FRANÇAIS :
Cet article montre le rôle modérateur des parents, en particulier sur l'influence des amis, concernant l'usage de drogues (tabac, alcool et cannabis) chez les adolescents. Les données proviennent d'une vaste enquête en milieu scolaire menée sur 18 mois dans l'Indiania, USA. Les déclarations des adolescents sont utilisées pour évaluer leurs attitudes vis-à-vis de l'usage de drogues. Plus grand est le nombre de parents et d'amis consommant des drogues, plus grand est le risque que l'adolescent en consomme. Avec l'augmentation de l'offre de produit, on observe une baisse de la capacité de résistance, capacité à dire non. Mais lorsque les parents ne sont pas consommateurs, l'influence des amis consommateurs est endiguée, et l'offre de produits a moins d'impact sur la capacité de refus. Les parents devraient donc être des cibles privilégiées pour les programmes de prévention de l'usage de drogues chez les adolescents.

ENGLISH :
Aims. A major question in national substance use prevention efforts is how much influence parents have on their children's substance use, answers to which could determine whether current school-based prevention programs that address peer pressure are sufficient or whether parents need to be involved. The present study examines whether parents act as gatekeepers by testing the moderator effects of parents' substance use on the relationships of friends' substance use to adolescent substance use (cigarettes, alcohol and marijuana). Design. A longitudinal school-based trial for prevention of substance use in adolescents. Setting. Indianapolis, Indiana. USA. Participants. As part of a large substance use prevention trial, entering middle school students were surveyed prospectively at baseline, 6 months and 18 months (n = 1807 from 57 schools). Measurements. A self-reported survey was used to measure behaviors, attitudes and social influences related to adolescent substance use. Findings. Results of logistic regression analyses and multiple group structural equation modeling showed that increasing numbers of parents and friends using substances were associated with greater risk of adolescent substance use, as were more substance offers and lower levels of refusal self-efficacy. Additionally, refusal self-efficacy mediated the effects of baseline use and substance offers on subsequent use. However, non-using parents had a buffering effect on friends' influences to use substances, such that friends' use did not affect adolescent use when parents were non-users, and the effects of substance offers on refusal self-efficacy were weaker. Conclusions. The findings suggest that parent substance use should be addressed in adolescent substance use prevention programs, and that continuing non-use by parents should be reinforced. (Author' s abstract)