Cannabis use initiation among adolescents: the predictive role of peers, alcohol, expectancies and internalizing factors.

Theoretical background: Cannabis is the most commonly used illicit drug among teenagers. Initiation is influenced by environmental factors and personal characteristics, events or experiences. An early initiation increases the risks of problems related to cannabis use (e.g. negative psychosocial effects, delinquent behaviors, mental disorders). Prevent initiation, through the identification of risk/protective factors and their interaction, seems crucial in terms of public health. Research question and significance: This study aimed to examine the progression of use, to identify protective and risk factors of cannabis initiation (including peers, alcohol, expectancies and internalizing factors) and to specifically focus on the influence of social anxiety and its moderators/mediators. Methods: A questionnaire was administered twice to 877 teenagers (49.94% female, M=15.61) with one year interval. Sex, age, demographic variables, peer cannabis use, cannabis-related variables (lifetime, frequency, problems and expectancies), alcohol use, social anxiety, trait-anxiety and depression were assessed through validated scales. Logistic regressions, mediation and moderation analyzes were performed. Results: During the follow-up period, 12.89% of the young participants initiated cannabis use. Several factors significantly predict initiation: alcohol use, peer users, perceptual enhancement and craving effect expectancies. Others factors significantly protect from initiation: negative behavioral effect expectancies and social anxiety. Gender, age, relaxation/social facilitation and cognitive impairment effect expectancies, trait-anxiety and depression do not significantly influence cannabis use initiation. In moderated mediation model, after controlling for relevant variables, social anxiety protected from initiation trough the mediating role of perceptual enhancement and craving effect expectancies. The number of peer users and alcohol use do not moderate this mediation. Negative behavioral effect expectancies do not significantly mediate the relation between social anxiety and cannabis initiation. Interpretation of findings: Through low positive expectancies, adolescents with social anxiety symptoms are less likely to initiate cannabis use than the others, whatever the number of peer users and the alcohol use. Findings are discussed in terms of risk and protective characteristics of relevant factors, in an overall and evolutionary approach including internalizing factors. Results support the identification of internalizing profile of adolescents concerned by prevention or treatment and the importance of social anxiety and expectancies in intervention