We tested a sequential regression model separately for each outcome controlling for family SES and cohort. In the first step we entered victimization status, race, and the controls (see Table 3). In the second step, we entered the interaction between race and victimization status to the main effects model. With the controls in the model, maltreatment was a significant predictor of depressive symptoms and violence but not heavy drinking in adolescence. Maltreatment was not significantly related to any outcome at age 24/25. Race was significantly associated with heavy drinking and violence during adolescence and all outcomes during young adulthood, with Whites reporting more heavy drinking and Blacks reporting higher depression and perpetrating more violence. Lower family SES predicted more violence. During adolescence, boys in the oldest cohort reported significantly more heavy drinking, violence, and depressive symptoms; during young adulthood the youngest cohort reported more symptoms of depression (note 7). There were no significant race-by-maltreatment interactions, suggesting that maltreatment played a similar role in adolescent depressive symptoms and violence for both Blacks and Whites.
