Association of Particulate Matter Exposure and Depression: A Systematic Review and M eta-Analysis

摘要

Background Impact of particulate matter (PM) air pollution exposure and mental health conditions is generating lot of curiosity among public health practitioners and policy makers. We aimed to generate evidence on daily PM2.5 and PM10 exposure and depression. Methods A systematic search was conducted for published studies in English till December 2017. Independent two electronic searches were carried out in Medline, Web of Science and Cochrane Library using key words "Airborne Particulate Matter" OR "Air Pollutants" OR "Air Quality", "Outdoor Air Pollution" "PM2.5" "PM10"" Depression" "Depressive symptoms" "Mood disorder." We used Medical Subject Headings (MeSH) terms: "Particulate matter" and "Depression" in Medline and Cochrane Search. We have included articles reporting adjusted relative risk (RR) or odds ratio (OR) PM exposure and depression. References of the selected articles were also traced. Meta-analysis was performed to calculate pooled estimate using fixed effect model. Quality of studies was assessed using Newcastle-Ottawa Scale for observational studies. Protocol of the systematic review was registered in PROSPERO. Results We have got 938 records from all searched databases. Of which, 14 records were included for full text review after screening abstract and removing duplicates. Finally, eight articles were included for meta-analysis. Most of studies were longitudinal studies and spread out geographically across the regions. Daily PM2.5 exposure was found to be associated with depression (RR=1.08, 95 CI: 1.02, 1.14) with acceptable heterogeneity (12: 56%). Daily PM10 exposure was not found to be associated with depression (RR = 1.0$195% CI: 0.97,1.04). Age, sex, household income, education, co-morbidity and road traffic noise were adjusted as confounders. Conclusion We have found modest strength of association for daily PM2.5 exposure and depression. Unknown confounders might have affected the pooled estimate from longitudinal studies in our analysis.