Smoking causes five million preventable deaths every year, a number predicted to rise to eight million by 2030. It is thought that around a quarter of young people may smoke by age 13-15. With a history spanning four decades, prevention programmes in schools try to tackle smoking at an early age before the habit becomes difficult to break. The systematic review aimed to resolve the uncertainty of whether the programmes are effective in preventing smoking.

The researchers analysed data from 134 studies, in 25 different countries, which involved a total of 428,293 young people aged 5-18. Of these, 49 studies reported smoking behaviour in those who had never previously smoked. The researchers focused on this group because it offered the clearest indication of whether smoking interventions prevent smoking. Although there were no significant effects within the first year, in studies with longer follow up the number of smokers was significantly lower in the groups targeted by smoking interventions than in the control group. In 15 studies which reported on changes in smoking behaviour in a mixed group of never smokers, previous experimenters and quitters, there was no overall long term effect, but within the first year the number smoking was slightly lower in the control group.

"This review is important because there are no other comprehensive reviews of world literature on school-based smoking prevention programmes," said Julie McLellan, one of the authors of the review based at the Department of Primary Care Health Sciences at the University of Oxford in Oxford, UK. "The main strength of the review is that it includes a large number of trials and participants. However, over half were from the US, so we need to see studies across all areas of the world, as well as further studies analysing the effects of interventions by gender."

Some smoking interventions aimed to develop social skills and competency or teach students to resist social pressure to smoke. In the longer term, subgroups of programmes that used a social competence approach or a combined social competence and resistance training approach had a significant effect on preventing smoking in young people who had never smoked at the time of the intervention. No effect was shown in studies that used resistance training alone. Equally, no significant effects were detected in those programmes that used information only or that combined tobacco education with wider school and community initiatives.

Booster sessions did not make any difference to the number of young people who went on to smoke. "We might expect booster sessions to reinforce the effects of the original programmes, but our review didn't find any evidence for this overall," said co-author Rafael Perera, also from the Department of Primary Care Health Sciences. "The exception was in studies focusing on general social competence and resistance training, which suggests that the way the curriculum is designed is more important than whether or not booster sessions are provided."