Predicting the life-time benefit of school-based smoking prevention programmes.

Aim: School-based smoking prevention programmes may delay the age of smoking initiation, but do not appear to achieve lasting reductions in smoking prevalence beyond school-leaving age. We explored whether delaying the age at which someone initiates smoking may have life-time benefits by increasing the likelihood of quitting in later life.

Design And Setting: Data from the General Household Survey of Great Britain were used in a logistic regression model to examine the association between age at which someone initiates regular smoking and the probability that the person will quit smoking later in life. The effect of confounding variables (sex, ethnicity, socio-economic class, education and geographical location) was taken into account. The predicted relationship was used in a cohort model to estimate the life-time reduction in smoking prevalence and all-cause mortality of a school-based smoking prevention programme.

Results: Age of regular smoking initiation was associated strongly with the probability of quitting later in life (coefficient -0.103, P < 0.001). The strength of the association was slightly reduced but still significant when confounding variables were included (coefficient -0.075, P < 0.001). An intervention that delays smoking initiation without decreasing smoking prevalence at age 18 may reduce adult smoking prevalence by 0.13-0.32% (depending on age) and all-cause mortality by 0.09% over the life-time of the sample.

Conclusion: School-based smoking prevention programmes have potential for a beneficial effect over the life-time of the participants even if they have no apparent effect at school-leaving age.