Association between tobacco control policies and smoking behaviour among adolescents in 29 European countries (2009)

AIMS: To investigate the associations between well-known, cost-effective tobacco control policies at country level and smoking prevalence among 15-year-old adolescents.
DESIGN: Multi-level modelling based on the 2005-06 Health Behaviour in School-aged Children Study, a cross-national study at individual level, and with country-level variables from the Tobacco Control Scale and published country-level databases.
SETTING: Twenty-nine European countries.
PARTICIPANTS: A total of 25 599 boys and 26 509 girls.
MAIN OUTCOME MEASURES: Self-reported regular smoking defined as at least weekly smoking, including daily smoking (dichotomous).
FINDINGS: Interaction effects between gender and smoking policies were identified, therefore boys and girls were analysed separately. Large cross-national differences in smoking prevalence were documented. Intraclass correlations (ICC) of 0.038 (boys) and 0.035 (girls) were found. In the final multi-level model for boys, besides the significance of the individual variables such as family affluence, country-level affluence and the legality of vending machines were related significantly to regular smoking [b(country affluence) = -0.010; b(partial restriction vending machines) = -0.366, P < 0.05]. Price policy was of borderline significance [b(price policy) = -0.026, P = 0.050]. All relationships were in the expected direction. The model fit is not as good for girls; only the legality of vending machines had a borderline significance in the final model [b(total ban vending machines) = -0.372, P=0.06].
CONCLUSIONS: For boys, some of the currently recommended tobacco control policies may help to reduce smoking prevalence. However, the model is less suitable for girls, indicating gender differences in the potential efficacy of smoking policies. Future research should address this issue.