Abstract
BACKGROUND: This study evaluated trends and risk factors over time for self-reported gun carrying among freshman and sophomore public school students in Chicago, New York City and Los Angeles, chosen as high profile cities with different levels of firearm violence. METHODS: The study used four biennial waves (2007-2013) of the Youth Risk Behavior Survey (YRBS), an anonymous, voluntary survey of public high school students. Analyses were restricted to freshman and sophomores given significant high school dropout rates among older students. School population weighted results are presented based on the YRBS complex survey design, including comparisons of reported gun carrying across survey waves and cities. A violence index was created from eight survey items that capture students' perceived threat level. Chi square tests and multivariable Poisson regression analyses were used to test the significance of differences across cities and over time in the likelihood of gun carrying controlling for sociodemographic characteristics, mental health risk factors and behavioral risk factors. RESULTS: The study included a total weighted population estimate of 1,137,449 students across the three cities and four survey waves. Mean self-reported gun carrying across all survey waves was 8.89% in Chicago, 4.09% in New York City, and 6.03% in Los Angeles (p < 0.001). There were no significant changes in gun carrying prevalence within each individual city over the survey waves. Multivariable Poisson regression estimates showed increased likelihood for gun carrying among males (IRR 1.41, CI 1.27-1.58), among non-Hispanic Blacks (IRR 1.26, CI 1.07-1.48), and among those who reported a higher violence index. Each additional violence index count increase was associated with a 1.74 times (CI 1.70-1.78) increased likelihood for gun carrying. CONCLUSIONS: There was a much higher self-reported rate of gun carrying and a higher burden of violence exposure in Chicago as compared to New York City and Los Angeles. Students' exposure to violence extended to other stressors illuminated by the YRBS including fighting, perceptions of safety, and other high-risk behaviors. Through the violence index we created, we are better able to categorize the most high-risk individuals and describe the magnitude of their increased likelihood to carry a gun.