Air Pollution Is Associated With Increased Risk Of Outpatient Visits For Psychiatric Disorders

Background:Ambient air pollution is exponentially increasing in modern society due to global heavy industrialization and technological advancements with only 1%of the worldwide population residing in locations that meet the air quality criteria levels from the World Health Organization(WHO).Industrial power plants,burning of fossil fuels,biomass combustion,agriculture,and transportation are the primary producers of air pollutants such as particulate matter(PM),ozone(O3),nitrogen dioxide(NO2),sulfur dioxide(SO2),and carbon monoxide(CO).In addition to causing 4.9 million premature deaths in 2017,emerging evidence has revealed exposure to air pollutants imposes several health risks including cardiovascular diseases and respiratory disorders.Furthermore,various experiments involving animal models discovered abnormal structural,morphological,and functional alterations of the central nervous system(CNS)induced by neuroinflammation,oxidative stress,and other neurotoxic effects from exposure to air pollution.As multiple studies consistently reported the detrimental neurological effect of ambient air pollutants,a relatively new field of research exploring the relationship between air pollution and mental health disorders has developed.This new line of study is crucial as the prevalence of psychiatric disorders is increasing annually worldwide.The two most recent Global Disease Burden(GBD)studies conducted in 2017 and 2019 revealed an increase of 178.1 million cases for psychiatric disorders over the three years.As of 2019,mental health disorders were the 7th leading cause of Disability-Adjusted Life Years(DALY),implicating the heavy global medical burden of these conditions.Treatment and productivity loss from these disorders evoke a heavy toll on the international economy as $2.5 trillions of USD are spent and lost each year.Instead of decreasing,this figure is expected to significantly increase to $6 trillion USD by the year 2030.However,due to the complex heterogeneity of psychiatric disorders,the underlying biomolecular pathogenesis of these conditions remains elusive.It is therefore critical to explore any potential risk factors that may be associated with these disorders,including the relationship between ambient air pollution and mental health.By identifying air pollutants that may have detrimental effects on mental health and implementing air quality control policy,the social and economic burdens caused by psychiatric disorders can be mitigated.The majority of results from epidemiological studies however are either non-conclusive,inconsistent,or irreproducible.Objectives:This retrospective case-crossover epidemiological study aims to investigate the association between ambient air pollutants and outpatient department visits for five major psychiatric disorders(anxiety disorder,depression,bipolar disorder,schizophrenia,and sleep disorders)across three Chinese cities.Methods:Daily outpatient data from the psychiatric department were collected from three hospitals in the three cities of Changsha,Chenzhou,and Shaoyang with a total duration from January 1,2013 to December 31,2020.However due to different information retrieval systems,psychiatric outpatient admissions were retrieved throughout three study periods:January 1,2013 to December 31,2020 for the Second Xiangya Hospital of Central South University;June 1,2016 to December 31,2020 for the Shaoyang Brain Hospital;and October 21,2013 to September 1,2018 for the Chenzhou Psychiatric Hospital.The Second Xiangya Hospital of Central South University in Changsha,the Shaoyang Brain Hospital,and the Chenzhou Psychiatric Hospital are all tertiary public hospitals that specialize in the treatment and prevention of psychiatric disorders and account for the majority of outpatient visits for mental disorders in each of their respective cities.Outpatient data from these hospitals consist of age,sex,first six digits of the Chinese identification(ID)number,date of outpatient visit,and clinical diagnoses.Based on the first six digits of their Chinese ID number,we can determine the general geographic location of these patients and locate the districts of the cities they reside in.Diagnoses of psychiatric disorders are all confirmed by senior psychiatrists at the three hospitals with the utilization of the Diagnostic and Statistical Manual,4th Edition(DSM-IV),DSM-5,and the International Classification of Diseases 1 10th Revision(ICD-10).Only patients with anxiety disorder,depressive disorder,bipolar disorder,schizophrenia,and sleep disorder were included in the study while patients with comorbid psychiatric disorders,substance misuse and a history of head trauma or neurological illnesses were excluded.Additionally,according to the first six digits of their Chinese ID number,only patients who reside in the three cities of Changsha,Chenzhou and Shaoyang were included.Air pollutant data are extracted from the Chinese Air Quality Reanalysis(CAQRA)and the Tracking Air Pollution in China(TAP)databases and consist of daily average concentrations of PM2.5,PM10,SO2,O3,NO2,and CO.The CAQRA database is a collaborative project involving several research institutions such as the Institute of Atmospheric Physics,Chinese Academy of Sciences(IAP/CAS),and the China National Environmental Monitoring Centre(CNEMC).The TAP database,derived from machine learning approaches,is a near real-time air pollutant concentration database collaboratively developed by several universities such as Tsinghua University and Peking University.The CAQRA database includes data throughout the entire country of China and has been verified through both cross-validation and independent data validation while the TAP database is constantly maintained and updated.As time-stratified case-crossover design has been utilized to assess the effects of air pollution on an assortment of health outcomes,it has been adopted for this current study.A time-series model is necessary since psychiatric outpatient visits,air pollution concentrations,and meteorological factors are all associated with dates.Unlike the majority of other studies where a healthy control group is required,each outpatient case in this present study is its own control on a set of predefined control days relative to the time of admission to the hospital.For each outpatient case,the day of the outpatient admission is defined as the case day.The control periods were selected from the same day of the week as the case’s hospital visit,in the same month and year.For each case,three or four controls were selected depending on the month.A Spearman correlation analysis was first performed to analyze the correlation between different environmental exposure factors.Conditional logistic regressions were then applied to analyze the odds ratio of the effects of air pollutants on the five major psychiatric disorders.Analysis was additionally carried out for both sex and age groups(Age groups:patients between 1 to 20 years old,patients between 21 to 40 years old,and patients over the age of 40).Different lag models for air pollutants were constructed from single day lags(lag 0 to lag 7)to moving average exposure lags(lag 0-1 to lag 0-7).After controlling for the meteorological factors of temperature and humidity,results are reported as odds ratios(ORs)and 95%confidence intervals(CIs)associated with an increase in interquartile range(IQR)of the air pollutants.A meta-analysis was finally conducted in order to achieve a pooled effect estimate.An association was considered statistically significant when p<0.05 in all presented statistical results.All analyses were performed using the R programming language.Results:A total of 613,010 cases of anxiety disorder,depression,bipolar disorder,schizophrenia,and sleep disorders were recorded in the three cities.More females(51.51%)visited the psychiatric outpatient departments than males(48.49%)while the majority of the patients visiting the psychiatric outpatient departments were between the ages of 21 and 40(86.92%).The lowest frequency of outpatient admissions occurred in the group of patients under the age of 21(6.06%).Most of the patients seeking medical attention in the outpatient departments are clinically diagnosed with schizophrenia(67.69%),followed by depression(11.88%)and sleep disorders(8.14%).Of the 3 cities,Changsha had the highest average air pollutant concentrations for PM2.5,PM10,NO2,and O3(54.19 μg/m3 for PM2.5,74.65 μg/m3 for PM10,36.37 μg/m3 for NO2,and 90.76 μg/m3 for O3).Shaoyang had the highest mean concentration for SO2(19.789μg/m3)while Chenzhou had the highest average concentration for CO(1.24 μg/m3).Spearman’s correlation test revealed that among all the air pollutants,the strongest positive correlation is established between concentrations of PM2.5 and that of PM10(correlation coefficient,r=0.95;p<0.001).Both concentration levels of PM2.5 and PM10 were moderately and positively correlated with concentrations of NO2 and SO2(r=0.670.78;p<0.001).All air pollutant concentrations have a negative correlation with temperature except for O3 with a moderate positive correlation(r=0.61;p<0.001).Among all the air pollutants,only CO concentrations have a positive correlation with relative humidity although it is very weak(r=0.14;p<0.001).Anxiety disorder:Statistically significant associations were present between anxiety outpatient admissions and all ambient air pollutants with the exception of SO2.Exposure to CO,NO2,PM10,and PM2.5 was associated with increased risk for anxiety outpatient admissions.The strongest effect estimates of CO occurred at lag day 0-3 and 0-4 days([1.079;95%CI:1.041-1.118]and[1.079;95%CI:1.04-1.12]respectively).NO2 have the largest effect estimate at lag day 0-5 days(1.167;95%CI:1.116-1.22)while the strongest effect estimates for both PM10 and PM2.5 occurred on lag day 0-1 days([1.036;95%CI:1.001-1.073]and[1.038;95%CI:1.008-1.07]respectively).Increased risk of anxiety outpatient visitations is significantly associated with an increase of concentrations in NO2 and CO for all lag days except for lag day 5(CO)and lag day 6 and 7(NO2).Exposure to PM10 and PM2.5 are statistically associated with increased risk for anxiety admissions on lag day 0-1 days;PM2.5 is additionally significantly associated with increased risk on day of exposure(lag day 0)and lag day 7.Exposure to O3 does not seem to impose a greater risk in anxiety outpatient visits as its strongest estimated effect on lag day 4 is only 0.956(95%CI:0.92-0.994).Depression:Increased concentrations of all ambient air pollutants are statistically significantly associated with outpatient admissions for depression with CO,NO2,and SO2 causing an increased risk.NO2 and SO2 have the strongest effect estimate on lag day 0-5 days(1.12;95%CI:1.088-1.152)and lag day 0-3 days(1.037;95%CI:1.003-1.071)respectively.Increased risk of depression outpatient visits is significantly associated with an increase of concentrations in NO2 for all lag days except for lag day 6 and 7 while SO2 is additionally significantly associated with increased risk on lag day 0 and lag day 0-3 days.Effects of CO exposure demonstrate inconsistent results that are statistically significant as some lag days show OR less than one(lag day 5,6,and 7)while other lag days display an increased risk(lag day 0,0-1,and 0-2 days).The largest and smallest significant effect estimate for CO occurs on lag day 0-1(1.03;95%CI:1.006-1.055)and lag day 7(0.957;95%CI:0.937-0.976)respectively.Exposure to O3,PM10,and PM2.5 do not cause increased risk for outpatient admissions for depression as their strongest effect estimates are 0.977(95%CI:0.956-0.999)at lag day 7,0.98(95%CI:0.962-0.999)at lag day 6,and 0.974(95%CI:0.9570.991)at lag day 7 respectively.Bipolar Disorder:Statistically significant associations were present between outpatient visitations for bipolar disorder and all ambient air pollutants with CO,NO2,and SO2 causing an increased risk.NO2 and SO2 demonstrate OR greater than 1 for all lag days except for lag days 6 and 7(NO2)and lag days 2,3,4,5,6,and 7(SO2).Increased risk for bipolar disorder outpatient visits is significantly associated with an increase of concentrations in CO at lag day 0 and lag day 0-1 days with the strongest effect estimates on day of exposure(1.041;95%CI:1.011-1.071).NO2 and SO2 have the largest effect estimate on lag day 0-4 days(1.197;95%CI:1.15-1.246)and lag day 0-5 days respectively.Exposure to O3,PM10,and PM2.5 do not cause increased risk for outpatient admissions for bipolar disorder as their strongest effect estimates are 0.943(95%CI:0.907-0.981)at lag day 7,0.951(95%CI:0.922-0.981)at lag day 7,and 0.974(95%CI:0.949-0.999)at lag day 5 respectively.Sleep Disorder:With the exception of PM2.5,all ambient air pollutants are statistically associated with outpatient admissions for sleep disorders.Among the air pollutants,only an increase in concentration of NO2 is associated with an increased risk for sleep disorder outpatient admissions.NO2 have the strongest estimated effect at lag day 0-5 days with every incremental 10 μg/m3 increase corresponding to an increase in risk for sleep disorder outpatient visits(1.149;95%Cl:1.099-1.202).There is significant association between increased risk for sleep disorder and increased NO2 exposure for all lag days except for lag day 6 and 7.Few statistically significant associations are detected between exposure to air pollutants consisting of CO,ozone,PM10,and SO2 and outpatient frequencies for sleep disorders.These ambient air pollutants do not impose an increased risk as their strongest effect estimates are 0.968(95%CI:0.938-0.998)at lag day 5 for CO,0.964(95%CI:0.93-0.998)at lag day 7 for O3,and 0.959(95%CI:0.927-0.992)at lag day 7 for SO2.PM10 have the strongest estimated effect for both lag day 0 and lag day 0-1 days([0.962;95%CI:0.928-0.997]and[0.962;95%CI:0.926-0.998]respectively).Age:In patients under 21 years old,there is statistically significant association between exposure to NO2 and increased risk for psychiatric outpatient admissions in all lag days except for lag day 5 and lag day 7.NO2 have the strongest estimated effect at lag day 0-2 days with every incremental 10 μg/m3 increase corresponding to an increase in risk for psychiatric outpatient cases(1.123;95%CI:1.091-1.155).Increased CO and SO2 exposure were also positively associated with increased risks for the same age group with the strongest effect at lag day 3(1.033;95%CI:1.01-1.056)and lag day 6(1.032;95%CI:1.006-1.059)respectively.In young adults(21 to 40 years old),statistically significant association is detected between all air pollutants and outpatient frequencies except for PM2.5.NO2,SO2,and CO exposure increase the risk for increased psychiatric outpatient admissions whereas O3 did not impose a greater risk in both cases and controls.Increased risk of psychiatric outpatient cases is significantly associated with an increase of concentrations in NO2 at all lag days except for lag day 7 with the strongest estimated effect at lag day 0-5 days(1.144;95%CI:1.126-1.162).Increased exposure to CO is associated with increased risk of psychiatric outpatient visitations at lag day 0,1,2,3,0-1,0-2,0-3,0-4,0-5,and 0-6 days while the statistically significant association between increased exposure to SO2 and increased risk of outpatient cases are present in all of the lag days except for lag day 5,6,and 7.CO and SO2 have the strongest estimated effect at lag day 0(1.033;95%CI:1.023-1.043)and lag day 0-4 days(1.048;95%CI:1.03-1.067)respectively.Exposure to both PM10 and O3 does not seem to impose a greater risk in psychiatric outpatient admissions as their corresponding strongest estimated effect on lag day 1 and lag day 5 indicates an OR less than 1([0.976;95%CI:0.962-0.99]and[0.986;95%CI:0.975-0.996]respectively).In individuals older than 40 years old,a statistically significant association is detected between all air pollutant types and risk of psychiatric outpatient admissions.Aside from ozone,increase concentrations of all air pollutants are associated with an increased risk for outpatient frequencies.Statistically significant association is present between increased risk of outpatient visitations and increased exposure to CO(at all lag day except for lag day 2,4,5,6,and 7 days),NO2(at all lag day except for lag day 5,6,and 7 days),SO2(at all lag day except for lag day 0,2,3,4,5,6,and 7 days),PM2.5(at lag day 2,0-3,and 0-4 days),and PM10(at lag day 1).The strongest effect estimates for CO,SO2,PM2.5,and PM10 are present on lag day 0-3 days(1.075;95%CI:1.032-1.119),lag day 0-5 days(1.069;95%CI:1.006-1.137),lag day 0-4 days(1.039;95%CI:1.001-1.079),and lag day 1(1.037;95%CI:1.003-1.073)respectively.NO2 have the strongest estimated effect for both lag day 0-2 days(1.206;95%CI:1.158-1.256)and 0-4 days(1.206;95%CI:1.152-1.263).Exposure to ozone does not seem to impose an increased risk for psychiatric outpatient cases as its strongest effect estimate is an OR of 0.944(95%CI:0.903-0.987)on lag day 3.Sex:With the exception of PM10,statistically significant association is established between air pollutant concentrations and risks of psychiatric outpatient visitations in males.Increased CO,NO2,and SO2 concentration levels is associated with an increased risk for hospital outpatient admissions at all lag days except for lag day 4,5,and 7 days(CO),lag day 7(NO2),and lag day 4,5,and 6(SO2).The strongest effect estimates for CO is on lag day 0-3 days(1.035;95%CI:1.019-1.051)while for NO2 it is on lag day 0-6 days(1.128;95%CI:1.104-1.153).SO2have the strongest effect estimates for both lag day 0-3 days(1.043;95%CI:1.021-1.066)and 0-7 days(1.043;95%CI:1.014-1.073).Ozone does not seem to confer any risk for increased outpatient visitations as its strongest effect estimates is an OR of 0.974 for both lag day(95%CI:0.956-0.992)and lag day 5(95%CI:0.957-0.992).PM2.5 additionally does not impose any increased risk as its only statistically significant association with psychiatric outpatient case frequencies demonstrates an OR of 0.986(95%CI:0.9750.997)on lag day 5.Statistically significant associations have been detected between all air pollutants and increased risk for outpatient admissions for the female sample size.Increased concentrations of CO,NO2,and SO2 show a significant OR greater than 1 for all lag days except for lag day 4,5,6,7,0-6,and 0-7 days(CO),lag day 6 and 7(NO2),and lag day 5,6,and 7 days(SO2).The strongest effect estimates are present in lag day 0 for CO(1.031;95%CI:1.018-1.043),lag day 0-5 days for NO2(1.156;95%CI:1.1351.178),and lag day 0-4 days for SO2(1.045;95%CI:1.023-1.067).Increased exposure to O3 and PM10 does not seem to increase risk for psychiatric outpatient visitations because their strongest effect estimates display an OR of 0.977(95%CI:0.96-0.994)at lag day 7 and an OR of 0.984(95%CI:0.972-0.997)at lag day 6 respectively.Similarly,to the male sample size,increased concentrations of PM2.5 displayed only one statistically significant OR of 0.987(95%CI:0.977-0.998)at lag day 5.Conclusion:This hospital-based study discovered acute exposure to ambient air pollutants,especially NO2,is statistically significantly associated with psychiatric outpatient admissions for five different major types of disorders across three cities in the Hunan Province of China.Exposure to NO2 and other air pollutants may therefore induce or exacerbate mental health disorders.Interventions that assist in improving air pollution may thus help prevent incidences and mitigate the severity of psychiatric disorders.