Not magic at all of course, but a consequence of the fact that substance use problems are closely related to other problems which often develop at early ages when substance use is just not on the agenda. The 2010 English national drug strategy and corresponding public health plans seemed to recognise this, breaking with previous versions to focus attention on early years parenting in general, and particularly among vulnerable families.

Though studies are few compared to approaches such as drug education (1) in schools (for which, see this Effectiveness Bank hot topic), this renewed emphasis on the early years has a strong theoretical rationale and some research backing. Child development and parenting programmes which do not mention substances at all (or only peripherally) have recorded some of the most substantial prevention impacts. Though mainly targeted at the early years, some extend to early teenage pupils and their families. The rationale for intervention rests partly on strong evidence that schools which develop supportive, engaging and inclusive cultures, and which offer opportunities to participate in school decision-making and extracurricular activities, create better outcomes across many domains, including non-normative substance use. As well as facilitating bonding with the school, such schools are likely to make it easier for pupils to seek and receive the support they need.

Understandably, such findings do not derive from random allocation of pupils to ‘good’ versus ‘bad’ schools, so are vulnerable to other influences the study was unable to account for. More convincing, if more limited in intervention scope, are studies which deliberately intervene and test what happens among young people randomly allocated to the focal intervention versus a comparator. An early example was a seminal Dutch drug education study of the early ’70s. So profound was its impact in Britain that we have analysed it in detail, contacting the original author. The surprising findings are outlined…………