Association Between Personal PM_2.5 Exposure And Lung Function In Wuhan Community Residents And The Role Of DNA Methylation

Fine particulate matter(the particulate with aerodynamic diameter?2.5?m,PM_2.5),is a widely concerned environmental pollutant in recent years.Ambient PM_2.5 can come from natural sources such as forest fire,volcanic eruption,and sea salt,and anthropogenic activities such as industrial emission,coal combustion,traffic exhaust,construction,and road dust.PM_2.5 can suspend in air and be inhaled through the respiratory tract,then go deep into the bronchioles and alveoli,damage the airway epithelial cells,and migrate into the circulation system,causing adverse health effects.The respiratory system is always the first to be damaged by PM_2.5 exposure.Lung function parameters,as sensitive indicators for respiratory health,can be used for identifying early pathological changes of respiratory system.The factors that affect lung function include not only individual characteristics such as age,gender,height,and weight,but also environmental factors such as particulate matter exposure.The effect and mechanism of PM_2.5 on lung function are still not fully understood which need to be studied.Published literatures have found the association between PM_2.5 exposure and lung function decline.However,most of the studies use ambient PM_2.5 concentrations as proxies of personal exposure,such as concentrations from fixed monitor stations or land use regression model,regardless of the influence of personal lifestyles such as smoking and cooking.Moreover,the current studies are mostly conducted in areas with relatively low PM_2.5 concentrations,for example Europe and the United States.While,most developing countries including China still have relatively high PM_2.5 concentrations.Therefore,it is necessary to assess the association between personal PM_2.5 exposure and lung function.Oxidative stress is thought to play an important role in the pathological damage caused by PM_2.5 exposure.PM_2.5 exposure can lead to mitochondrial dysfunction in alveolar type II epithelial cells,coupled with energy metabolism disorders and the production of reactive oxygen species,inducing the aggravation of inflammatory responses,activating a variety of cell death pathways,and leading to autophagy and apoptosis.However,the role of oxidative stress in the relationship between PM_2.5exposure and lung function is still not clear.In recent years,epigenetic changes such as DNA methylation are linked to adverse health impact caused by environmental pollutants,and the associations of epigenetic changes such as DNA methylation with lung function decline are also of concern.With the development of epigenetic detection technology,DNA methylation at the genome-wide level modified by PM_2.5 exposure can now be analyzed,while the role of DNA methylation at the genome-wide level in the association between PM_2.5 and lung function is still poorly studied.The present study was conducted among Wuhan community residents from the Wuhan-Zhuhai cohort,who were recruited in April to May,2011,and followed up in 2014and 2017.A panel among Wuhan community residents was selected to determine the personal PM_2.5 concentrations,and evaluate the impact factors and fit the estimated model for personal PM_2.5 exposure.The estimated model was applied among all Wuhan residents.Lung function and urinary oxidative damage makers,8-hydroxy-2?-deoxyguanosine(8-OHd G)and 8-iso-prostaglandin-F_2?(8-iso-PGF_2?)concentrations were determined to evaluate the short-term and long-term effects of personal PM_2.5 exposure on lung function,and the mediating role of oxidative stress.Last,a total of 402 never smokers were selected to evaluate the associations between PM_2.5 exposure and DNA methylation sites,and the associations between PM_2.5 related sites and lung function.This study was approved by the Ethics and Human Subject Committee of Tongji Medical College,Huazhong University of Science and Technology.All participants gave written informed consent.Quality control system was used to ensure the quality of data collection and laboratory detection This study includes three parts as follows:Part ? Factors and Estimation of personal PM_2.5 exposure in Wuhan community residentsObjectives:To explore the influence of ambient PM_2.5 concentrations,meteorological conditions,lifestyles,living conditions and occupational dust exposure on personal PM_2.5exposure and to establish an estimated model for personal PM_2.5 exposure in Wuhan community residents.Methods:A total of 191 residents living in Wuhan communities were included.Personal24-hour PM_2.5 concentrations were monitored using personal PM_2.5 samplers.Structured questionnaires and 24-hour activity logs were used to collect information including lifestyles(smoking,passive smoking,cooking,outdoor time and traffic exposure time),living conditions(distance from major roads,floor of residence),and occupational dust exposure,etc.Meanwhile,ambient PM_2.5 concentrations from fixed monitor stations and meteorological conditions in the same period were collected.Pearson correlation coefficients(r),Spearman correlation coefficients(r_s)or point biserial correlation coefficients(r_pb)were used to evaluate the impact factors of personal PM_2.5 exposure.A mixed linear model was used to fit the estimated model,and the stepwise selection was used for variables selection.Variable with variance inflation factor more than 10 was thought to be multicollinearity and was deleted.R²,10-fold cross validation(CV)R² and the ratio of estimated value to measured value were used to evaluate the model.All analyses were performed using SAS(Version 9.4)or R software(Version 3.6.1).Results:In this study,the mean personal PM_2.5 concentration of Wuhan residents was156.85?g/m³,with a range from 14.85 to 443.21?g/m³,which was about twice as high as the ambient PM_2.5 concentration from fixed monitor stations(17.39 to 242.55?g/m³).Personal PM_2.5 concentration and ambient PM_2.5 concentration are moderately positively correlated with a Pearson correlation coefficient of 0.57(P<0.001).In addition,personal PM_2.5 exposure was positively related to current smoking(r_pb=0.17),cigarette consumption per day(r_s=0.20)and passive smoking(r_pb=0.32),while negatively related to outdoor time(r_s=-0.15),atmospheric pressure(r_s=-0.28),relative humidity(r_s=-0.21),and average wind speed(r_s=-0.46).No significant correlation was observed of personal PM_2.5 exposure with cooking at home,domestic fuel,traffic exposure time,distance from major roads,floor of residence,and occupational dust exposure.In final,seven factors were selected into the estimated model,including ambient PM_2.5 concentration,cigarette consumption per day,passive smoking,the difference between atmospheric pressure and standard atmospheric pressure,temperature,relative humidity,and average wind speed.The final model explained 60.4%variation of personal PM_2.5 concentration.And R² of10-fold CV is 0.654.The ratio of estimated value to measured value were around 1,especially when the measured values more than 50?g/m³.Conclusions:The impact factors for personal PM_2.5 exposure included smoking,passive smoking,outdoor time,temperature,relative humidity,atmospheric pressure,and average wind speed.The final model explained 60.4%of the variation of personal PM_2.5concentration.Part ? Associations between short-term and long-term personal PM_2.5 exposure and lung function,and the mediating role of oxidative stress in community residentsObjectives:To assess the associations between short-term and long-term personal PM_2.5exposure and lung function in Wuhan community residents,and to explore the role of oxidative stress in such associations.Methods:The study population of this part is Wuhan community residents from the Wuhan-Zhuhai cohort,including baseline(year 2011),the first(year 2014)and the second(year 2017)follow-up.All visits were conducted during April to May to avoid seasonal differences.The estimated model established in part I was used to estimate the current-day concentration and 1?7 days moving averages of personal PM_2.5 exposure.Lung function was tested according to the guideline from American Thoracic Society.In baseline,lung function parameters included forced vital capacity(FVC),forced expiratory volume in 1 second(FEV₁),peak expiratory flow(PEF)and their percentages of the predicted values(FVC%pred,FEV₁%pred,PEF%pred).In the first and the second follow-ups,maximum mid-expiratory flow(MMF)and percentage of predicted MMF(MMF%pred)were added to evaluate small airway function.Mixed linear models were used to evaluate the association between short-term personal PM_2.5 exposure and lung function.In the evaluation of long-term exposure,the research subjects who participated in at least two lung function tests were selected.Participants were divided into high exposure group and low exposure group according to the median PM_2.5 level of all subjects in the same year,and then classified into persistently high exposure group(being in high exposure group in all periods),persistently low exposure group(being in low exposure group in all periods)or inconsistent exposure group.Mixed linear models and generalized linear models were used to evaluate the 3-year and 6-year lung function changes in persistently high exposure group compared with the persistently low exposure group,respectively.Morning urine was collected on the day of lung function tests.High performance liquid chromatography coupled with an electrochemical detector was used to determine urinary 8-OHd G concentrations.ELISA was used to determine the urinary 8-iso-PGF_2?concentrations.Mediation analyses were used to evaluate the roles of 8-OHd G and 8-iso-PGF_2?in the associations between PM_2.5 exposure and lung function parameters.All analyses were performed using SAS(Version 9.4)or R software(Version 3.6.1).Results:A total of 7685 observations were included to assess the association between short-term personal PM_2.5 exposure and lung function,with an average age of 55.69 years and 35.80%males.The average PM_2.5 concentration on the current-day of lung function test was 143.55?g/m³,with a range from 6.57 to 349.41?g/m³.Short-term personal PM_2.5exposure was associated with lung function decline,and the 7-day moving average had the strongest effect.After adjustment for age,gender,height,weight,smoking status,passive smoking,drinking status,education level,cooking at home and year,each 10?g/m³ increase in the current-day PM_2.5 exposure was associated with 2.36 m L,1.63 m L,14.53 m L/s and 0.20%declines in FVC,FEV₁,PEF and PEF%pred,respectively;each10?g/m³ increase in the 7-day moving average PM_2.5 exposure was associated with 9.29m L,0.28%,4.85 m L,30.78 m L/s and 0.43%declines in FVC,FVC%pred,FEV₁,PEF and PEF%pred,respectively.Stratified analyses showed the associations were stronger in females,participants with age more than 55 years,never smokers and never drinkers.Participants who participated in at least two lung function tests(N=4789)were selected to assess the association between long-term personal PM_2.5 exposure and lung function,with an average age of 56.08 years and 35.21%males.After adjustment for age,gender,height,weight,smoking status,passive smoking,drinking status,education level and cooking at home,compared with the persistently low exposure group,participants in the persistently high exposure group had increased 3-year declines in FEV₁/FVC(0.79%),FEV₁/FVC%pred(1.97%),PEF(128.60 m L/s)and PEF%pred(2.33%)with no significance,while significant increased 6-year declines in FEV₁/FVC(2.90%),FEV₁/FVC%pred(3.69%),PEF(370.86 m L/s)and PEF%pred(5.82%).Stratified analyses showed that compared with the persistently low exposure group,the lung function declines in participants in the persistently high exposure group were more obvious in females,participants with age no more than 55 years,never smokers and never drinkers.The role of oxidative stress in the association between personal PM_2.5 exposure and lung function was evaluated among 3895 participants who participated in baseline or the first follow-up completed the determinations of urinary 8-OHd G and 8-iso-PGF_2?.The median concentrations(P25,P75)of urinary 8-OHd G and 8-iso-PGF_2?were 58.31(20.05,134.09)?mol/mol Cr and 73.02(43.82,122.54)ng/mmol Cr,respectively.After adjusted for age,gender,height,weight,smoking status,passive smoking,drinking status,education level and year,there was neither a significant association between urinary 8-OHd G concentration and lung function,nor the mediated effect of 8-OHd G on the association between PM_2.5 exposure and lung function.A nonlinear association between PM_2.5 exposure and log-transformed 8-iso-PGF_2?concentration was observed.Among participants with high PM_2.5 exposure,personal PM_2.5 exposure concentration was linearly positively associated with 8-iso-PGF_2?concentration,with a 2.04%(95%CI:0.50,3.61)increase per 10?g/m³ 7-day moving average PM_2.5 exposure.Urinary 8-iso-PGF_2?partly mediated the association between personal PM_2.5 exposure and FEV₁(mediated proportion:15.70%)and FEV₁%pred(19.19%)declines.Conclusion:Short-term personal PM_2.5 exposure was associated with the declines in FVC,FVC%pred,FEV₁,PEF and PEF%pred in Wuhan community residents,and 7-day moving average had the strongest effect.Compared with the persistently low exposure group,persistently high exposure to PM_2.5 was associated with the increase of 6-year lung function declines.The associations between PM_2.5 exposure and lung function were more obvious among never smokers.Urinary 8-iso-PGF_2?partly mediated the association between personal PM_2.5 exposure and FEV₁ and FEV₁%pred declines.Part ? Association between personal PM_2.5 exposure and DNA methylation,andthe effect of DNA methylation on lung functionObjectives:To investigate differentially DNA methylated positions related to personal PM_2.5 exposure,and to explore the associations between the above sites and lung function.Methods:A total of 402 never smokers in the Wuhan community finished DNA methylation assay.The estimated method for personal PM_2.5 exposure levels on the day of lung function tests was same as part II.Lung function was tested according to the guideline from American Thoracic Society.Illumina Methylation EPIC Bead Chips were used to determine DNA methylation levels in whole blood,which can determine more than 850 thousand sites at the whole-genome level.R packages“minfi”and“wate Rmelon”were applied for quality control for samples and sites.After quality control for samples and sites,a total of 402 samples and 777153 sites were included in the epigenome-wide association study.Mixed linear models were applied to investigate differentially DNA methylated positions related to personal PM_2.5 exposure,and the associations between the above sites and lung function.All analyses were performed using R software(Version3.6.1).Results:The average age of participants in this part was 58.44 years,and the proportion of males was 12.19%.After adjustment for age,gender,BMI,passive smoking,drinking status,year,ratios of blood cells and batch effect,DNA methylation levels of 5 sites(cg10314909,cg08968107,cg18362281,cg24663971 and cg17834632)were negatively associated with personal PM_2.5 exposure,while the DNA methylation levels of cg24821877,cg24862131,cg23530876,cg11149743 and cg10781276 were positively associated with personal PM_2.5 exposure(FDR<0.05).And the top site was cg24663971(P=1.63(?)10^-9).Further analysis showed that DNA methylation levels of 3 sites,cg24663971(Intergenic),cg10314909(MMEL1)and cg11149743(HOXB7),were associated with lung function.The DNA methylation level of cg24663971 was positively associated with FVC%pred and FEV₁%pred;the DNA methylation level of cg10314909 was positively associated with FVC,FVC%pred and FEV₁%pred;the DNA methylation level of cg11149743 was positively associated with FVC%pred(all P<0.05).Stratified analyses suggested that the positive associations between DNA methylation level of cg24663971and FVC%pred and FEV₁%pred were more obvious among older participants.BMI had no effect modification on the associations between DNA methylation and lung function parameters.Conclusions:Ten differentially methylated positions were found to be associated with personal PM_2.5 exposure in the Wuhan community never smokers.And DNA methylation levels of 3 sites,cg24663971(Intergenic),cg10314909(MMEL1)and cg11149743(HOXB7),were associated with lung function.