Particulate Matter Pollution Of Different Particle Sizes And Stroke Mortality: Estimating The Associations And Attributable Risk In The Pearl River Delta Region

OBJECTIVE:Stroke is one of the most important diseases that threaten human health since the twentieth century.At present,its disability rate has ranked first in worldwide,and its ranking in the cause of death gradually advanced globally,especially in China.Currently,there were many overseas studies have showed the evidences that short-term exposure to particulate matter have effect on stroke mortality,while results were different or even opposite due to diverse researchers,methods,and subjects.In addition,the association between particulate matter and stroke is still uncertain in China,especially with different type of stroke.This study explored the exposure-response relationship between particulate matter and stroke mortality in six cities in the Pearl River Delta region,and further calculated the mortality burden caused by it.METHODS:We collected air pollution data,mortality data and daily meteorological data in the PRD region from2013 to 2016.We used time-series analysis to investigate the associations between different size particle matter and different types of stroke mortality.The generalized additive model was applied to analyze associations and temperature,humidity and other factors were adjusted,we also made a series of sensitive analysis to examined the robustness of the findings.We examined the associations with different lag structures including single-day lag and multi-day lags.We also drawn the concentration-response curves of particulate pollutants and different types of stroke mortality.RESULTS:1.The average concentrations of PM₁₀,PM_2.5-10,PM_2.5,SO₂,NO₂,O₃and CO ranged from 54.10μg/m³ to 66.65μg/m³,from 17.97μg/m³ to24.67μg/m³,and from 32.62μg/m³ to 44.00μg/m³,10.46μg/m³ to23.44μg/m³,32.14μg/m³ to 45.69μg/m³,48.59μg/m³ to 63.67μg/m³ and1.64μg/m³ to 1.99μg/m³ across the six cities from 2013-2016.The daily mean temperature ranged from 21.37℃to 23.29℃and the relative humidity ranged from 74.33%to 81.87%.2.In single-pollutant model,we found that each 10μg/m³ increase of PM_2.5,PM_2.5-10 and PM₁₀（lag03）was associated with an increase of 3.07%,5.72%and 1.88%in overall stroke mortality,an increase of 3.66%,6.82%and 1.88%in ischemic stroke mortality,and an increase of 2.04%,4.04%and 1.03%in hemorrhagic stroke mortality.And we estimated that 4.76%of stroke mortality（corresponding to 2584 stroke mortalities）were attributed to PM_2.5,1.82%stroke mortality（corresponding to 985 stroke mortalities）were attributed to PM_2.5-10,and 3.15%of stroke mortality（corresponding to 1708 stroke mortalities）could be attributed to PM₁₀.3.In sensitivity analysis,we observed statistically significant decrease ERs after adjusting other gaseous pollutants,while when the association between PM_2.5-10,PM₁₀ and hemorrhagic stroke mortality were non-statistically significant.4.In different lag structures,we observed a largest and robust effects of different particle size fractions on subtypes of stroke mortality on lag2and lag05.CONCLUSIONS:Short-term exposure to PM_2.5,PM_2.5-10 and PM₁₀0 were significantly associated stroke mortality,especially PM_2.5.The findings were robustness after conducted a series of sensitivity analyses.We found different mortality burdens from stroke among different particle size fractions of PM pollution,with the largest population attributable mortality and population attributable fraction caused by PM_2.5.