Pathways to depression: The impact of neighborhood violent crime on inner -city residents in Baltimore, Marylan

This dissertation examines the contextual impacts of disorder and crime on depressive symptoms in a population of current and former substance users living in an inner-city environment. Previous studies have not adequately explored mechanisms linking neighborhood conditions to mental heatlh outcomes. Hierarchical linear modeling, geographic clustering procedures, and path analysis are utilized to answer the following research questions: (1) When aggregating individual perceptions of neighborhood problems, what is the appropriate level of data aggregation? (2) Are higher levels of neighborhood problems, aggregated from individual perceptions to smaller and larger geographic areas, associated with higher levels of depressive symptoms? (3) To what extent do actual levels of violent crime correlate with perceptions of crime and neighborhood problems? (4) Are objective levels of neighborhood violent crime positively associated with individual experiences of violent crime? (5) Do perceptions of neighborhood problems mediate the relationship between objective levels of neighborhood violent crime and level of depressive symptoms, or is there a direct impact of objective levels of neighborhood violent crime on levels of depressive symptoms? Results indicate that aggregated neighborhood perceptions are modestly associated with individual level of depressive symptoms. Given the original sampling design, geographic clusters were found to be the most efficient means of aggregating neighborhood perceptions to form a contextual measure. Actual level of violent crime was found to be modestly correlated with both perceptions of neighborhood disorder and with individual experiences of violence. Finally, in path models, actual levels of violent crime were found to impact on level of depressive symptoms by modifying perceptions of the neighborhood, and by modifying actual exposure to violent events.