School smoking policies and smoking prevalence among adolescents: multilevel analysis of cross-sectional data from Wales

摘要

OBJECTIVE—To examine the association between school smoking policies and smoking prevalence among pupils.
DESIGN—Multilevel analysis of cross-sectional data from surveys of schools and pupils.
SETTING—55 secondary schools in Wales.
SUBJECTS—55 teachers and 1375 pupils in year 11 (aged 15-16).
MAIN OUTCOME MEASURES—Self-reported smoking behaviour.
RESULTS—The prevalence of daily smoking in schools with a written policy on smoking for pupils, teachers, and other adults, with no pupils or teachers allowed to smoke anywhere on the school premises, was 9.5% (95% confidence interval (CI) 6.1% to 12.9%). In schools with no policy on pupils' or teachers' smoking, 30.1% (95% CI 23.6% to 36.6%) of pupils reported daily smoking. In schools with an intermediate level of smoking policy, 21.0% (95% CI 17.8% to 24.2%) smoked every day. School smoking policy was associated with school level variation in daily smoking (p = 0.002). In multilevel analysis, after adjusting for pupils' sex, parents' and best friends' smoking status, parental expectations, and alienation from school, there was less unexplained school level variation, but school smoking policy remained significant (p = 0.041). The association of smoking policy with weekly smoking was weaker than for daily smoking, and not significant after adjustment for pupil level variables. Both daily and weekly smoking prevalence were lower in schools where pupils' smoking restrictions were always enforced. Enforcement of teacher smoking restrictions was not significantly associated with pupils' smoking.
CONCLUSIONS—This study demonstrates an association between policy strength, policy enforcement, and the prevalence of smoking among pupils, after having adjusted for pupil level characteristics. These findings suggest that the wider introduction of comprehensive school smoking policies may help reduce teenage smoking.


>Keywords: schools; smoking policy; smoking prevalence