Bridging Disciplinary Boundaries (January 11 - 14, 2007)

Purpose: The prevention of underage drinking has become an important priority in the United States, with increased attention given to identifying factors that protect youth from developing alcohol problems. This study examines the impact of three sets of protective factors (i.e., individual, familial, and extra-familial) on problem drinking among a nationally representative sample of adolescents. The aim is to determine the nature of these influences on drinking. We tested three types of protective effects—direct, cumulative, and moderating—cross-sectionally and over time. Methods: The sample includes 10,708 youth in grades 7-12 who participated in the 1st and 2nd waves of the National Longitudinal Study of Adolescent Health and whose parents completed interviews at wave 1. Two outcome measures, a composite of experiencing drinking-related negative consequences and intoxication at both waves are regressed on risk factors (i.e., alcohol availability in the home, peer alcohol use, delinquency, academic risk) and individual (i.e., positive outlook, high self worth, good decision-making), familial (parent availability, parent closeness), and extra-familial (caring adults, caring teachers, school connection, religious connection) protective factors at wave 1. Hierarchical linear regression was used to examine the relative direct, cumulative, and moderating impact of individual, familial, and extra-familial protective factors on problem drinking. Results: Three of the risk factors (peer use, delinquency, and academic risk) are strongly related to both outcome measures. These relations are much stronger than the protective factor measures. Alcohol availability is not related to problem drinking in any model. Controlling for risk factors and other demographic variables, the direct effects of protective factors vary by outcome measure. For the negative consequences outcome measure, positive outlook provides a protective influence at time 1 but not time 2. Mother availability is significant in all models and across both time points. Religious connection protects against experiencing negative consequences only at time 2. For the intoxication measure, high self worth, mother availability, caring teachers, and religious connection provide protective influences at time 1. Yet, only the extra-familial measures remain significant over time. Results of cumulative effects analyses show that none of the protective factor indices are related to the negative consequences measure but do protect against intoxication. Findings of moderation analyses also show differential influences by outcome measure. For the negative consequences composite, high self worth, mother availability, and religious connection reduce the impact of risk factors on problem drinking across time. Only the extra-familial protective factors, and not individual or familial protective factors, reduce the impact of risk factors on intoxication over time. Implications: These findings have important implications for protective factor approaches to preventing problem drinking during adolescence. First, results provide some support for previous evidence of the benefits of family based approaches to reducing problem drinking among youth. Second, at this stage in children's life cycles, extra-familial pro-social connections (i.e., with teachers or religious institutions) appear to be more influential than individual measures in protecting youth against problem alcohol use. Finally, interventions should include strategies to counteract the multiple ways in which youth engage in problem drinking.