The differential impact of a classroom-based, alcohol harm reduction intervention, on adolescents with different alcohol use experiences: a multi-level growth modelling analysis
Effectiveness bank home page. Opens new windowResearch analysis

This entry is our analysis of a study considered particularly relevant to improving outcomes from drug or alcohol interventions in the UK. The original study was not published by Findings; click Title to order a copy. Free reprints may be available from the authors – click prepared e-mail. The summary conveys the findings and views expressed in the study. Below is a commentary from Drug and Alcohol Findings.

Links to other documents. Hover over for notes. Click to highlight passage referred to. Unfold extra text Unfold supplementary text
Copy title and link | Comment/query |

The differential impact of a classroom-based, alcohol harm reduction intervention, on adolescents with different alcohol use experiences: a multi-level growth modelling analysis.

McKay M., Sumnall H., McBride N. et al.
Journal of Adolescence: 2014, 37, p. 1057–1067.
Unable to obtain a copy by clicking title? Try asking the author for a reprint by adapting this prepared e-mail or by writing to Dr McKay at

As with the original in Australia, an alcohol harm reduction programme adapted for secondary schools in Northern Ireland slowed down growth in drinking and related problems among the nearly half of pupils who before the lessons had already drunk without adults being present.

Summary The featured article further analyses results from a trial in Northern Ireland of the SHAHRP secondary school alcohol harm reduction programme, the overall impacts of which have previously been analysed for the Effectiveness Bank. Instead of overall impacts, this new analysis asked whether (as in Australia where the programme was developed) harm-reduction effects were greatest among the higher risk pupils who had already drunk without adult supervision. This account first reminds readers of the overall context and findings of the study before moving to the new analysis of which pupils were most affected.

Key points icon

Key points
From summary and commentary

An alcohol harm reduction programme retarded age-related growth in drinking and alcohol-related problems among pupils in secondary schools in Northern Ireland.

Impacts were mainly confined to the nearly half of pupils who before the lessons had already drunk without adults being present.

Findings broadly parallel those from the original programme in Australia and from a similar US programme, suggesting that in Western drinking cultures, harm reduction education helps reduce drinking and harm among the highest risk young teenagers.

Overall impact of the programme

Alcohol harm reduction approaches aim to decrease harm from drinking without requiring abstinence. In schools, such approaches seem most relevant at the ages when young people are first drinking unsupervised by adults and experiencing intoxication.

The programme tested in the School Health and Alcohol Harm Reduction Project (SHAHRP) project featured skills training, information and activities designed to promote behaviour which reduces alcohol-related harms. In the original evaluation in Australia the programme achieved this objective.

SHAHRP was adapted for secondary (or ‘high’) schools in Northern Ireland where it was implemented over two school years – the six lessons of phase one when pupils were in year 10 (ages 13–14) and the four of phase two the following school year. To test its effectiveness, starting in 2005 the featured study recruited 29 schools in the Belfast area. Nine carried on with normal alcohol education (the control schools), the remainder also implemented SHAHRP. In eight of these schools SHAHRP was delivered by the schools’ own teachers after being trained, in 12 by local voluntary-sector drug and alcohol educators. Allocation to the alcohol education options was intended to result in three sets of schools comparable in proportions of boys versus girls, socioeconomic profiles, and locations.

2349 pupils were surveyed at the start of the study; about 60% were girls and they averaged about 14 years of age. Surveys were repeated the following two years after the first and second phases of SHAHRP, and finally in March 2008 when lessons had ended at least 11 months before, at which time 2048 of the 2349 pupils (then averaging 16½ year of age) could be re-surveyed.

Generally the trends in how pupils drank and the harms they experienced were most favourable when SHAHRP lessons had been delivered by external specialists, next when delivered by the schools’ own teachers, and least favourable in the control schools. Compared to those in control schools, pupils offered SHAHRP lessons were more likely to have experienced virtually no harms during the study or a relatively low and stable level, rather than increasing and high levels of harm. When SHAHRP had been delivered by external specialists, pupils were more likely to have experienced virtually no harms than when delivered by the schools’ own teachers. Both types of SHAHRP delivery significantly improved on usual lessons.

The drinkers among the pupils were also asked how much they had drunk last time. Compared to those in control schools, at each follow-up pupils in SHAHRP schools were more likely say they had drunk very little than to have reported increasing and by the end of the study relatively high levels of drinking. Compared to the schools’ own teachers, when the lessons had been delivered by external specialists pupils were more likely to consistently have drunk relatively little.

Which pupils benefited most?

At the start of the trial about a fifth of the pupils indicated that they had not drunk alcohol, a third that they had but only under adult supervision, and just under half said they had already drunk ‘unsupervised’ without adults being present.

When outcome trends were analysed separately for each of these groups, compared to control school pupils, SHAHRP most consistently retarded growth in drinking and related harms among the unsupervised drinkers, especially when the school’s own teachers had delivered the lessons. Only among these more advanced drinkers was drinking itself significantly restrained (both frequency of use and amount drunk on the last occasion), an effect consistently apparent at the final follow-up but patchily before that point. When taught by their own teachers, at each follow-up there was also a retardant effect on the growth in the frequency/range of alcohol-related harms experienced by the pupils due to their own drinking. When external specialists were the teachers, this effect was less strong but still statistically significant at the last two follow-up points. Growth in harms experienced as a result of others’ drinking (such as being verbally abused) was also fairly consistently reduced by the lessons.

In contrast, of the 24 chances for SHAHRP to have registered impacts on drinking or harm among children who had only drunk when adults were present, just one was statistically significant. Among children who at the start of the study had not yet drunk at all, the tally was three statistically significant results, all at the first follow-up, and all in the ‘wrong’ direction – more frequent and heavier last-time drinking and more harm from one’s own drinking in teacher-delivered SHAHRP schools than in control schools.

While drinking and harms were consistently reduced only among unsupervised drinkers, across all the pupils SHAHRP generally improved alcohol-related knowledge and led to safer attitudes to and opinions about drinking. The major exception was among the children who before the lessons had not drunk at all, whose attitudes/opinions did not improve relative to control schools when SHAHRP had been taught by the schools’ own teachers.

The authors’ conclusions

Results of the present study suggest that regardless of whether they have tried alcohol or not and whether under adult supervision, SHAHRP has a significant and positive effect on pupils’ knowledge about and attitudes towards alcohol. To this extent its description as a ‘universal’ intervention is supported.

However, relative to usual education, significant reductions in drinking were seen only among children who at the start of the study had drunk alcohol without adults being present – an instance of the common finding that attitudes and knowledge do not predict behaviour. Among these pupils at increased risk of alcohol-related harm, beneficial effects persisted to the final follow-up, suggesting SHAHRP can help prevent the harms they would otherwise have experienced. These results were found whether the lessons were delivered by the schools’ teachers or by external experts, but the teachers seemed to create the strongest impact.

Findings from SHAHRP in Australia and Northern Ireland are broadly consistent with those reported elsewhere for other prevention programmes, where the intended impacts have been observed for those most at risk of alcohol-related harm.

Findings logo commentary Consistent findings in different countries with different sets of lessons ( below) suggest that alcohol harm reduction education reduces underage drinking and resultant harm in Western drinking cultures, and usually does most effectively where these effects are most needed – among youngsters already engaged in drinking in their early teens. Such findings make harm reduction education relevant to drinking cultures such as the UK where youth drinking (though declining) remains common. However, in Northern Ireland the effects were not substantial; despite some statistically significant differences in trends over time, whichever lessons their schools had been allocated to, unsupervised drinkers ended up at very similar levels of drinking, and difference in harm levels were small. SHAHRP-type lessons may help, but at a national level preventing youth or adult alcohol-related harm requires stronger levers than education in the teenage years.

Generally studies of harm reduction education have found no indication that this approach led more pupils to start drinking – usually the opposite. In Australia, though still very much in the minority, by the last follow up there were a third more abstainers among SHAHRP than control pupils. But in Northern Ireland some findings were consistent with harm reduction lessons from their own teachers initially giving non-drinkers ‘permission’ to try drinking. After the first set of lessons the drop in the proportion of non-drinkers had been steeper in SHAHRP teacher-led schools, where there were proportionately 60% fewer non-drinkers compared to 46% and 48% fewer in the other two sets of schools. The featured report found that relative to control schools, attitudes to and opinions about drinking among initial non-drinkers did not improve in teacher-led SHAHRP schools, and after the first year of lessons, these pupils also drank more heavily and experienced more harms. By the end of the study any possible counterproductive impacts on drinking and harm among initial non-drinkers had worked their way out and what remained were the statistically significant benefits among initial unsupervised drinkers.

Some methodological issues might have affected the results of the featured study. The general failure to find effects on drinking and harm among youngsters who had drunk, but only in the company of adults, may be due to their being below the risk threshold where SHAHRP’s impacts become apparent. This group was likely to have included many children who were not ‘drinkers’ at all, but had merely had a taste or sip of an alcoholic drink at a family meal, and were at low risk of developing harmful drinking patterns. Pupils in SHAHRP schools probably spent much more time in alcohol-related lessons than those in control schools, meaning that the programme’s impacts might have been due to the extensity of the teaching rather than its distinctive methods or content. In Australia too, compared to usual alcohol education SHAHRP occupied more classroom time overall and spanned two years rather than one. The non-randomised design of the study and the withdrawal of two control schools mean the results might be due to differences between schools and pupils not accounted for by the study. In particular, some of the extra steepness in the escalation of drinking and harm in control schools might have been due to pupils on average starting at lower levels and ‘catching up’ in ways which might have happened even without SHAHRP to help restrain trends in the other schools. Despite some significant trend differences, at the end of the study the difference between the levels the pupils ended up at was very small, and in respect of amount drunk on last occasion, slightly lower in control schools than when SHAHRP had been delivered by external experts. The measure of harm from the pupil’s own drinking included planning to get drunk, arguably not a harm but something which might lead to harm or even help avoid it depending on the planning, and excluded alcohol-influenced risky or later regretted sexual intercourse, common harms in the Australian study.

Findings similar in Australia and USA

Findings in Northern Ireland paralleled those from the original programme in Australia, where harm-reduction effects were greatest among the higher risk pupils who had already drunk without adult supervision; at each follow-up point they experienced about 20% fewer harms than control pupils. A derivative of SHAHRP has also more recently been evaluated in Australia, where compared to control schools it retarded age-related increases in the amount pupils drank and the resultant harms they experienced. As in Northern Ireland and the previous Australian study, the harm reduction programme seemed most beneficial in terms of drinking and resultant harms among children who started the trial most engaged with drinking – in this case, not defined as unsupervised drinkers, but the roughly a fifth of pupils who at the start of the trial usually consumed five or more drinks (each 10g of alcohol) when they drank. Though this was not the case in a smaller Australian pilot study, still the pattern of results is indicative of SHAHRP’s potential with these higher risk teenagers, who may see the lessons as more relevant to them and who may already have experienced the harms it aims to help them avoid.

Corroboration comes from the more restrictive youth drinking environment of the USA, where a programme forefronting alcohol problem reduction among its aims has produced similar findings to that in Australia. It retarded growth in alcohol problems (such as getting drunk or sick or complaints from parents and friends), but only among pupils who had already drunk without adult supervision, and only if the lessons did not occur too early to coincide with the development of this drinking pattern. After disappointing initial results, another US substance use education programme adopted harm reduction objectives. The revised programme resulted in a significant reduction in risky or harmful drinking.

Such findings contrast with unconvincing evidence from trials of substance use education in general and alcohol education in particular.

UK policy and practice

For the UK the most important guidance on alcohol education was issued in 2007 by the National Institute for Health and Clinical Excellence. It said education “should aim to encourage children not to drink, delay the age at which young people start drinking and reduce the harm it can cause among those who do drink”. Recommendations included ensuring alcohol education is an integral part of science and personal, social and health education (PSHE) curricula. The committee stressed that education should be adapted to its cultural context, noting that in the UK “alcohol use is considered normal for a large proportion of the population [and] a ‘harm reduction’ approach is favoured for young people”.

Inspections in 2012 of PSHE lessons suggest English schools are far from adequately implementing NICE’s recommendations, in particular in respect of education aimed at reducing alcohol-related harm. Only in just under half the inspected schools had pupils learnt how to keep themselves safe in a variety of situations, and the deficits were particularly noticeable in respect of drinking. Inspectors found that although pupils understood the dangers to health of tobacco and illegal drugs, they were far less aware of the physical and social damage associated with risky drinking. Some did not know the strength of different alcoholic drinks or make the links between excessive drinking and issues such as heart and liver disease and personal safety. The report attributed these deficiencies in part to inadequacies in subject-specific training and support for PSHE teachers, particularly in teaching sensitive and controversial topics.

Last revised 21 July 2015. First uploaded 13 July 2015

Open Effectiveness Bank home page

Top 10 most closely related documents on this site. For more try a subject or free text search

STUDY 2012 Reducing the harm from adolescent alcohol consumption: results from an adapted version of SHAHRP in Northern Ireland

STUDY 2015 Alcohol prevention and school students: findings from an Australian 2-year trial of integrated harm minimization school drug education

STUDY 2018 Alcohol prevention for school students: Results from a 1-year follow up of a cluster-randomised controlled trial of harm minimisation school drug education

STUDY 2012 Alcohol prevention: What can be expected of a harm reduction focused school drug education programme?

STUDY 2018 Steps Towards Alcohol Misuse Prevention Programme (STAMPP): a school-based and community-based cluster randomised controlled trial

REVIEW 2019 Family-based prevention programmes for alcohol use in young people

STUDY 2013 PROSPER community-university partnership delivery system effects on substance misuse through 6½ years past baseline from a cluster randomized controlled intervention trial

STUDY 2011 Effects of a school-based prevention program on European adolescents' patterns of alcohol use

STUDY 2004 School lessons reduce alcohol-related harm

STUDY 2010 Does successful school-based prevention of bullying influence substance use among 13- to 16-year-olds?