What Influences 11-year-olds to Drink? Findings From the Millennium Cohort Study

Yvonne Kelly; Alice Goisis; Amanda Sacker; Noriko Cable; Richard G. Watt; Annie Britton


BMC Public Health. 2016;16(169) 

In This Article


Our results suggest that nearly 14 % of 11 year olds in the UK have had an alcoholic drink. The odds of drinking were greater when their friends drank compared to when their parents drank: boys and girls who reported having friends who drank were five times more likely to report drinking themselves compared to those who reported having friends who did not drink. Having a mother who drank heavily was associated with an 80 % increased odds of drinking, however, fathers' drinking was not independently associated with children's drinking. Our results suggest that 11 year olds' perceptions of risk, their expectancies towards alcohol and relationships with their families were independently related to the likelihood of drinking.

Distinct strengths of this work are that we used data from a large sample representative of 11 year olds in the UK; we simultaneously examined relationships with parents and friends drinking; and we were able to take into account rich contextual information about young people's understanding of the risk of drinking alcohol, their expectancies, positive and negative, towards alcohol and family relationships. On the other hand, there are several limitations to acknowledge, including that the analyses were cross sectional as information on cohort member and friends' drinking, perceptions of harm and expectancies around drinking are only available from one wave of data collection thus causal inference cannot be drawn; the data on cohort member and friends' drinking were developed for the MCS survey making it difficult to compare prevalence rates with other studies, although closed questions as used in this study have been shown to be valid markers of alcohol consumption in adolescents;[25] the data on cohort member and friends' drinking were reported by the cohort member and thus may be prone to under or over estimation with one prior contemporary study suggesting a lower prevalence of drinking among British 11 year olds,[10] although this may be due to different survey questions; we were not able to distinguish those who had just tried one or two drinks ever from cohort members who are regularly drinking; also there were no data available on the context of cohort member drinking and so it was not possible to assess the circumstances in which, or with whom, 11 year olds drank.

Prior work has charted the prevalence of drinking among 11 year olds in the UK[10] and elsewhere.[26,27] To our knowledge this is the first UK study in this young age group to attempt a detailed exploration of family and peer influences, along with the young person's views about alcohol on the likelihood of drinking. Moreover, most prior work has been set in the US[28] and it may be that associations vary across contexts.[9] We examined associations between parent and friends' drinking and family relationships at the very start of the adolescent period, whereas prior studies have looked at these associations among older adolescents. For instance, Cable and Sacker's examination of 16 year olds from the 1970 Birth cohort suggests that negative expectancies are not protective.[13] However, we might expect to see the same pattern of association as adolescence proceeds with peer influences and associated social norms having a more profound effect on alcohol use in later than early adolescence.[13,15–17]

A recent Cochrane review[29] concluded there was limited evidence that school/education based intervention programmes were effective, and where they did work the focus was more holistic, not solely on alcohol. In keeping with this we found markers of other risky behaviours, including smoking and antisocial behaviours to be strongly independently related to drinking at age 11. Clearly, there are opportunities to intervene and help shape choices around risky behaviours including drinking. Our findings support policies working at multiple levels that incorporate family and peer factors. For example: compared with mother's drinking, father's drinking was not as strongly related to drinking in their 11 year olds but this may be because fathers are more likely to drink in settings other than the home. Our observations that greater awareness of the harms from alcohol and negative expectancies are associated with reduced odds of 11 year olds drinking support strategies to empower young people to say no to alcohol. This is particularly important, as undoubtedly, peer influences become stronger in shaping young people's behaviours as adolescence proceeds.

Our study was not able to examine contexts around drinking occasions among 11 year olds – who do they drink with? Where, when and what do they drink? How do they acquire alcohol and what are the broader social norms around drinking? One study that compared young people's drinking in Italy and Finland showed that Italian youth were more likely to drink with meals under family supervision, whereas Finnish youth were more likely to drink in settings that led to drunkenness.[9] Being able to investigate context in more detail would help inform alcohol harm prevention strategies. Longitudinal studies looking at changes in expectancies towards alcohol and how these relate to changes in young people's behaviours including potential clustering with other risky behaviours are important areas for future study.