Examining associations between street vegetation and child firearm violence at microgeographic locations in 3 US cities: a case-control study

Abstract Background Firearm violence is a leading cause of death among U.S. children under age 18. Urban vegetation, a vital component of the living environment, has been found to influence firearm violence, yet the relationship between different types of street vegetation and firearm violence with children as the victim at the microgeographic scale remains unclear. Methods We analyzed 2316 child firearm assault incidents from Los Angeles, New York City, and Philadelphia from 2015 to 2021. Using a spatial case-control approach, we matched assault locations with control locations within 1-km. For each location, we applied deep learning algorithms to Google Street View images to classify street vegetation into tall (mainly tree) and short (mainly grass and shrubs). We employed conditional logistic regression models to examine the associations between visible amount of different street vegetation types and child firearm incidents, while controlling for sociodemographic and socioeconomic factors. Results We found that the associations between the amount of different street vegetation and firearm violence with children as the victim varied across cities. Short street vegetation was associated with reduced child firearm assault risk in both Los Angeles (OR = 0.62, 95% CI [0.59, 0.76]) and New York City (OR = 0.82, 95% CI [0.70, 0.94]). Tall street vegetation was associated with reduced risk only in Los Angeles (OR = 0.67, 95% CI [0.59, 0.77]). No associations were identified between any street vegetation and child firearm assaults in Philadelphia. Conclusions Our findings highlight the relationship between street vegetation and firearm violence with children at a microgeographic scale. The impact of street greenery varies by city and vegetation type. The findings of this study would serve as a scientific reference for designing interventions to reduce firearm assaults in urban areas.

Abstract licence: CC BY-NC-ND

40 more references, whose wording the paper’s licence keeps private.

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.