Prospective study of the effect of exposure to other smokers in high school tutor groups on the risk of incident smoking in adolescence.

Adolescent smokers tend to have peers who smoke, but it is unclear whether this arises from self-selection of smoking peers or whether this is a causal effect on the uptake of smoking (incident smoking). The authors used school tutor group current smoking prevalence, an unbiased measure of peer smoking in high schools in the United Kingdom, to estimate the independent effect of peer smoking on incident smoking among high school students. In a prospective cohort study of children aged 13-15 years (grades 9 and 10) in 10 high schools in Nottinghamshire, United Kingdom, smoking behavior was surveyed in 2,881 students in 2000, and the survey was repeated in 2,109 students (73%) in 2001. There were 267 incident smokers (15%) among the 1,766 nonsmokers in 2000. The adjusted odds of incident smoking were significantly higher in girls, in students with parents or siblings who smoke, and in relation to school tutor group current smoking prevalence in 2000 (relative odds for highest relative to the lowest quartile of prevalence = 1.78, 95% confidence interval: 1.20, 2.64). This tutor group effect was independent of having a best friend who smoked in the 2001 study. Incident smoking is therefore increased among students exposed to other students who smoke, and preventing smoking at school may reduce adolescent smoking.