In recent years increasing levels of hazardous and harmful alcohol use, along with their negative health consequences, have received international attention as issues requiring immediate action. Thus, the World Health Organization (WHO) has identified excessive alcohol use as one of the most important risks to health 1 and the World Health Assembly (Resolution WHA58.26) has requested Member States to implement effective interventions to tackle alcohol misuse 23. Within the WHO European Region, where per capita alcohol consumption and related disease burden are double global averages 45, two successive alcohol action plans (covering 1992 to 2005) have been followed by a Regional Framework providing strategic guidance to member states on the development and implementation of alcohol policy 6. Further strategic support in the European Union (EU) has been provided through the EU strategy to assist member states in reducing alcohol-related harm 7. In both, particular attention is focused on the increasing levels of alcohol intoxication and the disproportionate burden of alcohol-related harms in European youth.

Among schoolchildren (aged 15–16) in many European countries, hazardous and harmful drinking patterns have escalated with increases in both binge drinking (here drinking five or more alcoholic drinks in one drinking occasion) (e.g. UK, Norway, Czech Republic) and drunkenness (having ever experienced more than 20 drunken episodes) (e.g. Ireland, Estonia, Slovenia) 8. In the long term continued heavy alcohol consumption, including frequent bingeing, is linked to pathologies including liver disease, cancers and cardiovascular disease 91011. However, binge drinking by youths is also associated with acute problems including alcohol-related violence and other anti-social behaviours 1213. Furthermore, it contributes to unprotected sex, drug consumption and educational problems including truancy 131415.

In the UK, levels of both alcohol-related health burdens (e.g. liver disease) and crime are rising 1617. Approximately 7.1 million residents in England are now thought to be hazardous or harmful users of alcohol 18 with around 1.1 million being alcohol dependent 19. Further, UK youth now have among the highest levels of alcohol consumption and binge drinking in the European Union (e.g. prevalence of binge drinking in the past 30 days among girls aged 15–16: UK 56%, Belgium 44%, France 23%) 8, and experience drunkenness at an earlier age than most of their European counterparts 20. Associated problems including anti-social behaviour 13 and unintended pregnancy 21 are also high 2223. Nationally, alcohol-related problems are estimated to cost approximately £20 billion per annum 24.

The reasons why alcohol creates a disproportionate health and social burden in some cultures (e.g. UK) are likely to be multi-faceted. While debate continues over whether any alcohol consumption by youths can be beneficial 25, studies suggest that drinking behaviours early in life can not only be of immediate detriment to health but may also relate to adult alcohol consumption patterns and their related health consequences 26272829. In particular, binge drinking and unsupervised underage drinking in public places can result in negative health consequences (e.g. accidents) and are strongly linked to anti-social behaviour 3031. However, where alcohol is typically integrated into family life (e.g. in Mediterranean countries such as France and Italy) 25, children display lower levels of binge drinking and alcohol-related anti-social behaviour than their Northern counterparts (e.g. UK and Scandinavian countries) 8. The consequences of frequent drinking by youths are less clear yet several studies have associated frequent youth drinking with increased health risks including involvement in physical fights 3132.

Previous research has identified a range of factors associated with risky drinking behaviours in youths. These include those related to: the individual (e.g. sensation seeking, anxiety, positive alcohol expectancies 33), their relationships (e.g. parental and peer drinking behaviour 3435), and the drinking context (e.g. cultural norms, alcohol availability 35). Here however, we use a school-based (aged 15–16) sample of 10,271 children and focus specifically on those that drink (n = 9,029). We examine relationships between risk-related drinking behaviours and alcohol consumers' demographics, personal economics and access to alcohol and leisure activities. Finally, we explore how an understanding of such relationships can be used to reduce alcohol-related harm.

Within England, the North West Region (population 6.8 million) 36 has the highest number and greatest prevalence of hazardous/harmful drinkers (1.25 million) of all English Regions 19. Previous work within the Region has already identified both substantial consumption of alcohol by young people and frequent illegal alcohol sales to those under 18 (UK minimum age of purchase is 18) 373839. To provide more intelligence on both issues, Trading Standards North West (the organisation responsible for enforcing legislation and regulations governing the sale of goods and services, including alcohol, in the North West of England) undertook a survey of 14 to 17 year old schoolchildren in the North West Region. Questions were anonymous and explored: basic demographics of respondents; drinking frequencies (over the past six months), types and quantities of alcohol consumed (drinks were counted in the number of bottles, cans, glasses and pints consumed to accommodate existing definitions of youth bingeing 8); sources of alcohol; income available to youths; and personal views on alcohol consumption. The questionnaire consisted almost entirely of closed questions with the exception of one requesting suggestions for additional facilities for young people. The questionnaire was made freely available though local Trading Standards offices (n = 22 covering the North West Region of England) to schools located within their administrative areas. In schools agreeing to participate, questionnaires were distributed by Trading Standards Officers to school staff (teachers and assistants) who then delivered the questionnaire to students within normal school lessons. Research methods and materials complied with the Declaration of Helsinki. Compliance levels were not recorded in each class as the sample was not intended to be representative but was opportunistic, with analyses focusing on relationships between variables recorded by individual participants. Further, while demographics such as ethnicity were included, only broad categories were used in order to obtain meaningful sample sizes in each group.

From 147 North West secondary schools (in 21 local authorities) that participated in the survey, a total of 12,883 questionnaires were returned during the sampling period (from 1^st February to 31^st March 2005). From these, 43 were excluded immediately due to completion by 18 year olds. In order to assign an approximate level of deprivation to each school, Index of Multiple Deprivation 2004 (IMD) 40 was calculated based on the Middle Super Output Area (MSOA) in which each school was located. For the North West, MSOA are geographical areas with an average total population of 7,380 (500 for those aged 15–19). Based on the IMDs of all MSOAs across the North West, each MSOA was allocated to a regional quintile of deprivation and students were allocated to quintiles based on the MSOA of their school. At this stage any schools that were not identified as local state schools (i.e. tax funded and recruiting in general from the local population) were removed from the analysis (n = 15 schools), leaving 132 general secondary schools out of a total of 474 such schools in the North West 41. In addition, as only 1% of pupils were 17 years old and 2% 14 years old, analyses were limited to 15 to 16 year olds where adequate numbers were available. The final sample for analysis was 10,271 respondents.

Initial analyses (Table 1) examined factors predicting whether individuals drink alcohol at all (here defined as drinking at least once every six months). However, most analyses (Tables 2 and 3) focused on identifying predictors of whether those that consume alcohol: usually have five or more alcoholic drinks per session (here, binge drinking); usually drink at least two or more days each week (here, frequent drinking); and drink in public settings (including drinking in streets, parks, pubs and clubs and excluding drinking at home, in friends' houses and on special occasions with family and friends). From these three risk factors, two composite variables were also calculated: "any drinking risk" (having any one of the three risk factors); and "all drinking risks" (having all three risk factors). Initial chi square statistics were calculated in SPSS 42. However, to account for confounding relationships between variables and potential underestimation of variance associated with cluster sampling 4344 STATA was utilised for logistic regression with appropriate cluster sample corrections applied 45.