SOCIAL NETWORKS, DRUG USE, AND DRUG ABUSE HELP-SEEKING: A TEST OF THE NETWORK EPISODE MODEL AMONG AFRICAN AMERICAN WOMEN

Untreated substance use disorders are a major public health concern that has costly consequences at both the societal and individual level. Identifying the characteristics and resources of those who seek help for substance abuse problems in order to inform more effective intervention and treatment techniques is therefore an important research objective. Using the Network Episode Model (NEM) as a theoretical framework, this dissertation examines both substance abuse help-seeking (i.e. inpatient/outpatient treatment and 12-Step meeting attendance) and patterns of drug use over time among low-income African American women, with a special focus on the role of the social network system in shaping these outcomes. Drawing on social network theory, critical race theory, and health service utilization research, this test of the Network Episode Model addresses the relative absence of work examining the connections between network characteristics and help-seeking in multiply marginalized groups. The core relationships proposed by the NEM are systematically tested using longitudinal data gathered for the Black Women in the Study of Epidemics Project (N=643). Findings of multilevel models indicate strong support for the Network Episode Model. Specifically, measures of social influence, social control, and social integration significantly predict both patterns of drug use and help-seeking. Importantly, having contact with and receiving health advice from a physician emerged as a significant predictor of a number of positive outcomes, including quitting or abstaining from illicit drug use during the study and attending 12-Step meetings. Results also reveal that experiences specifically related to low-income African American women’s multiply marginalized status – such as experiencing gendered racism – significantly predict patterns of drug use over the study timeframe and may be an important risk factor for substance abuse. In all, this research reveals the important contributions of both traditional predictors and social network predictors on substance abuse help-seeking and patterns of drug use over time. Conclusions suggest that given the limited financial and material resources of multiply marginalized groups, learning how to mobilize or effectively build upon available social network resources to encourage substance abuse treatment may be a particularly fruitful strategy to explore