Adult Support and Substance Use Among Homeless Youths Who Attend High School
Background and Purpose: Despite high rates of substance use among homeless youths, little is known about the interaction of substance-use risk and protective factors. Further, limited research exists on substance use by school-attending homeless youths, as extant studies have largely relied on street- and shelter-based samples. The purpose of this study thus was to examine how risk and protective factors influence school-attending homeless youths’ substance use as well as how protective influences can mediate and moderate the impact of risk factors on substance use. A theoretical model was constructed using empirical precedents on adolescent substance use, the risk and resilience framework, and social capital theory. The study hypotheses were that greater substance use would be reported by school-attending homeless youths who: 1) reported gang involvement; 2) experienced partner abuse; 3) skipped school more frequently; and 4) had lower levels of adult support. It was further hypothesized that adult support would both mediate and moderate the relationship between the aforementioned risk factors and substance use.
Methods: Using the Mplus software, authors analyzed data from the California Healthy Kids Survey (CHKS) dataset, which includes a representative, statewide sample of 729,626 high-school-attending youths. A sub-sample of 2,146 homeless youths was drawn from the full sample. Measures for the independent variables (gang membership, intimate partner abuse, truancy, and adult support) and dependent variable (substance use) were derived from the survey. The theoretical model was tested via structural equation modeling (SEM) using maximum likelihood parameter estimation. To test the mediation effects of adult support, SEM and the four standard criteria of MacKinnon, Krull, and Lockwood (2000) were used. To test the moderation effects of adult support, SEM was utilized to create three moderation models and to examine the direct effects of the independent variable, moderator, and interaction term on the dependent variable.
Results: The full model represented an acceptable fit to the data (X2=987.42, df=83, p<0.001, CFI=0.968, RMSEA=0.071, PCFI=0.669). The magnitude of standardized path coefficients from the three predictors and the mediating variable to substance use ranged from 0.139 to 0.311 (p<0.001). Collectively, gang membership, partner abuse, truancy and adult support explained 36% of the variance in homeless youths’ substance use. Greater substance use was associated with gang membership, partner abuse and truancy. Lower levels of substance use were associated with higher levels of adult support. Additionally, adult support acted as both a mediator and moderator between the hypothesized risk factors and substance use.
Conclusions and Implications: Findings highlight the mediating and moderating effects of adult support on substance-use risk factors. This study establishes baseline levels of substance-use prevalence for school-attending homeless youths from which future longitudinal research can build to illuminate the causal pathways among substance-use risk factors, adult support, and actual use. Understanding risk and protective factors will help school personnel assess for substance-use risk as well as limited adult supports in determining which particular students are most in need of intervention. Further, findings underscore the importance of incorporating adult community supports into existing violence and substance-use prevention efforts for school-attending homeless youths.