Ambient Air Quality And The Risk Of Residents' Non-accidental Death:Two Counties Study

Objective:Due to the current lack of study on the impact of ambient air pollution in county-level administrative regions,this paper aims to study and describe the character of ambient air pollution and the effects of ambient air pollution exposure on non-accidental daily mortality risks in two counties,and explore the similarities and differences between the chosen regions and nearby metropolises.This paper also explores the feasibilities of study on the relationship between ambient air pollution and the daily non-accidental mortality of residents at the county level,which aims to provide a scientific basis for environmental management in rural China.Methods:Shunyi District of Beijing and Fengning County of Chengde were selected as study areas,and related data were collected from January 1,2014 to December 31,2015.The air quality data were collected from the local environmental monitoring stations,including daily average levels of specific pollutants such as SO2,NO2,PM10,PM2.5,CO,and O3.In addition,the AQI and the average of 6 air pollutants AQI sub-indices(referred to as AAQI in the following)were used as air pollution index.Meteorological data including daily average relative humidity and daily average temperature were collected from the local weather monitoring station;daily mortality data of Shunyi District and Fengning County residents were obtained from the local disease prevention and control center.Non-accidental death refers to those encoded by ICD10 first letter from A-U.Grade of AAQI is divided using frequency distribution to obtain the corresponding level with AQI.Both AQI and AAQI were used as air pollution indicators to analyze the air quality in Shunyi District and Fengning County and its relationship with non-accidental mortality.After controlling the meteorological factors,long-term trends and week day,the Generalized Additive Models of Poisson Regression was used to establish the concentration-response relationship between air pollution and non-accidental morality.The relative risk(RR)of inhabitants' unintentional death in the both regions was estimated by using the whole curve segment estimation and the binding piecewise linear function.The excess risk(ER)was used to denote the impact of air pollution on non-accidental death.Result:1.The basic situation of air quality,meteorological factors and residents' daily non-accidental death.The average annual temperature of Shunyi District in 2014-2015 was 14.0 ?,which was 40C higher than that in Fengning County.The relative humidity of Shunyi District was 53.3%,which was about 2%lower than that of Fengning County.The two selected regions had similar meteorological factors as their nearby metropolises.The pollution characteristics of Shunyi and Fengning were different.Shunyi District was polluted by particulate matters,while Fengning County was polluted by coal soot and O3 due to summer radiation.In 2014-2015,the daily average of the PM10 and PM2.5 in the Shunyi area was 108.8?g/m3 and 83.3?g/m3 respectively,significantly higher than the corresponding average values of 80.3?g/m3 and 51.3?g/m3 in Fengning County.SO2 and O3 pollution in Fengning County was more severe,and the daily average values were 33.7?g/m3 and 106.9?g/m3 respectively,much higher than the daily average values of 13.8?g/m3 and 59.8?g/m3 in Shunyi District.The daily average values of air pollution of Shunyi District were close to the daily averages of Beijing.The SO2 indicator of Fengning County was significantly higher than that of Chengde while the other air pollution indicators were similar.The daily average number of non-accidental death count was 10.7 in Shunyi District and 5.8 in Fengning County.2.The relationship between AQI and AAQI and non-accidental death in two regionsThere were positive correlations between the AQI and AAQI grades and the daily count of non-accidental death,with AAQI having a stronger correlation in two regions.For Shunyi District,the correlation of different subgroups of non-accidental mortality with AQI(R2 ranges from 0.66 to 0.87)was weaker than the correlation with AAQI(R2 ranges from 0.72 to 0.91).For Fengning County,the correlation between non-accidental mortality and AQI was relatively weak(R2=0.16),while the correlation with AAQI showed positive correlation(R2=0.96).Thus AAQI could reflect comprehansive environmental air quality better.3.The air quality of the two districts and the estimated risk of daily non-accidentalmortalityThe results of time series analysis in Shunyi District showed that an increase of 10?g/m3 of NO2(Lag4)was associated with 1.45%increase in daily non-accidental mortality.Respectively,an increase of 10?g/m3 of PM10(Lag1),PM2.5(Lag01)and O3(Lag01)increased non-accidental mortality rate by 0.47%,0.45%and 1.35%.The results of time series analysis in Fengning County showed that an increase of 10?g/m3 of SO2(Lag01)was associated with 1.90%increase in daily non-accidental mortality.An increase of 10?g/m3 of NO2(Lag01)was associated with 3.95%increase in daily non-accidental mortality.Respectively,an increase of 10?g/m3 of PM10(Lag1),PM2.5(Lag01)and O3(Lag01)caused the non-accidental mortality rate to increase by 1.25%,1.20%and 1.25%.The risk of non-accidental mortality correlated with AAQI was more significant than that with AQI.For Shunyi District,an increase in 10 units of AQI(Lag3)and AAQI(Lag4)was associated with an increase in non-accidental mortality by 0.41%and I.15%;for Fengning County,non-accidental mortality rate increased by 1.17%and 3.41%respectively.The risk in Fengning County was relative higher compared to that in Shunyi District,and also higher than that of related domestic reports and studies.4.Characteristics of concentration-response curve of air quality and residents' daily non-accidental death in two regionsThe curve of concentration-response relationship of air pollution index and inhabitants'non-accidental mortality in Shunyi and Fengning showed a general trend of non-linearity.The shapes of the curve were C-shaped,S-shaped,and J-shaped.The threshold value of SO2 was not observed for neither region,nor was the threshold values of NO2 and CO observed for Shunyi District.Threshold for all other pollutants were observed in the curve.The inflection point of NO2 for Fengning County was located within the range of 11-14?g/m3.The distribution of PM10 was found with in the range of 49?g/m3-53?g/m3,for PM2.5 within the range of 28?g/m3-32?g/m3.The inflection point of CO was located within the range of 0.51-0.59mg/m3,and for O3 with in the range of 68-79?g/m3 and 140-145?g/m3.For Shunyi District,the inflection point for PM10 is located within the range of 81-96?g/m3,for PM2.5 within the range of 58-94?g/m3,and for O3 around 40?g/m3.Conclusion:1.The trend of annual and seasonal changes of ambient air quality in the selected regions in 2014 and 2015 is basically the same as that of nearby metropolises,as well as the pollution intensity.Major pollutants were coal soot and severe O3 pollution(in the summer)in Fengning County,and particulate matters in Shunyi District.The SO2 pollution and O3 pollution in the summer is more sever in Fengning County than in Shunyi District,while NO2,PM10,PM2.5 and CO pollution were lower than those in Shunyi District.2.In general,AAQI is better able to characterize the correlation of comprehensive ambient air quality with inhabitants' non-accidental mortality rate than AQI;at the same time AAQI is more closely related to non-accidental mortality rate.The correlation of AAQI with daily non-accidental mortality rate(R2=0.96)is stronger than that of AQI in Fengning County(R2=0.16)?3.The results of time series analysis show that increases of air quality indexes had increased the inhabitants' non-accidental mortality rate,and differences in risks exist between the two compared regions.The risk in Fengning County was relatively high,which was similar to the results of foreign researches,while the relatively low risk of Shunyi District was more similar to the results of domestic research.4.In this study the concentration of each of the air quality index and non-accidental mortality-response relationship curves were nonlinear.They were mostly C-shaped,S-shaped,and J-shaped;segment risk estimation result is generally higher than that of the entire plot.The relative risk was higher in the low concentration and the narrow range,and the relative risk was lower in the high concentration and wide range.