Abstract
AIM: The current study sought to identify unique and common demographic and parental predictors of adolescent tobacco and alcohol initiation behaviors. METHODS: A total of 1023 middle school students (Grades 6-8) and their parents from six Rhode Island schools were enrolled in a larger study on substance use initiation and progression, with the current sample representing those dyads with data at baseline and at a 12-month follow-up (n=814 dyads). There was a relatively even split with regard to adolescent sex (nfemales=444; 55%). Comparisons were made between covariate and predictor associations with corresponding substance use behaviors (e.g., ever puffed a cigarette vs. ever sipped an alcoholic drink; ever smoked a whole cigarette vs. ever drank a full alcoholic drink). RESULTS: At the bivariate level, a host of demographic and parent-related variables were associated with each adolescent substance use behavior. Adolescent reports of parental monitoring variables were consistently more related to use than parent reports. In multivariate logistic regression analyses, adolescent reports of parental monitoring variables were more frequently associated with tobacco use behaviors than alcohol use behaviors. Tobacco use behaviors were more strongly predicted by perceived availability of tobacco than alcohol use behaviors were predicted by perceived availability of alcohol. CONCLUSIONS: The distinctive predictors observed for cigarette versus alcohol use make it problematic for new and existing programs to assume that efforts targeting specific individual or parental characteristics will impact both substances with equivalent efficacy.