Danielle C. Hill and Sylvie Mrug, Ph.D. University of Alabama, Birmingham, AL


Alcohol and cigarette use during adolescence is a serious public health concern (Shih et al., 2010). Epidemiological studies show consistent racial and gender differences in substance use (Johnston et al., 2009), but the reasons for these differences remain poorly understood. Many of the risk factors for substance use, parent-child communication, parenting, and antisocial behavior of the children and their friends also evidence gender and racial differences (Bray et al., 2001; Dishion & Owens, 2002; Hawkins et al., 1997; Stoker & Swadi, 1990), suggesting that they may explain racial and gender differences in substance use. It is important to examine these differences in early adolescence, because early age of substance use initiation is associated with detrimental outcomes in adulthood (Odgers et al., 2008). The current study investigates whether there are racial and gender difference in early adolescence substance use. Additionally, we examine whether specific child, family, and peer factors explain these differences. Participants included 703 early adolescents who took part in Wave 1 of the Birmingham Youth Violence Study. The mean age of the students was 11.8 years old; the sample was 53% male, 76% African American, and 22% Caucasian. Adolescents reported on their alcohol and cigarette use, friends’ deviant behavior, victimization by bullying, parenting, and parental conflict. Primary caregivers provided information about children’s antisocial behavior, their own level of education, and family income. Logistic regression and independent sample t-tests analyses were performed to evaluate racial and gender differences in early adolescent substance use and potential mediators of these relationships. Results revealed gender differences in alcohol use but not smoking, with boys being more likely to report using alcohol than girls. No racial differences in substance use emerged. Of the potential mediators, adolescents’ antisocial behavior, delinquent peers, and parental conflict differed between both boys and girls and alcohol users and non-users. When these variables were added to the logistic regression model, gender differences in alcohol use disappeared. The results suggest that higher parental conflict, antisocial behavior, and friends’ deviant behavior explain gender differences in early alcohol use.