LAUSR

Background As ecological factors are getting attention as important determinants of suicide, it is important to identify the unit at which the largest variation exists for more tailed strategy to prevent suicide. We examined the relative importance of two administrative levels for geographic variation in the suicide rate between 2014–2016 in Seoul, the capital city of Korea. Methods Two-level linear regression with Dongs (level 1) nested within Gus (level 2) was performed based on suicide death data aggregated at the Dong-level. We performed pooled analyses and then year-stratified analyses. Dong-level socioeconomic status and environmental characteristics were included as control variables. Results The overall age- and sex- standardized suicide rate across all Dongs decreased over time from 24.9 deaths per 100,000 in 2014 to 23.7 deaths in 2016. When Dong and Gu units were simultaneously considered in a multilevel analysis, most of the variation in suicide rate was attributed to within-Gu, between-Dong differences with a contribution of Gu-level being small and decreasing over time in year (Variance partitioning coefficient of Gu = 5.3% in 2014, <0.1% in 2015 and 2016). The number of divorce cases per 100,000 explained a large fraction of variation in suicide rate at the Dong-level. Conclusions Findings from this study suggest that ecological micro-area unit is more important in reducing the geographic variation in the suicide rate. More diverse ecological-level data needs to be collected for targeted area-based suicide prevention policies in Korea.