Methods: Data and Samples: We conducted a secondary data analysis using the Chafee Independent Living Evaluation Project. Our sample consisted of 429 foster youth (175 males, 254 females) residing in a large metropolitan area in California. Measurement: At two assessment points (ages 17 and 18), participants reported their past-year delinquent behavior (e.g., destroying property, stealing, fighting), as well as past-year substance use (e.g. alcohol, marijuana, illicit drugs). Both variables were dichotomized as either “present” or “absent” for use in subsequent analyses. Data Analysis:We examined the relative prediction of each risk behavior (delinquency, substance use) to the other over time by conducting separate binary logistic regression analyses. The first model examined substance use at age 17 as the primary predictor of delinquency at age 18, and the second model examined delinquency at age 17 as the primary predictor of substance use at age 18. We utilized key demographic variables (e.g., race/ethnicity, placement instability, child maltreatment) as controls in both models.
Results:Nearly half of youths (n=193, 46.1%) engaged in past-year delinquent behavior at age 17, most commonly, avoiding paying for purchases (22.7%) and acting loud/unruly in a public place (17.8%). Approximately half of youth (n=211, 49.8%) reported past-year substance use at age 17. The most frequently used substances were alcohol (41.0%), marijuana (26.7%) and prescription drugs used for non-medical purposes (8.3%). At age 18, youths reported similar likelihoods and patterns for past-year delinquency and substance use as reported at baseline. In multivariate analyses, the strongest predictors of both substance use and delinquency at age 18 were engagement in these risk behaviors at age 17. However, substance use at age 17 was not a significant predictor of future delinquency once baseline delinquency was controlled for in Model 1. In contrast, delinquency at age 17 remained a significant predictor of future substance use, even after controlling for baseline substance use (OR=2.07, p=.005) in Model 2.
Conclusions and Implications: The findings from this study support prior research on the co-occurrence of delinquency and substance use among high-risk youth. However, our results demonstrate that delinquency may be a more robust predictor of future risk behavior compared to substance use. Given the considerable stability of both behaviors, interventions targeting these risks, and in particular, development of interventions targeting constructive behavior change for delinquency among youth aging-out of foster care are needed.
Note.References omitted for space.