Send email for updates

Effectiveness bank home page. Opens new windowReview analysis

This entry is our analysis of a review or synthesis of research findings considered particularly relevant to improving outcomes from drug or alcohol interventions in the UK. The original review was not published by Findings; click Title to order a copy. The summary conveys the findings and views expressed in the review. 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 |

Family-based prevention programmes for alcohol use in young people.

Gilligan C., Wolfenden L., Foxcroft D.R. et al.
Cochrane Database of Systematic Reviews: 2019, Issue 3. Art. No.: CD012287.

Findings of this comprehensive review seem to almost entirely deflate what in the mid-2000s was a bubble of enthusiasm for parental programmes as a way to prevent or reduce drinking among teenagers – but despite this overall verdict, some interventions have had remarkable results.

Summary This review updates one published in 2011 analysed for the Effectiveness Bank, and extends its analyses of family-based interventions to prevent or reduce drinking among young people aged up to 18 beyond ‘universal’ interventions to those classed as ‘selective’ or ‘indicated’.

Key points icon

Key points
From summary and commentary

Comprehensive review amalgamated findings from studies of family-based interventions which assessed whether they prevented or reduced drinking among school-age young people up to 18 years old.

Overall the review found little evidence that interventions with parents or families applied universally to all children, to high-risk groups, or to children already drinking, reduced adolescent drinking compared to no intervention, standard care, or a child-focused intervention alone.

There were some more promising findings, but these derived from few studies conducted by even fewer lead researchers and their results cannot be relied on as an indication of what might be achieved if the ‘best’ interventions were selected to be widely implemented, or drinking was assessed over a longer period than the up to four years analysed for the review.

In the context of the sector being investigated, ‘universal’ interventions are those which target all the parents within a population regardless of the risk that their children might become drinkers, usually aiming to delay the start of drinking among the children or to reduce how often or how much they drink. With similar aims, ‘selective’ interventions target parents whose children have an elevated risk of substance use due to social or family factors. ‘Indicated’ interventions target parents whose children are already known to drink, generally with a view to reducing consumption, the frequency of heavy drinking, or alcohol-related harms.

Approaching prevention via the parents is thought to work by promoting appropriate parenting strategies likely to develop positive social norms in the children and help them resist negative influences among their peers and the broader society. These strategies include rule-setting, appropriate communication, monitoring the children’s activities, and conveying positive values and attitudes.

Such programmes are often appended to school-based prevention curricula for the children, but may also be standalone interventions. Commonly they focus on parent-child communication and relationship building, developing the child’s social competence and self-regulation skills, and intensifying the parents’ involvement with their children.

To assess whether overall these types of programmes reduce drinking, the reviewers searched for studies which had randomly allocated school-age children (aged up to 18) and their parents to a parent-based alcohol prevention intervention versus either an alternative programme or no special programme at all. Preference was given to outcomes reported or estimated across the entire sample rather than just children followed up or whose families completed the interventions, and to drinking as assessed at the latest follow-up within four years of the intervention. When studies adjusted their results for other influences on drinking, these adjusted figures were used. Rather than individual children, many studies allocated groups such as classes or schools to the intervention and the comparator. If the study did not adjust for the bias this can generate, the reviewers did, and used the adjusted figures in their analyses.

In total 46 trials (29 from the USA) were found involving 39,822 participants or families. In 35 trials the comparison was with usual care or no intervention. A further 12 compared the effects of an intervention aimed at the children with the same intervention coupled with family/parent components. Universally applied interventions were trialled in 27 studies, selective in 12, and indicated in seven. Children were on average older in trials of selective (about 13 years) or indicated (about 15.5 years) than universal interventions (about 12 years).

Universal and selective interventions were delivered to parents in a variety of ways, including print, CDs or videos, via internet/computers, presentations or workshops at the child’s school, face-to-face group, individual or family sessions at the school or a community venue, or parent/family sessions in the home or at a healthcare setting. All indicated interventions were delivered through face-to-face sessions with parents and children separately and/or together. In half the trials the children too were involved in the intervention through classroom curriculum or other adolescent-focused resources, or face-to-face sessions in individual, group or family formats.

To analyse the outcomes of these trials the reviewers amalgamated data from studies where interventions and outcome measures were considered similar using meta-analytic techniques. The outcomes pooled were:
prevalence of alcohol use, ie, the proportion of children who had drank alcohol either ever, in the last six months, or in the past week;
frequency of drinking, including the number of occasions in the last 30 or 90 days; and
volume or amount of alcohol consumed, including the number of drinks in the previous 30 days, or a score generated from the quantity typically drank on each occasion and frequency measures.

Main findings

Findings are presented first for studies which compared the evaluated intervention with no special intervention at all, then which compared the effects of adding family/parent components to an intervention aimed at the children. In each case, amalgamated findings are presented for all studies, and in so far as this is appropriate and feasible, separately for those testing universal, selective or indicated interventions.

Parental interventions versus no intervention or usual care

Generally these analyses found no statistically significant differences between the outcomes of parental interventions versus no intervention or usual care, meaning that it could not be concluded that parental interventions reduced drinking.

Across all the studies the only statistically significant finding on the prevalence, frequency or volume of drinking, was a very small difference indicating a reduction in volume after parental interventions compared to no intervention or usual care. This became marginally insignificant when the analysis was confined to studies at lesser risk of producing biased results. A non-significant, but relatively large, reduction in frequency of drinking after parental interventions compared to no intervention or usual care became marginally reversed when studies at a higher risk of bias were excluded.

Findings were similar when the analysis narrowed in on studies of interventions applied universally to all children regardless of their risk of drinking or drink-related problems. There were no statistically significant results for prevalence or frequency of drinking, and again the only statistically significant finding was a small difference indicating a reduction in volume after a parental intervention compared to no intervention or usual care. This was found across three studies whose results could be amalgamated. Another three whose results could not be included in the analysis also reported statistically significant reductions in the volume of drinking after a parental intervention compared to no intervention or usual care.

Since there were so few of these studies, results from those trialling selective or indicated interventions were pooled. There were no statistically significant differences in respect of any of the drinking outcomes either across all studies or those involving different types of children/families, after different follow-up periods, or more or less intensive interventions. The picture was similar across studies whose results could not be included in the amalgamated findings. The most promising finding was a non-significant and therefore possibly chance finding of a reduction in the frequency of drinking across five studies. Within these studies, findings from the three targeting ethnic minority children registered a large reduction in frequency which came close to being statistically significant, though findings substantially differed between the studies.

Impacts of adding parental components to a child-focused intervention

These studies tested whether ‘added value’ was generated by supplementing a child-focused intervention with family/parent components. No such findings emerged which might not have been due to chance fluctuations.

No studies reported on volume of alcohol. Both prevalence and frequency of drinking were lower when family/parent components had been added, and remained lower when the analysis was confined to studies at lesser risk of producing biased results. Though the size of these differences were appreciable, with variable findings and so few studies (at most four) included in the analyses, the differences were not statistically significant, meaning that the possibility that these were chance results could not be excluded. The most promising findings were a relatively substantial but still non-significant reduction in prevalence across two studies judged at low risk of bias, and a near-significant reduction in frequency of drinking across three studies at low risk of bias.

Findings were similar when the analysis narrowed in on studies of interventions applied universally to all children regardless of their risk of drinking or drink-related problems. Frequency and prevalence measures favoured adding parental components, but not consistently enough to result in a significant finding across so few studies.

Only frequency of use could be analysed for selective and indicated studies, and then in only two studies of indicated interventions whose results amalgamated to virtually no difference from adding parental components.

Studies whose results could not be amalgamated with the others did not contradict these findings.

The authors’ conclusions

Overall this comprehensive systematic review with meta-analyses found little evidence that universal, selective, or indicated interventions with parents or families reduce adolescent drinking compared to no intervention, standard care, or a child-focused intervention alone. Some evidence suggests that under certain circumstances such interventions may be effective. However, in light of the number of analyses conducted, variation in effects, and the high risk of bias across the studies, the overall interpretation of outcomes indicates no effect. There seemed no clear differences between the interventions which did and did not work.

Findings on frequency of use suggested that interventions aimed at low-risk children can be counterproductive, while more targeted selective and indicated interventions aimed at higher risk children are more likely to reduce frequency. Among these higher risk children there is more scope to reduce frequency and arguably it is a more relevant measure than prevalence, since preventing use altogether may be less achievable than reducing drinking. However, this speculation is not supported by findings on volume of drinking.

Findings logo commentary These findings seem to almost entirely deflate what in the mid-2000s was a bubble of enthusiasm for parental programmes in the form of the Strengthening Families Program, leading to its being adapted for the UK. However, to reach a verdict on parental programmes in general, the analyses undertaken by the reviewers pooled results from different programmes as if these were the same intervention implemented in different circumstances. In fact, of course, they differ, and some may actually be effective, at least in certain circumstances. Arguably this is the more practically relevant issue, since commissioners and services do not mount a ‘parental programme in general’, but a particular one, and would want to choose one which has a good record, even if this is not enough to outweigh less effective programmes when their results are pooled.

To address this issue we can look at the studies which generated the review’s most promising findings. This closer look at the original studies shows that even the more promising results found by the review cannot be relied on as an indication of what might be achieved if the ‘best’ interventions were selected to be widely implemented, or drinking was assessed over a longer period, reinforcing the review’s ‘no reliable effect’ implications.

Where did the promising findings come from?

The only statistically significant results were for volume of drinking across all trials and across trials of universal interventions. Possibly this was due to the narrower range of studies which could be included in these analyses – just five across all trials compared to 12 for prevalence and eight for frequency of drinking. It seems possible that had all the studies which reported prevalence or frequency also reported volume, there would have been no significant effects, since logically volume is a composite of frequency and amount drunk on each occasion. Additional to these significant findings, there was also a near-significant finding indicating that frequency of drinking might be reduced by universal interventions.

Just three studies were responsible for these more encouraging findings – all from the same lead researcher, likely to have come from a highly self-selected sets of children and parents, and in two studies limited to mother-daughter pairs. In one case the reviewers seem to have mistaken a frequency measure for volume, in another the volume measure covered the last 30 days of the intervention, with no indication whether the effect lasted, and in another the finding on frequency was dependent on the follow-up period chosen for the review; in the context of negative findings across the whole follow-up period, the positive findings at this point can be seen as anomalous.

There was also a non-significant reduction in prevalence arising from a single study of a universal intervention, but again this was an artefact of the review’s choice of which follow-ups to focus on. Across the entire follow-up period there was no reduction in drinking attributable to the family/parenting intervention.

Among studies of selective or indicated interventions, the most promising finding was a non-significant reduction in the frequency of drinking, which came close to being statistically significant across the three trials targeting ethnic minority children. In both cases a single study was responsible, conducted among a very distinct group – drug and/or alcohol-using Mexican-American adolescents with links to ‘gangs’ – and the findings were at risk of bias due to the high number of participants who could not be followed up.

To unpack this summary of the key studies unfold Unfold supplementary text the supplementary text. It includes consideration of whether the Swedish Örebro intervention is an exception to the general ineffectiveness of parental interventions.

Thanks for their comments on this entry in draft to research author Conor Gilligan, Senior Lecturer in the Discipline of Health Behaviour Science at the University of Newcastle in Australia. Commentators bear no responsibility for the text including the interpretations and any remaining errors.

Last revised 20 May 2019. First uploaded 13 May 2019

Give us your feedback on the site (two-minute survey)
Open Effectiveness Bank home page
Add your name to the mailing list to be alerted to new studies and other site updates

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

REVIEW 2014 Interventions to reduce substance misuse among vulnerable young people

REVIEW 2012 Universal alcohol misuse prevention programmes for children and adolescents: Cochrane systematic reviews

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

STUDY 2014 The differential impact of a classroom-based, alcohol harm reduction intervention, on adolescents with different alcohol use experiences: a multi-level growth modelling analysis

STUDY 2012 Brief intervention for drug-abusing adolescents in a school setting: outcomes and mediating factors

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

STUDY 2010 Long-term effects of the Strong African American Families program on youths’ alcohol use

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

REVIEW 2015 Prevention of addictive behaviours