Scope and definitions

The scope of this review was to present a summary of selected relevant literature that we have encountered through our work on childhood alcohol use in the last decade. We did not apply systematic review methods, nor was the review intended to be exhaustive, but it summarises important findings related to alcohol use and its consequences in a neglected age group and raises important questions for further research. We focused on childhood use of alcohol and have therefore not included prenatal exposure or unintentional intake. Ages and stages of childhood include a myriad of definitions, with inconsistent age ranges divided into infancy, early, middle, and late childhood, preteens, adolescence and youth. In this review, we have adhered to the definitions applied by the WHO-UNICEF-Lancet commission ‘A future for the world’s children?’,11 where ‘children’ were defined to be 0–18 years, and ‘adolescents’ were defined to be 10–18 years. We applied the term ‘younger children’ to those aged 0–10 years (see table 1). However, we acknowledge that different studies use varying and overlapping age categories that may not fit these definitions. Similarly, there is no consistent cut-off for ‘early initiation’ or ‘early use’ of alcohol. Therefore, we have not applied a definition for what is considered ‘early’ but adhered to what was used by the respective studies.

The epidemiology of childhood alcohol use

Worldwide, there are large recurring cross-sectional studies in school-going children that monitor the prevalence of alcohol and substance use, time trends and associated factors among adolescents. These include the Global School-Based Student Health Survey (GSHS),12 Health Behaviour in School-Aged Children (HBSC),13 and others. These large surveys are important, but the inconsistency in definitions and reporting of questions used to measure alcohol use complicates comparisons of the prevalence estimates they yield. The GSHS has defined a drink to be ‘a drink other than a few sips’12 and the HBSC has defined it as ‘more than a small amount’,13 without further specification. In some studies, children were asked whether they have had a drink or not without specifying a cut-off,14 and others have focused specifically on the sips and tastes of alcohol.15 Moreover, the recall time frame varies between weekly, monthly, yearly and lifetime use. The inconsistent use of definitions hampers the comparability of studies, and the opportunity for undertaking meaningful systematic reviews. While some systematic reviews have investigated the epidemiology among specific subgroups, such as street-connected children,16 comparing youth within and outside foster care17 and migrant adolescents in Europe compared with native peers,18 a consistent challenge has been the heterogeneity imposed by inconsistent definitions and measurements, making it difficult to execute and interpret meta-analyses. To improve comparability between studies, a consensus is needed for terminologies and measurements of alcohol use. To achieve this, one can explore existing tools, such as the publicly available GSHS questionnaire,12 or the shorter Car-Relax-Alone-Forget-Family and Friends-Trouble (CRAFFT) screening tool,19 as well as their relevance and validity across different contexts and age-groups.

In addition to the noted epidemiological challenges, there is a need for age-specific clinical screening methods and cut-offs. Donovan and colleagues have been important contributors to research on alcohol use involving younger children, termed ‘really underage drinkers’, aged 12 and below.10 15 As a response to a lack of age-appropriate definitions and screening instruments for children, Donovan used estimated blood alcohol concentration to suggest age-specific and gender-specific cut-offs for ‘binge drinking’; ≥3 drinks for 9–13 years old boys, ≥4 for 14–15 years old boys, ≥5 for 16–17 years old boys and ≥3 for 9–17 years old girls.20 Such attempts at establishing clinically meaningful cut-offs are welcome, and more research is needed to inform valid screening methods and establish thresholds for levels of concern for younger children, although any alcohol use in this age can be argued to be of concern.

Determinants of early alcohol use

Initiation of alcohol use is a complex process with multiple determinants that vary over time and by geographical location. A full exploration of these factors and their interplay is beyond the scope of this paper, but identified key determinants include family environment, genetic factors, mental stressors, sociocultural factors and policy. These often occur in combination. A summary of the determinants of early alcohol use discussed in this review can be found in table 3.

Family influence is an important determinant for early alcohol initiation, especially by younger children.29 A systematic review of longitudinal studies found a consistent positive association between parental and child drinking, but concluded that causal inferences could not yet be established.30 While genetics can play a role, the variance in drinking initiation and drinking patterns attributable to genetic factors has been found consistently lower in childhood and early adolescence, and increasing into adulthood.31 Social pathways have shown more promising results in explaining the link between parental and child drinking, and important concerns have been raised related to the increased exposure to parental drinking during the Covid-19 pandemic, where parents and children have spent more time at home.32 The link is sometimes indirect, through mechanisms of social learning, but also direct, by provision of alcohol in a family setting. One study investigating GSHS data from 45 LMIC found that familial alcohol supply was associated with earlier onset of alcohol intake and a higher frequency of drinking.23 Furthermore, a 2011 Australian secondary school survey found that 33% of last-week-drinkers received their latest drink from their parents and 23% from friends; only 10% took the alcohol from home or purchased it themselves.33 Studies have shown that parents' own alcohol use is associated with early childhood drinking as well as an increased chance of experiencing negaitve alcohol-related consequences. A study on 7000 European pupils aged 12–14 years, found a strong relationship between parents’ permissiveness and adolescent alcohol use.34 Perceived parent approval of sipping, parental drinking status and frequency, and child attitude toward sipping was associated with sipping alcohol by 8–10-year-olds in the USA.35 In a birth-cohort from England, almost 7000 children were followed up at age 10, finding that maternal drinking in early childhood was the most important influence on child alcohol-use in the past 6 months, while paternal drinking was associated with a lower risk for child drinking (possibly due to residual confounding).26 A study from Burkina Faso, Ghana, Uganda and Malawi found a relationship between living with someone before age 10 who had an alcohol use disorder and self-reported drunkenness in adolescence.36 Adding to the influence of learning from a family member that drinks, one also has to keep in mind that parental drinking is related to less family support, reduced parental monitoring and increased risk of experiencing mental stressors, such as domestic violence3 and neglect.17 37

Harmful alcohol use is closely related to mental stressors and disorders,3 including in younger age groups. This can be related to coping with hardship, or part of a symptom picture, with for example reduced impulse control or conduct issues. In the aforementioned English birth cohort, drinking at age 10 had a weaker association with conduct problems compared with depression, but the relationship was stronger in adjusted analysis.26 In a small study from Uganda (n=148), 10 out of the 11 children aged 5–8 years diagnosed with harmful alcohol use and dependence had psychiatric comorbidities, including conduct disorder, suicidality, panic disorder and depression.24 In a study from the USA, investigating suicidal ideation among more than a thousand 8-year-olds at risk for or victims of maltreatment, 21% had used substances, including alcohol. Further, children that used substances were four times more likely to report suicidal ideation, and this association was stronger among maltreated children.38 However, some studies that have compared community samples with subgroups that are vulnerable for mental stressors show inconsistent results. One example is a systematic review that compared non-foster care youth and current and former foster care youth aged 12 years and older.17 While experiences of mental stressors, including neglect and post-traumatic stress disorder, were among the identified predictors of use, also among foster youth, the effect of being in foster care was not uniform. Overall, foster care youth initiated alcohol use earlier, and had higher lifetime use, but past year use was comparable between foster care youth and non-foster care peers, where youth outside foster care sometimes reported higher intake.17 Moreover, among migrant adolescents aged 11–29 years in Europe, a systematic review found consistently lower rates of substance use among migrants compared with native European peers. This was explained by cultural factors and religious beliefs, as well as the ‘migration paradox’, where recent immigrants tend to do better than their native peers in health, education and other outcomes, despite poorer socioeconomic conditions, exposure to discrimination and other stressors.18 This demonstrates how patterns of substance use and risk factors, even in younger age groups, are complex and multifaceted.

In addition to individual-level explanatory factors, we must take into consideration environmental and anthropological aspects. In some contexts, traditional forms of home brewing has made alcoholic brews affordable and available also to young children.28 37 Family location has been reported to be part of the social environments that influence alcohol consumption among adolescents. One study compared adolescents aged 12–17 years from urban and rural areas in Australia, finding that adolescents from rural areas were at higher odds of receiving their first alcohol from parents, receiving their current alcohol of use from parents, and drinking at home.39 Additionally, children from rural areas were more likely to engage in harmful drinking patterns, with heavy drinking and weekly drinking, and their parents were also more likely to have harmful drinking patterns.39 This does not mean that rural residence always predicts more harmful drinking patterns, but that where you live matters, and drinking patterns vary between and within countries depending on sociocultural and economic factors. Further, cultural norms, social influences from peers and social media intersect to create an environment around alcohol use during mid-adolescence and the early years, driving consumption at this stage. A qualitative exploration of sociocultural influences of adolescent alcohol use found that alcohol consumption was associated with being cool, mature and popular, while enabling escape from reality and boosting confidence and enjoyment.40 Positive expectancies, alongside opportunity, contributed to motivating initiation, but social influences were paramount, with participants describing a need to ‘fit in’ with friends to avoid social exclusion. Such influences positioned drinking at parties as a normative social practice, providing opportunities for social learning and the strengthening of peer norms. Social media presented young people with positive alcohol-associated depictions of social status, enjoyment and maturity. This intersection of influences and norms generated a pressurised environment and a sense of unease around resisting pressures, which could elicit stigmatising insults.40

Interventions, policy gaps and opportunities

Interventions for childhood alcohol use can target different stages of drinking, namely: (1) initiation of alcohol use, (2) problematic drinking and (3) chronic or clinical alcohol use disorders. A large proportion of the research on interventions for harmful alcohol use among adults focus on clinical alcohol use disorders, requiring treatment and rehabilitation within health facilities. However, the complexity of childhood alcohol use demands multidisciplinary interventions, and it is imperative to provide systemic interventions to prevent alcohol use at the primary level, i.e., the initiation stage. To encompass the needed focus on parenting, social support, coping skills, and emotional functioning family-based and school-based intervention and prevention programmes have proven promising.41 In an overview of systematic reviews on interventions for adolescent substance use, eight systematic reviews on alcohol use were included.41 While the findings from the interventions had mixed results, they showed that school-based interventions and prevention efforts were overall associated with a reduction of alcohol consumption. Family-based interventions, focusing on parenting and nurturing behaviours, had small but long-lasting effects.41 While these studies focused mostly on older adolescents, positive results have been found regarding the feasibility and acceptability of a programme for screening, brief intervention, and referral to treatment of substance use among youth in a school-setting as early as in sixth grade.42 In settings where access to treatment for alcohol use disorders is scarce, innovative interventions are needed. One example is mobile facilities for inpatient treatment and psychoeducation for adults which have shown promising results in Uganda, demonstrating the need for accessible treatment options close to the community.43 Age-appropriate and context-appropriate interventions for preventive and treatment purposes for younger ages are needed, both integrated within child mental health services and within children’s contexts. Rigorous research that identifies effective interventions and their implementation strategies is necessary to enable governments to counter early childhood alcohol use.

Beyond the immediate environment of family and school, prevention of access to alcohol by children needs attention on a wider societal level. While narcotic substances are illegal in most countries, political priority and regulations are necessary to protect children from the harm related to the alcohol available to them in communities. The WHO has defined the most cost-effective strategies to prevent harmful alcohol use across country income levels, termed ‘best buys3’ (see table 4). These include regulating alcohol availability by implementing and enforcing a minimum legal drinking age and restricting hours, days and places of sale, restricting and enforcing bans on alcohol marketing, and regulating prices and taxes.3 Among countries reporting to the 2018 WHO Global Status Report on Alcohol and Health, 93% had implemented a minimum age for on-premises alcohol purchase, where 18 years was the most common, while some reported limits as low as 13, 14 and 15 years, and some LMICs had no formal age limit.3 Regulation of days for sales and alcohol outlet density, however, was reported by less than a third of reporting countries. Relevant targeted restrictions on public drinking included a total ban on intake in educational buildings (60%) and sporting events (30%).3 Children and youth are especially sensitive to marketing, and the need to regulate commercial marketing of harmful products to children has been emphasised in the WHO-UNICEF-Lancet commission ‘A future for the world’s children?’, where social media and the internet were considered increasingly important platforms.11 Yet, social media and internet regulations were the least common form of market and advertising regulation in the 2018 WHO Global Status Report on Alcohol and Health, and LMICs were most commonly reporting to have no restrictions across media platforms.3 Overall, the report revealed large disparities across the globe relating to the political priority and the available resources for implementing effective measures to prevent early alcohol exposure. Table 4 summarises key opportunities and characteristics of needed interventions, as well as established recommendations for relevant public health policy.