Correlates and predictors of violent behavior among adolescent drinkers

PURPOSE: To examine a wide range of demographic characteristics and psychosocial factors to determine the cross-sectional correlates of violence and longitudinal predictors of violent initiation among adolescent drinkers. METHODS: We conducted secondary analyses of the 1995 (Time 1) and 1996 (Time 2) in-home surveys of the National Longitudinal Study of Adolescent Health (Add Health). This study included a nationally representative school-based sample (N=18,924) of adolescents in grades 7-12. The analyses were restricted to adolescent drinkers (n = 8885). Two logistic regression models were constructed using a backward elimination procedure to identify statistically significant cross-sectional correlates of violence and prospective predictors of violence initiation. RESULTS: Half (49%) of all adolescent drinkers reported violent behavior at Time 1 and 15% of those who were not violent at Time 1 reported initiating violent behavior at Time 2. A total of 14 significant cross-sectional correlates of violence were identified that included measures of alcohol use, drug use and selling, exposure to drugs, delinquency, and poor school functioning. Four variables (high-volume drinking, illicit drug use, low grade point average, and having been suspended and/or expelled from school) were significant longitudinal predictors of the initiation of violent behavior. CONCLUSIONS: The factors significantly associated with violence pertain mostly to alcohol use, drug use and selling, exposure to drugs, delinquency, and poor school functioning. However, most of these problems and behaviors tend to occur in closer temporal proximity to violent behavior (i.e., within a year) and do not seem to developmentally precede initiation in violent behavior.