Individual and contextual determinants of perceived peer smoking prevalence among adolescents in six European cities.

Background: Young people perceiving a high peer smoking prevalence are more likely to initiate smoking. It is unclear which factors contribute to perceived peer smoking prevalence and if these factors vary according to education. This study aimed to assess the determinants of perceived smoking prevalence and assessed its variation at school and country-level.

Methods: Data of 10,283 14-17-year-old students in 50 secondary schools in six European cities were derived from the 2013 SILNE survey. The outcome was the perceived smoking prevalence score among peers at school (0-10 scale, 10 represented 100% smoking prevalence). Multilevel linear regression models estimated the associations of factors with perceived prevalence score and variance at school and country-levels. Analyses were also stratified by academic achievement of the adolescent and parental education.

Results: Determinants of a higher perceived prevalence score were female sex, ever smoking, having friends who smoke, low academic achievement, low parental educational level, and higher actual prevalence of smoking in the school. The perceived prevalence score was not associated with school policies or with the availability of cigarettes near the school. Determinants were very similar across levels of academic achievement and parental education. Perceived prevalence scores substantially varied between schools and countries: 10% and 11% of total variance was related to schools and countries respectively.

Conclusion: Across educational levels, perceptions of peer smoking are strongly determined by both individual characteristics and school and national contexts. Future studies should assess why perceived smoking prevalence varies between schools and countries and identify modifiable factors.