Air pollution and increased number of psychiatric emergency room visits : A case-crossover study for identifying susceptible groups

Background: Ambient particulate matter is a leading risk factor for disease globally. Particulate matter 10 (PM10) and particulate matter 2.5 (PM2.5) are derived from different sources, including operating motor vehicles as well as from industrial activities. In this study we investigate the association between increased concentrations of PM and total daily visits to the psychiatric emergency unit (PEV). Further, the aim is to identify specific risk groups who are more susceptible to the effects of air pollution exposure by studying sex, age, ongoing psychiatric follow-up and diagnoses of depression/anxiety or substance use. Material and methods: The sample was comprised of data from 2740 days to 81 548 PEVs at Sahlgrenska University Hospital in Gothenburg and daily mean concentrations of PM10 and PM2.5. A time-stratified case-crossover design was used to analyse associations between air pollution and PEVs. Results: Mean number of daily PEVs were 35 and sex distribution was even. PM exposure was associated with total PEV at lag 0 (the same day), by RR 1.016 (95% confidence interval [CI] 1.004–1.028) and RR 1.020 (95%CI 1.003–1.038) per 10 μg/m3 increase in PM10 and PM2.5, respectively. In females, PEV were increased at lag 0 and lag 1, and in males at lag 1 and lag 2. In the age-stratified analysis, PEVs significantly increased following PM exposure amongst individuals aged 35–65 years by lag 0–2 and in individuals who had contact with outpatient care at lag 0 to lag 1. There were no associations between air pollution and PEVs for any specific diagnostic group evaluated (amongst depression, anxiety and substance use disorder). Conclusions: The results indicate that acute exposure to PM10 and PM2.5 may trigger acute worsening in mental health in both males and females, especially among 35–65 year old individuals. However, in subgroups of the most common psychiatric diagnoses, we did not observe statistically significant associations with PM exposure