females, respectively. The largest death effect of PM_2.5 was found at low temperatures. Females were at the highest risk in higher and higher temperatures, with RRs of 1.021 (95% CI: 0.999, 1.042) and 1.021 (95% CI: 0.993, 1.050) at lag 0 and lag 2, respectively. Males have the highest risk in lower temperatures at lag 1, with an RR of 1.006 (95% CI: 1.000, 1.012). In the multi-pollutant models, the effects of PM_2.5 were mostly increased when NO₂, SO₂, or O₃ were added to the models. The results of the study can provide a scientific basis for public health-friendly urban planning and long-term preventive measures in northeast China.