Abstract
Several studies reported the roles of short-term exposure to fine particulate matter (PM2.5) on violent behaviors; however, existing findings had a limitation in assessing the population-representative association between violence and PM2.5 due to the limited data availability: most studies have been based on homicides in monitored urban areas. This study collected violence data from the National Hospital Discharge In-depth Injury Survey in South Korea (2015-2019), based on population-representative samples. To cover unmonitored areas, we used the daily modeled PM2.5, the predicted result driven by a machine-learning ensemble model covering all inland districts in South Korea (R2>0.94). We evaluated the national association between short-term exposure to PM2.5 and violence cases with a time-stratified case-crossover design. A total of 2,867 violence cases were included. We found an approximately linear association between short-term exposure to PM2.5 (lag 0-2 days) and an increased risk of violence, with an estimated odd ratio (OR) per 10 μg/m3 of PM2.5 of 1.07 with 95% CI: 1.02-1.12. This relationship was more prominent in males and individuals aged 64 years or less than in females and individuals aged 65 years or older for the most part. The estimated excess fraction of violence cases attributable to PM2.5 was 14.53% (95% CI: 4.54%-22.92%), and 6.42% (95% CI: 1.97%-10.26%) of the excess violence was attributable to non-compliance with the WHO guidelines (daily PM2.5 > 15 μg/m3). Our findings might be evidence of the need to establish elaborate action plans and stricter air quality guidelines to reduce the hazardous impacts of PM2.5 on violence in South Korea.