Can certain neighborhood features such as streets and natural surroundings influence the likelihood of homicide for teens? A study released online today from JAMA Pediatrics found street lighting, parks, public transportation, and maintained vacant lots were associated with lower odds of homicide among those from the ages of 13 to 20 in Philadelphia.

On the flip side, the chance of homicide were four times higher in locations with stop signs, nine times higher when security bars/gratings were present on houses, and three times higher when a house had private bushes or plantings compared to neighborhoods or homes that did not have the feature. (This doesn't mean these features cause homicide, but that it is more likely to occur when these conditions are there.)

The study included 143 homicide victims (average age 18 years) and 155 matched participants who were not victims of homicide from the city. The researchers then used pictures to create 360-degree panoramic images of the neighborhood from the street corner closest to each homicide and the location of matched participants at the time of the homicides for comparison. They looked for 60 environmental elements in each photo.

We asked one of the study's authors, Alison J. Culyba, MD, MPH, from The Children's Hospital of Philadelphia to tell us more about the findings.

What led you focus on neighborhood features for the study?

Homicide has been the second leading cause of death among all adolescents and the number one cause of death among African American adolescents for more than a decade. Youth violence is an incredibly complex public health problem. Research has taught us a lot about individual, family, and socioeconomic factors that are important in homicide. However, until recently, much less attention has focused on the role of the physical environment in shaping risk for severe violent injury and death.

As an adolescent medicine doctor, I work to help kids and families heal after violent injuries by connecting with teens and brainstorming strategies to keep them safe. However, I am also acutely aware of my limited ability to address larger contextual factors, such as a lack of safe spaces to play, from the context of individual doctor-patient relationships.  My frustration has fueled my commitment to adolescent research to understand risk-factors and postulate effective interventions so that we can safeguard youth.

How was using 360 panoramic images helpful?

Our research team stood on the street corner of each homicide and control location (where a homicide did not take place) and took a series of individual pictures. We then used specialized software to create a high-resolution 360 degree interactive image of each location. We were able to zoom in to see incredibly rich details of each location. For example, we could count the individual pieces of trash strewn along the entire sidewalk.

Did some aspects of the study results surprise you such as the odds of adolescent homicide were higher in places with stop signs and houses with security bars/gratings?

Overall, many of the study findings are consistent with prior work examining associations between green space, vacant and neglected buildings, and other less serious crimes. Some of these results were somewhat surprising, and it is hard to know exactly how to interpret them. However, in terms of the findings related to stop signs, we think this could be related to the fact that stop signs tend to be located on less busy streets in residential neighborhoods. Homicides may be more likely to occur in these locations because perpetrators may think there is a better opportunity to covertly commit crimes in less busy settings.

In regards to the security bars/gratings, I think it is quite possible that residents in less safe neighborhoods install them in response to existing high crime rates, rather than security bars/gratings causing crime. From our current study, we are unable to tell which came first. That's why it is so important that we do further research to test whether intervening on any of the features we identified may be able to protect kids from homicide.

What were some limitations of the study and how can they be addressed in future studies?

The main limitation of the current study is that we aren't able to understand whether the features we identified, such as parks, maintained vacant lots, street lighting, and public transportation actually protect kids from homicide. We can only tell that certain features of physical environments are related to homicide, not that they cause or reduce homicide. In order to see if we can actually protect kids, we need to do further studies where we design neighborhood interventions to improve some of the features we identified and then study how changing those features impacts on homicide rates.

For now, what can we take away from the study's findings?

Multiple modifiable features of streets, buildings, and natural surroundings were significantly related to adolescent homicide. This is exciting because it helps us identify what neighborhood features might be important in homicide risk. We should consider future studies of place-based interventions to improve street lighting, pedestrian infrastructure, public transportation, parks, and vacant lot greening to see if we can reduce youth violence by improving neighborhood context.

Have a question for the Healthy Kids panel? Ask it here. Read more from the Healthy Kids blog »