Dennis Grevenstein1, Christoph Nikendei2, Ede Nagy2. 1. Psychological Institute, University of Heidelberg, Heidelberg, Germany. 2. Department of General Internal Medicine and Psychosomatics, Center for Psychosocial Medicine, University Hospital Heidelberg, Heidelberg, Germany.
Abstract
Background: A number of correlates to adolescent substance use have been shown, yet their unique predictive influence is unclear. We investigated the incremental validity of demographics, family background, school variables, risk perception, parental alcohol attitudes, age of first alcohol use, peer group influence, and behavioral variables as they concurrently predicted last month alcohol use, binge drinking, and drunkenness experience frequency. Methods: Hierarchical multiple regression analyses in a sample of N = 743 adolescents (mean age = 15.01). Results: All predictors explained 26-40% of the total variance. Alcohol use was predicted by age, gender, having two working parents, academic family background, relationship to parents, relative risk perception, parental acceptance of alcohol use, age of first use, talking about positive aspects of alcohol, normality of alcohol use among peers, time spent with friends, and going out to clubs. Binge drinking was predicted by age, working parents, school problems, relative risk perception, parental alcohol acceptance, age of first use, talking about positive aspects of alcohol, time spent with friends, and going out to clubs. Drunkenness was predicted by age, relationship to parents, school problems, relative risk perception, age of first use, talking about positive aspects of alcohol, and going out to clubs. Conclusions: Researchers need to take the complexity of adolescent substance use into account when designing studies and interventions. Relative risk perception emerged as the strongest (positive) predictor, indicating that adolescents are able to rather accurately assess their own risk and risk awareness alone does not suffice to engage in protective behavior.
Background: A number of correlates to adolescent substance use have been shown, yet their unique predictive influence is unclear. We investigated the incremental validity of demographics, family background, school variables, risk perception, parental alcohol attitudes, age of first alcohol use, peer group influence, and behavioral variables as they concurrently predicted last month alcohol use, binge drinking, and drunkenness experience frequency. Methods: Hierarchical multiple regression analyses in a sample of N = 743 adolescents (mean age = 15.01). Results: All predictors explained 26-40% of the total variance. Alcohol use was predicted by age, gender, having two working parents, academic family background, relationship to parents, relative risk perception, parental acceptance of alcohol use, age of first use, talking about positive aspects of alcohol, normality of alcohol use among peers, time spent with friends, and going out to clubs. Binge drinking was predicted by age, working parents, school problems, relative risk perception, parental alcohol acceptance, age of first use, talking about positive aspects of alcohol, time spent with friends, and going out to clubs. Drunkenness was predicted by age, relationship to parents, school problems, relative risk perception, age of first use, talking about positive aspects of alcohol, and going out to clubs. Conclusions: Researchers need to take the complexity of adolescent substance use into account when designing studies and interventions. Relative risk perception emerged as the strongest (positive) predictor, indicating that adolescents are able to rather accurately assess their own risk and risk awareness alone does not suffice to engage in protective behavior.