Date of Award


Degree Type


Degree Name

Doctor of Philosophy (PhD)

Graduate Group

Epidemiology & Biostatistics

First Advisor

Douglas J. Wiebe


Annually, over 600,000 youth are treated in Emergency Departments for assault-related injuries; African American male youth are disproportionately affected. Research suggests that adolescent-adult connections may protect youth from violence in the general U.S. population. However, research among male youth in under-resourced urban neighborhoods is limited, based on self-report, and has not accounted for the complex interplay between individual, family, and neighborhood contextual factors. This dissertation leverages data from a recent population-based case control study that enrolled adolescent male gunshot (n=135) and non-gunshot (n=194) assault cases and community-based controls (n=274) to measure the nature of adult connection, assault injury, and neighborhood exposures. First, we examined associations between supportive adult connections (defined by brief survey questions and detailed family genograms) and objective measures of assault injury using conditional logistic regression, stratified by prior violence involvement and adjusted for individual and contextual confounders. Among youth with high levels of prior violence involvement, reporting at least one supportive adult family member was associated with higher odds of gunshot assault injury (OR=4.01,p=0.01) and non-gun assault injury (OR=4.22,p=0.01). No significant associations emerged among youth with low prior violence involvement. Second, we compared conventional versus novel methods of measuring environmental exposures among youth during daily activities. We found that defining environmental exposures based on participant home address resulted in significant misclassification compared to gold standard measures of detailed participant activity path data. Using a novel method that divided participant activity paths into origin-destination segments, we demonstrated that calculating environmental exposures based on shortest possible travel routes compared to actual travel paths may result in exposure misclassification for point-level environmental data. Third, we applied these novel spatial methods, and found no significant evidence that adult connection was protective against exposure to neighborhood risk factors during daily activities. This work demonstrates that despite high levels of connection, families struggle to protect male youth in low resource neighborhoods from violent injury. Interventions to improve neighborhood contexts may play a significant role in violence prevention efforts. Novel spatial methods explored herein can be utilized in future research to more accurately quantify environmental exposures and associations with injury outcomes.

Included in

Epidemiology Commons