Although this is one of the largest samples of youths to have been examined for drinking behaviour in England, it was not designed to be representative of the population. In addition, in order to ensure truthful responses the survey asked a minimum on personal demographics and consequently deprivation was assigned utilising school locations rather than more sensitive individual measures. Such factors limit both the analyses possible within the study and the opportunity to extrapolate data to wider populations. Further, as a cross-sectional study our results do not address cause and effect. However, the design of the study has allowed detailed examinations of relationships between risky alcohol consumption, sources of alcohol, places of consumption and other individual behavioural and demographic characteristics.
In fact, levels of at least any alcohol consumption by study participants (here, 87.9%) were broadly consistent with national data that show a prevalence of any alcohol consumption in the last twelve months of 91% (aged 15–16) . However, rather than conducting detailed analysis of factors relating to any alcohol use, we have examined factors predicting risky drinking among youths that consume alcohol. Such analyses acknowledge that the vast majority of 15 to 16 year old youths in the UK drink alcohol (at least occasionally), that alcohol consumption in controlled environments (e.g. with family) may not always be detrimental to health [31, 46] and that identifying factors relating specifically to alcohol misuse (c.v. consumption) are a critical part of developing harm reducing interventions.
Results identify binge, frequent and public drinking as all being strongly related to amounts of spending money youths have available. Such information offers at least three possible points for intervention. Firstly interventions could aim to reduce money available to young people or advise parents on improving their monitoring of what youths spend money on. Currently, teenagers (aged 12–16) in the UK typically receive almost £10 pocket money a week from parents  while over a third (37%) of 14 to 15 year olds work in a regular paid job during school term time . However, public health considerations of how providing money to youths affects behaviour or of how parents may better control expenditure are poorly developed. Secondly, increasing the cost of alcohol may reduce access to alcohol and thus consumption. Unfortunately, in real terms affordability of alcohol has increased in the UK  and despite good evidence that increased costs can reduce alcohol consumption, particularly among young people, there appear to be no moves to increase alcohol taxes  or allow areas to manage alcohol prices locally through other means . Finally, organised social and sporting activities which are attractive to young people should be made more widely available as an alternative to getting drunk. Currently there is little international evidence that providing specifically alcohol-free diversionary activities will reduce youth alcohol consumption . However, our results (Table 3) at least support findings elsewhere that individuals involved in youth groups and sporting activities are less likely to exhibit risky drinking behaviours . Some interventions have been developed in the UK and other countries with activities such as sports, theatre and crafts provided to encourage youths away from anti-social behaviour, including substance use [53, 55, 56]. However, such programmes currently affect only limited numbers of individuals in deprived areas [55, 56] and our results (Table 3) suggest risky drinking behaviours occur across all deprivation quintiles. Consequently, rather than tightly focused interventions, further exploration is required of wider population approaches including school-based after-school activities and parental roles in encouraging participation in sports and other social ventures. Further, with membership of such groups potentially associated with greater binge drinking (Table 2), provision of better entertainment options for youths should incorporate safer drinking messages for participants.
Related to available income but independently associated with increased binge, frequent and public drinking is individuals buying their own alcohol. Those who purchase their own drink are nearly six times more likely to drink in public settings, three times more likely to drink frequently and twice as likely to usually binge (Table 3). However, all such alcohol purchasing should be controlled. It is illegal to sell alcohol to anyone under 18 in the UK and a range of voluntary identity schemes (e.g. Challenge 21) are available to ensure that those trying to buy alcohol are aged 18 or over . Local enforcement bodies (e.g. Trading Standards, police) can punish and even force closure of establishments that persistently sell alcohol to those underage. However, despite evidence that such interventions are effective , at least historically such powers have been used infrequently. Even with increased enforcement in recent years, over a fifth of targeted test purchases still result in a positive sale to a minor (England and Wales). Moreover, in our sample approximately a third of the individuals had bought alcohol for themselves (39.6% of those that drink). With greater investment in enforcing underage sales restrictions these proportions could be radically reduced. Further, partnership working between alcohol retailers, health services and enforcement sectors could develop interventions to help those identified trying to buy alcohol underage.
Not all alcohol consumed by youths is purchased by them. In this study many individuals used older siblings, friends and even strangers outside alcohol retail outlets to obtain alcohol. Obtaining alcohol from older siblings was related to binge, frequent and public drinking and again consideration should be given to informing parents of additional risks faced by younger siblings from alcohol provided by older brothers and sisters . Obtaining alcohol from adults outside shops was especially related to drinking in public settings. Not only does this allow youths to get drunk but it also requires young people to interact with adult strangers, potentially exposing children to the risk of sexual abuse. However, partnership working could allow adults who buy alcohol for underage youths to be identified and laws preventing such purchases enforced.
Finally, and in stark contrast to other forms of obtaining alcohol, youths whose parents buy them alcohol were less likely to binge and only half as likely to drink in public settings. Moreover, they were also significantly less likely to be in the most harmful group showing all three risk behaviours (Table 3). Such findings suggest that parental provision of alcohol, at least to youths who already consume, may reduce their immediate risk of hazardous and harmful consumption behaviours. These findings are consistent with other studies which identify drinking with parents as protective against higher alcohol consumption . Further, although early initiation into alcohol use appears to be related to later problem drinking, those initiated in a family environment are much less likely to become problem drinkers than those initiated outside the family . Models of family drinking in other (e.g. Mediterranean) countries, where alcohol is routinely consumed in the family environment but levels of binge drinking and anti-social behaviour associated with alcohol are lower than in the UK, are also at least consistent with a positive effects of parental alcohol provision. Such family consumption may help open up an early dialogue about alcohol between parents and children. Furthermore, it allows youths to experiment with alcohol in a family setting with positive parental role models rather than outside the family with pressure from peers to consume to excess. More research is needed on any positive effects of consuming alcohol in family environments. In the meantime however, it is essential that public health messages do not discourage parents from consuming modest amounts of alcohol with their children as such changes may actually increase drinking behaviours most damaging to youths' health.