Moderators of Trauma Exposure and Mental Health Diagnoses among Homeless Urban Youth

Purpose. Homeless youth have higher rates of victimization and mental health / substance abuse diagnoses than do other youth. These patterns vary by gender and by sexual orientation, and take place in the context of complex service histories (Gwadz, Nish, Leonard, & Streauss, 2007; Thompson, McManus, & Voss, 2006). This presentation examines gender, sexual orientation and service moderators of trauma exposure and mental health diagnosis among homeless youth. Methods. 178 urban homeless youth were interviewed as part of a university-agency partnership. Variables collected included socio-demographics, mental health diagnoses (computer-based, E-Mini measure), trauma exposure (CREV exposure scale and self report of specific experiences) and prior service contacts and trajectories with formal and informal, community services. Trained MSW and PhD students collected data with youth in three settings (shelter, drop-in center and transitional living) using a combination of computer based and paper reporting. Interviewing continued until saturation at each site. Given the nature of the population, participation rates are difficult to assess, however incentives were used, and interest among youth in each site was high. Results. Youth ranged in age from 15 to 25. As expected, homeless youth report significant trauma histories with regard to familial abuse and other types of violence exposure. More than 25% report child neglect, physical abuse and sexual abuse, with more than half reporting at least one type of maltreatment. More than 50% also report having been threatened or physically hurt by others (not family), and 20% report non-familial sexual assault. These patterns vary by gender and sexual orientation. Respondents also report numerous service experiences over much of their lifetime, including high rates of school, child welfare and juvenile justice service. Homeless youth have multiple mental health diagnoses, with more than a third reporting current major depression, alcohol and marijuana dependence/abuse, and more than 60% meeting criteria for at least one current diagnosis. In logistic regression models, trauma exposure is the strongest predictor of several mental diagnoses, including PTSD, anti-social personality disorder, and most alcohol and drug dependencies. Depression and dysthymia present an exception to this pattern. Some demographics and service histories moderate the relationship between trauma exposure and current diagnoses, in both positive and negative directions, including differing patterns by gender and sexual orientation in combination with some community interventions. For example, for sexual; minority youth, prior service patterns in religious settings appear to increase the strength of relationship between trauma and diagnosis of anti-social personality disorder, while child welfare intervention has a buffering effect. We will present logistic models detailing interactions that describe the moderating effects between trauma histories and current diagnoses. Implications for Practice or Policy. Our findings indicate that community service histories are associated with current mental health functioning among certain groups of homeless youth. Understanding how systems such as schools, churches and child welfare agencies may be associated with differing trajectories for certain groups of youth is important for strengthening homelessness prevention efforts in these settings.