A spatial analysis of the moderating effects of land use on the association between alcohol outlet density and violence in urban areas.

AIMS While there is substantial evidence of an association between alcohol outlet density and assault, it is unlikely this association is constant across the urban environment. This study tested the moderating influence of land use on the outlet-violence association. DESIGN Cross-sectional ecological study that controlled for spatial autocorrelation. SETTING, PARTICIPANTS AND MEASUREMENTS: Police-recorded data on simple and aggravated assaults were obtained for all 302 block groups (mean population = 1038) in Cincinnati, Ohio, USA. Addresses of alcohol outlets for Cincinnati were obtained from the Ohio Division of Liquor Control, geocoded to the street level, and aggregated to census block groups. Data on eight categories of land use were obtained from the Cincinnati Area Geographic Information System, with location quotients computed for each block group. FINDINGS We found substantial evidence that the impact of total alcohol outlet density, bar density and carryout density on assault density was moderated by land use. CONCLUSIONS By taking into account local characteristics, policy-makers can make more informed decisions when regulating the placement and density of alcohol licenses in urban areas. Similarly, more systematic knowledge of how the association between alcohol outlet density and assault varies across the urban landscape should reduce harm and promote responsible retailing. Nevertheless, ours is one of the first studies to address the moderating effect of land use and we encourage further research to test the stability and generalisability of our results.