Prevalence of parental supply of alcohol to minors: a systematic review

Abstract Parental supply of alcohol to minors (i.e. those under the legal drinking age) is often perceived by parents as protective against harms from drinking, despite evidence linking it with adverse alcohol-related outcomes. This systematic review describes the prevalence of parental supply of alcohol, as reported in the international literature. The review was registered with PROSPERO (CRD42020218754). We searched seven online databases (Medline, Embase, PsycINFO, CINAHL, Scopus, Web of Science and Public Health Database) and grey literature from January 2011 to December 2022 and assessed the risk of bias with the JBI Critical Appraisal Checklist. Among 58 articles included in narrative synthesis from 29 unique datasets, there was substantial variation in the definition and measurement of parental supply of alcohol. Overall prevalence rates ranged from 7.0 to 60.0% for minor-report samples, and from 24.0 to 48.0% for parent-report samples. Data indicate that parental supply prevalence is generally proportionately higher for older minors or later-stage students, for girls, and has increased over time among minors who report drinking. Literature on the prevalence of parental supply of alcohol is robust in quantity but inconsistent in quality and reported prevalence. Greater consistency in defining and measuring parental supply is needed to better inform health promotion initiatives aimed at increasing parents’ awareness.


INTRODUCTION
Parents often assume that the provision of alcohol to underage youth, particularly under supervision and in moderate quantities (Jones et al., 2016), has a protective effect against future alcohol consumption (Jones, 2016).Previous studies have highlighted widespread support among parents for introducing alcohol in the home to teach their underaged child(ren) about responsible alcohol consumption (Jones et al., 2016;Roberts et al., 2010).However, these views are not supported by current evidence.Parental supply of alcohol to minors, including via sips, is associated with earlier alcohol initiation and subsequently alcohol-related harms in the long-term (Sharmin et al., 2017)-for example, studies of parental supply to minors show increased quantity and frequency of consumption, risky drinking behaviour and higher levels of alcohol use later in life (Ryan et al., 2010;Sharmin et al., 2017;Yap et al., 2017;Aiken et al., 2020).Studies also highlight that early sipping of alcohol permitted by parents can lead to a normalization of alcohol use in later adolescence, leading to increased frequency and quantity of alcohol consumption, and increased alcohol-related problems in later adolescence (Jones et al., 2016;Colder et al., 2019).
Alcohol use is the single leading cause of death and disability in those aged 15-24 years globally (Mokdad et al., 2016).Among young people, alcohol use is associated with considerable individual and societal impacts, including alcohol-related injury and assault (Quigley et al., 2019), high risk sexual behaviour (Stueve and O'donnell, 2005;Chan et al., 2016a, b), sustained neurocognitive effects (White and Swartzwelder, 2005;Zeigler et al., 2005;Lisdahl et al., 2013;Lees et al., 2020) and increased risk of suicide or premature death (Miller et al., 2007).Despite the potential harms related to alcohol use during early life, parents are one of the most common sources of alcohol for underage drinkers (Australian Institute of Health and Welfare, 2020).
Legal purchasing age policies, among other health promotion interventions, reduce alcohol access and associated harms (European Union Agency for Fundamental Rights, 2017).There is variation between the minimum ages set between countries, for example, 18 years in Australia and most of Europe (with further variation between 16 and 20 years between some European countries) and 21 years in the USA, and within some jurisdictions by alcohol volume percentage [e.g.where a higher age is set for distilled spirits (Callaghan et al., 2013)].However, alcohol use among minors (i.e.those below the minimum age set) is still common in many countries, suggesting widespread alcohol access through sources outside of legal purchase.The Youth Risk Behaviour Survey found that among US high school students, 29.0% reported currently drinking alcohol and 14.0% reported binge drinking in the past 30 days (i.e.drinking >4 or >5 alcoholic drinks on a single occasion for females and males, respectively) (Centers for Disease Control and Prevention, 2020).Similarly, in Australia 27.0% of those aged 12-17 reported drinking alcohol in the past month, and over one-third (38.0%) of those 16-17 year olds reported consuming five or more drinks on a single occasion in the past week (Guerin and White, 2020).
While some systematic reviews have focussed on associations between parental supply of alcohol and minors' outcomes, to our knowledge, this is the first systematic review of the prevalence of parental supply (Ryan et al., 2010;Sharmin et al., 2017).A clearer understanding of the variation within supply prevalence by study (e.g. reported by minors as opposed to parents) and sample characteristics (e.g.age, gender and country) may inform future health promotion intervention designs and initiatives that reduce parental supply and associated alcohol use by minors (Munn et al., 2014).This systematic review aimed to describe the prevalence of parental supply of alcohol to minors, as reported in the international literature.

METHODS
This systematic literature review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (Moher et al., 2010).The protocol was registered in the online database PROSPERO (registration number: CRD42020218754).

Eligibility criteria
Studies were eligible if they met the following inclusion criteria: (i) were full-text articles published in peer-reviewed journals, or research reports published by research agencies or government bodies that provided enough information on design, conduct and analysis for study quality to be assessed; (ii) included as participants children and/or adolescents (aged <18 years), or parents/formal guardians (hereafter We restricted eligibility to studies that were published after 2010, to reflect more recent behaviours and given the substantial changes in adolescent alcohol consumption in many high-income countries in recent decades (Vashishtha et al., 2020b).The original review protocol was amended to also include studies conducted in jurisdictions where the minimum purchasing age was above 18 years (e.g.South Korea, Sweden, the USA and Canada), and therefore including minors aged 18-20 years old, to increase coverage.

Search strategy
The

Selection of studies
After duplicate removal, retrieved articles were selected through two phases, conducted by two independent reviewers via an online reference management database (Covidence).First, titles and abstracts were checked for potential relevance.Then, full-text articles were retrieved and screened for eligibility.At each stage, screening was performed by one author (S.K.) in full, and a second author (A.B.) independently screened 10% of studies in duplicate.The agreement on inclusion/exclusion between the two reviewers was 94% (κ = 0.76) and 80% (κ = 0.59) for title/abstract screening and full-text screening, respectively, with any disagreement resolved by consensus and discussion.Contact with authors of primary studies was attempted where appropriate (e.g. to seek further information where eligibility for inclusion was unclear).

Data extraction and management
One author (S.K.) independently extracted the following key characteristics from included studies using a Microsoft Excel spreadsheet: citation details, country, year of data collection, study design, outcome definition and measurement, participants' characteristics (sample size, demographics and response rate) and prevalence of parental supply (%).Data on secondary outcomes, including frequency of alcohol supply by parents to minors, and volume of alcohol supply by parents to minors (converted to grams of ethanol), were also extracted.Our prospectively registered protocol also specified extracting the age of minor at first alcohol supply by parents, but this outcome is not reported here due to insufficient data in included studies.A second author (N.J.H.) reviewed a subset of the data extraction (100% of prevalence rates and 10% of all other data).

Critical appraisal
Included studies were independently assessed in duplicate for risk of bias by three authors [S.K. (100%), N.J.H. (65%) and J.B. (35%)], using the JBI Critical Appraisal Checklist for Prevalence Studies (Joanna Briggs Institute, 2017) suggested by a recent review (Ma et al., 2020).Any discrepancies in the duplicate assessment of checklist items were resolved through discussion between two authors and further consulted with a third author when necessary.The overall risk of bias for each study was calculated using the percentage of relevant items rated as including essential quality characteristics (i.e. the number of 'yes' items recorded on the checklist divided by the number of items, excluding items considered to be 'not applicable').A quality score of ≥70% was deemed to indicate a low risk of bias; scores between 50 and 69% and <50% indicated moderate and high risk of bias, respectively (Joanna Briggs Institute, 2014;Lu et al., 2021).

Synthesis of results
Our protocol specified that if studies were sufficiently similar, a random-effects meta-analysis would be conducted to examine the pooled prevalence of parental supply, including 95% confidence intervals.However, formal quantitative meta-analyses and pooled mean estimates were deemed inappropriate given the heterogeneity of supply measurement and reporting in the included articles.Hence, a narrative synthesis of data was undertaken instead.
Prevalence data is presented below by minor-report (Table 1) and parent-report (Table 2) separately, and further split into subgroups based on other key study and sample characteristics (i.e.age, gender, country and over time).Where comparable subgroup data were available from two or more datasets, the range of parental supply of alcohol prevalence across relevant studies is depicted; other summary statistics (e.g.weighted mean estimates) are not reported due to heterogeneity and the small number of studies reporting prevalence data in most subgroups.For consistency, all prevalence rates are reported here to one decimal place.

Study selection
After the removal of duplicates, 4216 titles and abstracts were screened and 167 articles were assessed for eligibility based on full text (refer to Supplementary Appendix A for a list of exclusion reasons).A total of 44 articles were considered eligible.Of the 44 articles, 15 were based on overlapping databases, resulting in 29 unique datasets that were included in the narrative synthesis.Based on searches for additional data collection years of national surveys, 14 additional articles meeting the review criteria were identified, resulting in 58 articles from a total of 29 datasets.No additional studies were deemed eligible through backward searching.Figure 1 summarizes the results of the search.

Definitions of parental supply
Nineteen out of 29 datasets (65.5%) asked participants about lifetime parental supply of alcohol, either explicitly (i.e.asked whether they had ever obtained alcohol from their parent or supplied alcohol to their child) or implicitly (i.e.asked participants whether they had received alcohol from their parent or supplied alcohol to their child).Ten of the 29 datasets (34.5%) asked participants whether they had received alcohol from their parents (i.e.minor-report), or asked parent participants whether they had supplied alcohol to their child (i.e.parent-report), over a past period of time (e.g.past 12 months or last 4 weeks).

Risk of bias
Using JBI's critical appraisal tool, 12 of 29 datasets (41.4%) were deemed as having low risk of bias, 9 out of 29 (31.0%) as moderate risk of bias and 8 out of 29 (27.6%) as high risk of bias.Studies with the lower risk of bias, and that were considered to provide the most robust estimates on the prevalence of parental supply, generally used a nationwide sampling frame or large cohort sample.These studies were conducted in Australia (Kelly et al., 2012;Chan et al., 2016a, b;Kelly et al., 2016;Chan et al., 2017;Australian Institute of Health and Welfare, 2020), England (Health and Social Care Information Centre, 2011, 2013, 2015, 2017, 2019) and the USA (Substance Abuse and Mental Health Services Administration, 2015, Substance2016, Substance2017, 2018, 2019, 2020, 2021;King et al., 2016;Vidourek et al., 2018).The studies with the higher risks of bias generally reported cross-sectional surveys with community based, and often smaller, convenience samples (refer to Supplementary Appendix B for details of the sampling methodology).A summary of the risk of bias of the 29 datasets is provided in Tables 1 and 2.

Overall prevalence of parental supply
Twenty-five out of 29 datasets (86.2%), including 51 of 58 studies (87.9%), reported prevalence rates based on minor-report.Overall prevalence rates varied widely across studies, ranging from 7.3 (Brunborg et al., 2019) to 60.1% (Aiken et al., 2020).Two distinct denominators were used to calculate prevalence rates: (i) all minors (i.e. both those who reported drinking and those who reported abstaining), ranging from 8.1 to 45.7%; and (ii) only minors who reported drinking, ranging from 10.8 to 60.1% (Table 1).Two studies (Friese and Grube, 2014;Brunborg et al., 2019) assessed parental supply of alcohol specifically in the context of parties, ranging from 7.3 to 9.0%.The range of prevalence rates, including by key study and sample characteristics, where comparable subgroup data is available, is shown in Figure 2.

Prevalence rates by age
Over a third of the datasets (37.9%) and nearly half (44.8%) of studies reported minor-report prevalence rates by age (Supplementary Appendix C).Studies that reported prevalence rates for all minors, all found prevalence rates higher for those who were older (indicated by either age in years or school grade) (Health and Social Care Information Centre, 2011, 2013, 2015, 2017;Stafström, 2014;Shaw et al., 2018); ranging between 32.0 and 38.0% for those who were 15 versus 4.0-8.0%for those who were 11 years old (Figure 2).The Australian Parental Supply of Alcohol Longitudinal Study (APSALS) (Aiken et al., 2017;Mattick et al., 2017;Mattick et al., 2018;Clare et al., 2019;Aiken et al., 2020;Boland et al., 2020;Clare et al., 2020;Najman et al., 2021) also found that prevalence rates increased with age from 9.3% in 2010-11 (mean age: 12.9 years) to 10.5% in 2013-14 (mean age: 15.8 years) for all minors who reported parents as their only alcohol source, and from 15.2% in 2010-11 to 45.7% in 2014-15 (mean age: 16.9 years) for those reporting any parental supply (i.e.including those additionally reporting supply via other sources).Prevalence rates could not be compared for other older age groups, as few data were available for individual age groups (as opposed to age ranges), and due to differences in denominators.
For only minors who reported drinking, prevalence rates varied.Compared with those who were younger, 8 out of 16 studies (50.0%) found lower rates in those who were older (Pilatti et al., 2013;Substance Abuse and Mental Health Services Administration, 2015, Substance2016, Substance2017, 2018, 2019, 2020, 2021), 6 studies (37.5%) found similar rates (White and Bariola, 2012;Asante et al., 2014;White and Williams, 2016;Health andSocial Care Information Centre, 2017, 2019;Guerin and White, 2020) and 2 studies (12.5%) found higher rates (Clark et al., 2013;Lam et al., 2017a, b).Prevalence rates ranged between 46.0 and 72.0% for 11 year olds, 22.0 and 72.0% for 12 year olds, 51.8 and 73.0% for 12 year olds, 52.8 and 70.0% for 14 year olds, and 58.1 and 72.0% for 15 year olds (Figure 2).Prevalence rates by gender A total of 11 datasets (37.9%), including 23 (39.7%) studies, included minor-reported prevalence rates by gender (Supplementary Appendix C).Prevalence rates ranged from 17.2 to 57.9% for males, and from 16.0 to 61.9% for females (Figure 2).Overall, the prevalence of parental supply was higher for female than male respondents in 15 out of 23 studies (65.2%) (Clark et al., 2013;Asante et al., 2014;Strandberg et al., 2014;Substance Abuse and Mental Health Services Administration, 2015, Substance2016, Substance2017, 2018, 2019, 2020, 2021;White and Williams, 2016;Health and Social Care Information Centre, 2017;Lam et al., 2017a, b;Guerin and White, 2020;Murphy et al., 2021).For studies reporting prevalence rates by gender for all minors, four out of eight studies (50.0%) found no significant differences in prevalence rates between male and female respondents (Danielsson et al., 2011;Health and Social Care Information Centre, 2011, 2013, 2015), two studies (25.0%) found slightly higher prevalence rates in females than males (Strandberg et al., 2014;Health and Social Care Information Centre, 2017) and one study found slightly higher prevalence rates in males than females (Murphy et al., 2021).Another study (Najman et al., 2021) reported parental supply separated by age and the individual parent who supplied alcohol (i.e.father or mother); boys reported higher rates of supply compared with girls when alcohol was received from the father.The opposite trend was observed when alcohol was received from the mother; girls reported higher prevalence rates than boys.For studies describing prevalence rates for only minors who reported drinking, of which one study (Health and Social Care Information Centre, 2017) also reported prevalence rates for all minors, 13 out of 16 studies (81.3%) found higher prevalence rates amongst females than males (Clark et al., 2013;Asante et al., 2014;Substance Abuse and Mental Health Services Administration, 2015, Substance2016, Substance2017, 2018, 2019, 2020, 2021;White and Williams, 2016;Health and Social Care Information Centre, 2017;Lam et al., 2017a, b;Guerin and White, 2020), two studies (12.5%) found higher prevalence rates amongst males than females (White and Bariola, 2012;Pilatti et al., 2013) and one study (6%) found no difference by gender (Health and Social Care Information Centre, 2019).

Prevalence rates over time
Four datasets (13.8%), including 22 studies (37.9%), reported data from repeated, cross-sectional national surveys (Health and Social Care Information Centre, 2011, 2013, 2015, 2017, 2019, 2022;Kelly et al., 2012;White and Bariola, 2012;Substance Abuse and Mental Health Services Administration, 2015, Substance2016, Substance2017, 2018, 2019, 2020, 2021;Chan et al., 2016a, b;Kelly et al., 2016;King et al., 2016;White and Williams, 2016;Chan et al., 2017;Vidourek et al., 2018;Australian Institute of Health and Welfare, 2020;Guerin and White, 2020).Secondary analyses of the Australian National Drug Strategy Household Surveys (NDSHS) found that parental supply prevalence rates among all minors declined between 2004 and 2013 (Kelly et al., 2016;Chan et al., 2017), from 16.2 to 8.1% for minors aged 12-17 years and from 21.2 to 11.8% for minors aged 14-17 years, respectively.In contrast, the SDDU in England showed relative stability in supply prevalence over the time period 2010-21, from 20.0 to 23.0%, respectively, for all minors (Health and Social Care Information Centre, 2011, 2013, 2015, 2017, 2019, 2022) (Table 1).Studies that reported prevalence rates using a denominator of only minors who reported drinking, at variously defined time periods, all found that the prevalence of supply increased over time.In Australia, primary reports of the NDSHS also showed that prevalence rates of parental supply as a usual source of alcohol increased from 25.0% in 2010 to 41.6% in 2019 (Australian Institute of Health and Welfare, 2020).Similarly, the Australian Secondary Students' Alcohol and Drug Survey (ASSAD) found an increase in supply prevalence among current drinkers over time (White and Bariola, 2012;White and Williams, 2016;Guerin and White, 2020); from 32.9% in 2011 to 43.0% in 2017.In the USA, the National Survey on Drug Use and Health (Substance Abuse and Mental Health Services Administration, 2015, Substance2016, Substance2017, 2018, 2019, 2020, 2021) also found an increasing trend in parental supply of alcohol to minors who reported drinking from 7.9% in 2013 to 14.5% in 2020 (Supplementary Appendix C).

Parent-report of parental supply of alcohol
Six datasets (20.7%), including nine studies (15.5%), reported prevalence rates based on parent-report, of which two datasets reported rates based on both minorand parent-report (Table 2).All studies calculated prevalence rates using the total parent population as denominator.All studies were conducted in Australia, and one (Gilligan et al., 2014a) also included Canadian as well as Australian participants.The overall parentreport prevalence rate of parental supply to minors ranged from 24.4 (Wadolowski et al., 2016) to 48.0% (Jongenelis et al., 2018).There were three crosssectional studies that reported prevalence rates by age (Ward and Snow, 2011;Jongenelis et al., 2018;Shaw et al., 2018); all studies found higher prevalence rates with increased age.Similarly, data from the APSALS cohort (Aiken et al., 2017) showed increased parental supply of alcohol with increased age over time, both in sips and full drinks (Supplementary Appendix C).

Frequency of parental supply
This outcome was reported for two datasets, including three studies (5.2%; Supplementary Appendix D).In the APSALS cohort (Mattick et al., 2017;Clare et al., 2019), supply frequency increased with age, particularly for monthly supply, from 24.1% in 2010-11 (mean age: 12.9 years) to 35.5% in 2014-15 (mean age: 16.9 years).Another Australian study (Lam et al., 2020) reported 'slightly' more frequent supply to minors for parties where alcohol consumption would be consumed without direct parental supervision, as opposed to those where parents would be present.

DISCUSSION
The aim of the current systematic review was to describe the prevalence of parental supply of alcohol to minors as reported in the international literature.The findings suggest there is a robust interest in research concerning the prevalence of parental supply of alcohol to minors, with 58 studies and 29 unique datasets reporting parental supply.The studies included in this review reported prevalence rates ranging from 7.3 to 60.1% in minor-report, and from 24.4 to 48.0% in parent-report samples.Due to heterogeneity between studies, no overall prevalence rates for the supply of alcohol from parent to child could be estimated for both minor-and parent-report.
Although some reviews have found a decrease in the prevalence of parental supply to minors over time (Kelly et al., 2016), our findings suggest the supply of alcohol from parent to child is still high and has in some cases increased over time.In particular, our review found increasing trends for those studies that reported parental supply for only minors who reported drinking which is cause for concern, and provides an impetus for governments to act.In contrast, those studies reporting on all minors either found a stable or decreasing trend.However, studies that found a decreasing trend in parental supply of alcohol over time [e.g.(Kelly et al., 2012;Kelly et al., 2016;Chan et al., 2017)], typically include the increasing population of young people who do not drink, and inherently have no source of alcohol supply (Vashishtha et al., 2020a).This decreasing trend of parental supply also seemed to be the case for those reporting on lifetime supply of alcohol.The review also found higher overall prevalence rates in parental supply of alcohol to females compared with males.It has been suggested this is because females are seen as more responsible than males (Strandberg et al., 2014).However, evidence shows females are more likely to report negative consequences related to alcohol use, compared with males, such as being robbed and having unprotected sex (Strandberg et al., 2014).Particular attention is thus required to reduce supply amongst minors who drink, who remain at increased risk of alcohol-related harms, and in particular, females.
It has recently been hypothesized that alcoholspecific parenting factors (e.g.alcohol-related parental approval and communication, in addition to supply), rather than general parenting factors, could be closely related to changes in minors drinking over time (Vashishtha et al., 2022).Addressing these factors may enhance health promotion interventions targeting supply behaviours specifically.Particularly as parents remain the most common source of alcohol supply to minors who consume alcohol in some jurisdictions (Australian Institute of Health and Welfare, 2020; Health and Social Care Information Centre, 2022), and given the established link between parental supply of alcohol and earlier initiation of alcohol (Sharmin et al., 2017), it is plausible that the implementation of health promotion interventions and policies focussed on reducing and preventing parental supply could further decrease rates of underage drinking.Consistent with this proposition, evaluations of mass media education campaigns targeting parental supply of alcohol in the state of Western Australia have been promising.These campaigns focussed on raising awareness amongst parents of alcohol's neurocognitive effects for adolescents, and reinforcing national alcohol guidelines that discourage alcohol consumption by minors (Johnston et al., 2018).Early evaluations have found that these campaigns are likely to influence parents' discussion of alcohol-related issues with their child(ren), although this did not always result in a reduction of parental supply (Johnston et al., 2018).However, an additional study, recently conducted in Western Australia, found a decline in parental supply of alcohol over the time period 2013-19, corresponding to the implementation of parent-targeted state-wide mass media campaigns (Booth et al., 2023).Importantly, parental supply campaigns may rely on prevalence data to address parents' perceptions of social norms (i.e. that supply of alcohol is practised by many or most other parents) and reinforce widespread non-supply among similar parents.For example, recent Alcohol.Think Again messaging in the state of Western Australia highlighted that '2 in 3 parents choose not to provide alcohol and it's reducing alcohol-related harm ' (Alcohol. Think Again, 2022).
The wide range in prevalence rates is likely related to several issues.First, variability across measured outcomes between studies is most likely caused by inconsistency in defining parental supply of alcohol.For example, one study asked minors who their usual source of alcohol is and found a prevalence rate of 60.1% (Clark et al., 2013), whereas another study found a prevalence of 33.8%, when minors were asked who purchased them alcohol the last time they consumed it in the previous 12 months (Rowland et al., 2014).Second, variability was observed in the number of response options (i.e.numerator) and analysis of responses.The APSALS study, for example, asked minors to report receiving alcohol from their mother and/or father in the past 12 months, for which the prevalence was found to be 16.1% for those reporting multiple sources including parents (Aiken et al., 2020), and 9.0% for those reporting only parents (Boland et al., 2020).The APSALS study also assessed parental supply of sips versus whole drinks, for which the prevalence was found to be 14.6% for those receiving sips and 1.5% for those receiving whole drinks at age 12.9 years, but 3.9 and 64.9% at age 18.8 years, respectively.Although evidence suggests that minors who receive sips from parents are less likely to report subsequent binge drinking and alcohol-related harms, compared with those receiving whole drinks from parents (Aiken et al., 2020), those receiving sips from parents are still at higher risk compared with those not receiving any alcohol from parents or others.Other reported differences in numerator between studies included the use of Likert-type scales or binary options (e.g.yes/ no).Third, variability occurred when studies used a different denominator; all minors versus only minors who reported drinking.In addition, minor-and parent-report prevalence rates of alcohol use are often inconsistent, which may reflect the latter being more susceptible to social desirability reporting biases (Kypri et al., 2005).These variabilities, coupled with earlier calls (Jones, 2016), further highlight the need for valid and reliable tools to measure parental supply of alcohol, as well as a consensus on definitions used, to enable comparability between original studies and future data collection in population surveys.It is recommended that future studies report parental supply by both all minors and those minors reporting drinking, quantify the amount of alcohol supplied by parents (e.g.sips versus whole drinks) and minimize recall bias by reporting recent or current alcohol supply, rather than ever or lifetime supply.In addition, we recommend that future studies collect data on minors' age at which alcohol is or was first supplied by parents.Our findings indicate a lack of current studies reporting this data.However, such nuanced information could be valuable for decision-makers and health promotion practitioners who design, implement and evaluate prevention and early intervention initiatives, for both minors and parents, to help ensure that the right age groups are targeted early enough, to reduce unnecessary costs and resources.
It is also important to note that nearly all studies included in the review were conducted in highincome Western countries, with nearly half of them from Australia and New Zealand alone (total 45.0%), and only one study from the South America region (i.e.Argentina) (Pilatti et al., 2013) and two studies from Asia (i.e.Thailand and South Korea) (Asante et al., 2014;Prasartpornsirichoke et al., 2022).As only English-language studies were included in the review, it is possible that additional studies from these regions were not identified in our searches.However, a recent systematic review investigating risk factors for drinking among young people in Thailand also concluded there was a dearth of studies on parental supply in that country, resulting in a crucial gap in current evidence (Luecha et al., 2020).This highlights the need for comprehensive prevalence data on parental alcohol supply from a wider range of countries to inform health promotion efforts, particularly given projected increases in alcohol consumption in middle-income countries over the next decade (Manthey et al., 2019).Future research could explore the factors shaping the prevalence of parental supply across countries, including the impact of different jurisdictional purchasing age policies, for example, which may provide further insight into how parental supply might be reduced.

Strengths and limitations
This systematic review provides an overview of current estimates of parental supply of alcohol and a comprehensive description of the characteristics of individual studies.To the best of our knowledge, this was the first systematic review that describes the prevalence of parental supply of alcohol to minors, as reported in the international literature.Over 70.0% of included studies were considered as having low risk or moderate risk of bias, indicating that a large proportion of reported rates are likely to represent valid prevalence estimates.Furthermore, our inclusion/exclusion criteria (e.g.minimum sample size) were applied to reduce the potential impact of reporting biases, and our searches were based on a wide variety of databases plus the grey literature, although it is acknowledged that additional studies may have been missed.
However, some limitations should be noted.Given the diversity of geographical and epidemiological settings, as well as the wide range of definitions and measurements used to measure parental supply of alcohol, substantial heterogeneity was observed in prevalence estimates between studies.Accordingly, pooled quantitative summaries were deemed inappropriate, which may have reduced our ability to synthesize the available data and to assess the weighted contribution of prevalence estimates (i.e. based on study sample size).In the absence of formal meta-analysis and the generation of funnel plots, publication bias could also not be assessed.

CONCLUSION
Raising parents' awareness of the harms associated with providing alcohol to minors, as to reduce parental supply of alcohol and the risk of alcohol consumption among youth worldwide, should be a priority.Future research should address the following issues: (i) utilize a universal definition of parental supply to be able to compare studies on parental supply of alcohol; (ii) use standardized measures to measure the supply of alcohol; and (iii) investigate factors associated with reporting in minors-versus parent-report of parental supply of alcohol.
A.B., C.M. and J.B. are associated, was one of 22 organizations that funded jurisdictional data collection for the article included in this review by Guerin and White (Guerin and White, 2020).

ETHICAL APPROVAL
This study was deemed exempt from ethical approval by the Flinders University Human Research Ethics Committee.

Fig. 1 :
Fig. 1: PRISMA diagram of database search and record screening.

Fig. 2 :
Fig.2: Range of parental supply of alcohol prevalence reported in included studies by key study and sample characteristics.Range is presented here where prevalence data is available from comparable subgroups from multiple datasets or independent studies (i.e.subgroup categories based on a single study, or multiple reports from repeated waves of the same cross-sectional survey, are not shown).
OR suppl* OR offer OR furnish OR source* OR rule OR allow* OR permi* OR agree* OR buy* OR approv* OR host*)].Search strategies were adapted to each database by combining subject keywords and related database-specific terms with the assistance of an experienced health research librarian.
following search terms were used in the Scopus online database: [alcohol* OR drinking AND (child* OR adolescen* OR youth* OR underage* OR minor OR teen* OR juvenil* OR pubescen* OR schoolchild OR youngster OR offspring OR kid* OR puber*) AND (parent* OR mother* OR father* OR guardian* OR custodian*)] AND ALL [(parent* OR mother* OR father* OR guardian* OR custodian*) W/6 (provi*

Table 1
Summary of included studies of parental supply by dataset,

as reported by minors Dataset ID (name) Article authors Study design; study name Setting Sample size (data collection year) Adolescent age (mean; range (years)) Response rate (%) Numerator; response options Denominator Overall prevalence (%) Prevalence subgroups a Quality appraisal: risk of bias
-6 S. van der Kruk et al.

Table 1 .
ContinuedPrevalence of parental supply of alcohol to minors 7

Table 1 .
ContinuedPrevalence of parental supply of alcohol to minors

Table 1 .
Continued 10 S. van der Kruk et al.

Table 1 .
ContinuedPrevalence of parental supply of alcohol to minors

Table 1 .
Continued a Some studies did not report an overall prevalence rate of parental supply, but reported prevalence rates by subgroups, or reported this in addition, which can be found in Supplementary Appendix D. b Response rate represents proportion of established cohort participants providing data at relevant timepoint.c Prevalence rate was calculated by authors based on the number of cases and sample size or subgroup proportions reported in article.d Response rate was reported for the total survey population (all ages involved), not only adolescents.e Prevalence calculations were based on a subset of this full study sample.f Response rate was reported for those aged 12-17 years and not 12-20 years.g Reported sample size rounded to the nearest interval of 10 respondents, aged 12-20 years.h Data were provided by the author of included article.-= data item not reported in article.n/r = sample size not reported for adolescent sample (14-17 years) specifically.NSDUH = National Survey on Drug Use and Health; YAARS = Young Australians' Alcohol Reporting System; SDUSAP = Student Drug Use Survey in the Atlantic Provinces.Table 1.Continued 14 S. van der Kruk et al.

Table 2
Summary of included studies of parental supply by dataset, as reported by parents a Some studies did not report an overall prevalence rate of parental supply, but reported prevalence rates by subgroups, or reported this in addition, which can be found in Supplementary Appendix D. b Response rate represents proportion of established cohort participants providing data at relevant timepoint.c Prevalence rate was calculated by authors based on the number of cases and sample size reported in article or subgroup proportions reported in article.d Data were provided by the author of included article.-= data item not reported in article.

Table 2 .
Continuedof parental supply of alcohol