Association Of Ambient Fine Particulate Matter And Its Chemical Constituents With Risk Of Parkinson’s Disease

Background and ObjectivesThe health effect of fine particulate matter(aerodynamic diameter≤2.5μm;PM2.5)is of great public health concern.Since 1980s,numerous epidemiological studies have shown a documented relationship between long-term exposure to PM2.5 and mortality,cardiovascular and respiratory disease.However,PM2.5 is complex mixture with different kinds of chemical constituents.It remains unknown which exact chemical constituent contributed to the health effect of PM2.5.As different chemical constituents have different toxicological effects and come from different sources,exploring the association between PM2.5 chemical constituent and health-related outcomes would benefit the etiology of certain disease,and at the same time provide evidence for the control strategy of air pollution.Parkinson’ s disease is the second most common neurodegeneration disease worldwide after Alzheimer’s disease.One of the key points to reduce the disease burden of Parkinson’ s disease is identifying its modifiable environmental risk factors.In recent years,attention has been paid to the association between ambient PM2.5 exposure and risk of Parkinson’s disease.However,current epidemiological evidence is inconclusive,and most of the studies so far coming from the Western countries where the PM2.5 exposure level was relatively low.There is no prospective cohort studies from the Chinese population yet.Herein,we aimed to explore the spatial-temporal distribution of PM2.5 chemical constituents in Zhejiang province and build the individual exposure predicting model of PM2.5 chemical constituents using land-use regression models.The land-use regression(LUR)models would be further applied in environmental epidemiological studies in Zhejiang province.Based on the established general population-based cohort,we would explore the association of long-term exposure to PM2.5 and its chemical constituent with the incident of Parkinson’ s disease.Materials and methodsPM2.5 chemical constituent(sulfate,nitrate,ammonium,chlorine,aluminum,arsenic,beryllium,cadmium,mercury,lead,manganese,nickel,selenium,antimony,thallium)monitoring data in Zhejiang province from Jan 1st,2015 to December 31th,2019 were used in the current study.Spatial-temporal LUR models for each chemical constituent were generated via generalized additive model using land cover type data,traffic data,population density,altitude and meteorological data as predictors.Hold-out cross-validation,K-fold cross-validation and leave one out cross-validation were used to assess the accuracy and robustness of the LUR models.We used data from a prospective cohort on 47,516 participants recruited between June 2015 to January 2018 in Yinzhou district,Ningbo,China.Incident Parkinson’s disease were retrieved through the Regional health information system.Long-term exposure to PM2.5 and PM10 and nitrogen dioxide(NO2)estimated by land-use regression models,road proximity and surrounding green assessed by Normalized Difference Vegetation Index(NDVI)were calculated based on the residential address for each participant.Cox proportional hazard models were used to analyze the individual and joint effects of air pollution,road proximity,and surrounding green on PD,adjusted for age,sex,education level,income,smoking status,alcohol consumption status,tea consumption status and physical activity.Two-exposure models were also conducted to test the potential independent effect of PM2.5 on PD.We further calculated the long-term exposure to PM2.5 chemical constituent by applying the LUR models established in the first part of the study.Cox proportional hazard models were used to analyze the association between PM2.5 chemical constituents and risk of PD.False Discovery Rate(FDR)method was used to correct for multiple comparison.We further used two-exposure models(adjusted for PM2.5 mass or PM2.5 mass residuals)to explore the potential independent effect of certain PM2.5 chemical constituent.ResultDuring the study period of 2015 to 2019,the major components of PM2.5 in Zhejiang Province were sulfate,nitrate,aluminum,lead,manganese.On the time trend,the distribution of the chemical composition of PM2.5 fluctuates and decreases,and the spatial distribution shows that the pollution of inland cities is higher than that of coastal cities.Concentrations of PM2.5 chemical composition were general higher in winter compared with that in summer.The LUR model performances for ammonium,chlorine,cadmium,thallium were good with R-squares higher than 0.7,model performance for sulfate,beryllium,arsenic,lead,manganese,selenium were moderate,with R-squares ranging from 0.5 to 0.7,model performance for antimony,aluminum,mercury,nickel were not good with R-squares 0.45,0.44,0.44,0.49,respectively.R-squares from cross-validation were generally similar to the model adjusted R-squares,indicating the models were robust.A total of 46,839 participants were included in the prospective cohort analysis with a median follow-up period of 3.5 years.206 incident cases of Parkinson’s disease were identified.After adjusted for sex,age,education level,income,BMI,smoking status,alcohol consumption,tea consumption and physical activity,long-term exposure to PM2.5 was found to be associated with risk of PD(HR=1.41(95%CI:1.14-1.75)per interquartile range increment).Sensitivity analyses using different exposure time windows(one-year average concentrations preceding the baseline investigation,two-year average concentrations preceding the baseline investigation and ten-year average concentrations)generated essentially the same results.Subsequent analysis on the 15 kinds of PM2.5 chemical constituents showed that long-term exposure to ammonium,cadmium,lead were positively associated with risk of PD(HR=1.13(95%CI:1.04-1.23),1.21(95%CI:1.05-1.39),1.11(95%CI:1.00-1.22),respectively).The association estimates remained significant after adjusted for multiple comparison and adjusted for PM2.5 mass and PM2.5 mass-constituent residuals.Subgroup analysis showed that older adults(65 years and older)might be more susceptible to the adverse effect of certain PM2.5 chemical constituents.ConclusionIn the current study,we established LUR models for PM2.5 chemical constituents with high spatial resolution using monitoring data in Zhejiang Province,and explore the association between long-term exposure to PM2.5 and its chemical constituent in a prospective cohort.The major conclusions were as follows:(1)During the study period of 2015 to 2019,the concentrations of PM2.5 chemical constituent generally decreased;(2)Long-term exposure to ambient PM2.5 is associated with increased risk of Parkinson’s disease,which is independent of other air pollutants(PM10 and NO2),traffic exposure(road proximity)and surrounding greenness;(3)Ammonium,cadmium,lead within PM2.5 contributed to the health effect of PM2.5 on Parkinson’s disease.