OUP user menu


R. Harnett, R. Herring, B. Thom, M. Kelly
DOI: http://dx.doi.org/10.1093/alcalc/34.5.672 672-677 First published online: 1 September 1999


This paper reports the findings of an AUDIT questionnaire administered to a sample of young (16–24-year-old) white males. Of the sample, 65% were drinking at potentially harmful levels. Averaging 45 U/week, 18–21 years was the age of highest alcohol consumption. Compared with 16–17-and 22–24-year-olds, this group recorded the highest proportion of hazardous drinkers and most negative consequences of drinking.


At the same time that young people's illicit drug use has become more ‘normalized’ in recent years in Britain (Parker et al., 1995), there has been an increased concern in health promotion circles over their changing drinking practices (Health Education Authority, 1998). Evidence suggests that young men aged 18–24 years are the heaviest consumers of alcohol in the population (Office for National Statistics, 1998). Studies have found that alcohol consumption rises with age and peaks at 18–21 years of age (Plant et al., 1990; Health Education Authority/MORI, 1992). In this age group, the number of non-drinkers declines and the number of heavy drinkers increases (Goddard and Ikin, 1988; Goddard, 1991). We also know that, for a large proportion of young men, alcohol consumption patterns centre around a periodic ‘binge’ (Plant et al., 1990; Moore et al., 1994; Miller and Plant, 1996; Webb et al., 1996) and that, while for many bingeing is seen as ordinary drinking behaviour, drinking every day, even in small amounts, is considered to be a sign of dependence (Health Education Authority, 1998). These perceptions are, however, at odds with the revised sensible drinking message, which recommends daily benchmarks of 3 to 4 U/day (1 U = 8 g of alcohol) for men instead of the previous 21 U weekly limit (Department of Health, 1995).

This paper aims to advance knowledge of the pattern of alcohol consumption practices among young people by presenting the findings of research which set out to explore differences in drinking patterns of a sample of 16–24-year-old white males in one relatively deprived electoral ward in east London. In addition to being amongst the heaviest consumers of alcohol in the population (Office for National Statistics, 1998), it was decided to focus on young men from a deprived area because they have been largely ignored in research which tends to focus on school, college, and university populations. By focusing on this group, the report describes the alcohol consumption practices of these young people, and details important differences found in their drinking behaviour across the age range. In terms of promoting appropriate health messages, an understanding of the differences in the pattern of alcohol consumption among different groups of young people, in particular an appreciation of change in drinking practices over time, is crucial if they are to be targeted effectively.



As part of a larger study, whose aim was primarily to collect qualitative data on young men's drinking (see Harnett et al., 1998), 60 screening questionnaires which contained the Alcohol Use Disorders Identification Test (AUDIT) were administered between October and December 1997. AUDIT was developed by the World Health Organization (WHO) to identify hazardous and harmful alcohol consumption (Saunders et al., 1993). In tests, a score of 8 and above on the 10-item questionnaire has been found to be sensitive in predicting alcohol-related harms and current problematic drinking (Conigrave et al., 1995). In addition to AUDIT, five questions about the context surrounding usual drinking events were included and socio-demographic information was collected. To facilitate analysis, three age groups have been compared. At the time of the study, 16–17-year-olds were below the legal age to buy alcohol; based on previous evidence (Plant et al., 1990; Health Education Authority/MORI, 1992), 18–21-year-olds were considered to be at the peak alcohol consumption age, and 22–24-year-olds were the remainder of the sample in the heaviest drinking age range identified in the literature (Office for National Statistics, 1998).

Research sample

Using the London Research Centre (1995) findings on the level of deprivation in electoral wards in England and Wales, a research location was selected to represent the behaviour and views of young men living in a relatively deprived inner city area about their drinking. The selected ward was situated in London, because it afforded convenient access for the researcher. Seventeen 16–17-, 24 18–21-, and 19 22–24-year-olds completed the questionnaire in community settings in the chosen ward. The settings used were four public houses (17), a bench outside a sports centre (8), a bar with a liquor licence inside a leisure centre (8), the staff room of a youth club (16), a college refectory (4), and the homes of a network of friends (7). Twenty-four were students working towards General and National Vocational Qualifications (11), A levels (7) or higher qualifications (6), 17 were unskilled manual workers, seven were clerical and managerial, seven unemployed, three skilled manual, and two had professional occupations. Most of the young men were not married/cohabiting (54) and had no children (59). Thirty-seven lived with their parents; the remainder were renting privately (17), staying with friends (2), owner occupiers (2), and housing association tenants (2).


Because the questionnaire was part of a larger study, the sample was recruited over a 3-month period. The researcher went into the field in sessions averaging 4 h in duration at different times of the day and on all days of the week. Young men who appeared to meet the criteria of the chosen study group were approached and asked if they would like to participate in the study. This approach led to the completion of 53 questionnaires and five refusals. In addition, an indigenous 21-year-old white male was given training and asked to recruit members of his social network (7). All of these young men were aged between 21 and 24 years, were in employment, and none of those approached using this method refused to participate.

Data analysis

A number of measures were compared. The first four questions on the questionnaire aimed to examine the social context in which drinking was ‘usually’ situated. Using AUDIT, the following measures were examined: (a) drinking frequency; (b) number of drinks consumed on a typical occasion; (c) frequency of ‘binge’ drinking (the WHO definition of six or more drinks on one occasion was used); (d) drinking problems and consequences (AUDIT questions 4–10). In order to establish how many units of alcohol each respondent was drinking, an additional question was added to AUDIT asking respondents to indicate what their ‘usual drinking measure’ was. By interpreting the responses ‘bottle’ and ‘glass’ as 1 U and ‘pint’ and ‘double’ as 2 U, a calculation of typical mean weekly alcohol consumption [a sum of (a) and (b) above and ‘usual drinking measure’] was made and is shown in Table 1. In order to compare the differences in hazardous and harmful alcohol consumption across the age range, two measures were used. In Table 2, the AUDIT scores have been summarized according to the proportion of respondents with ‘low’ (<8) and ‘high’ (>7) scores in each age category, and Table 3 shows the ‘low’ and ‘high’ mean AUDIT scores of each age group.

View this table:
Table 1.

Mean weekly alcohol consumption of 16–17-, 18–21-, and 22–24-year-olds

View this table:
Table 2.

Percentage of sample with ‘low’ and ‘high’ AUDIT scores by age group

Age group (%)
AUDIT score16–1718–2122–24Total
Low (<7)53253235
High (>8)47756865
View this table:
Table 3.

Mean ‘low’ and ‘high’ AUDIT scores of 16–17-, 18–21-, and 22–24-year-olds

Age group (score)
Mean AUDIT score16–1718–2122–24
Low (<7)
High (>8)11.918.211.8


Alcohol consumption patterns and the level of hazardous and harmful drinking

Across the total sample, over 60% (38/60) drank premium lager (5% alcohol by volume or above) as their main usual drink, which was the most popular drink among all age groups. The most popular heavy drinking occasions of all 16–24-year-olds were Friday and Saturday nights (58/60), drinking in mixed sex groups was reported as the most usual combination of heavy drinking companions (44/60), and ‘the pub’ was the most popular place to drink in heavily (38/60).

The majority (65%; 39) scored above the AUDIT cut-off score, which predicts hazardous and harmful alcohol consumption. Their mean AUDIT score of 11.2, was also well above the critical score and the mean weekly alcohol consumption for the whole sample was 26 U/week. In all of these measures there were differences between age groups.


In the 16–17-year-old group, the proportion of respondents in ‘low’ (53%) and ‘high’ (47%) AUDIT categories was roughly equal (Table 2) and their mean AUDIT score was 8 (Table 3). On average, weekly alcohol consumption among this group was 11 U/week (Table 1). Their drink preferences included ciders and alcopops, they tended to use Saturday as their heavy drinking day and the majority drank most of their alcohol ‘at friends’ (10/17). All 16–17-year-olds reported drinking between one and five drinks on typical occasions, and most drank two to four times a month or less (12/17). Two binged weekly. Their most popular drinking measures were bottles (6/17) and pints (6/17).


Of the 18–21-year-olds, 75% were located in the ‘high’ AUDIT score bracket (Table 2) and their mean AUDIT score was 14.6 (Table 3). The mean weekly alcohol consumption of this group was 45 U/week (Table 1). In addition to premium lager, this age group's other most popular drink was spirits. Most said they drank heavily on Friday (14/24) and Saturday (15/25) and the majority reported drinking most alcohol ‘in the pub’ (15/24). On typical occasions, the majority of 18–21-year-olds said that they drank between five and 10 or more drinks (17/24), drinking this amount between two and four or more times a week (21/24). Half of them binged weekly or more often. Pints were the favoured drinking measure (13/24).


Sixty-eight per cent of 22–24-year-olds scored >8 on AUDIT (Table 2). The mean AUDIT score for this age group was 9.7 (Table 3) and their mean weekly consumption was 16 U/week (Table 1). Again, 22–24-year-olds preferred spirits as a second choice drink, most said that Friday was their heavy drinking day (14/19) and the majority drank most alcohol ‘in the pub’ (17/19). Most reported drinking between one and six drinks on typical occasions (17/19) and drinking between two and four or more times a week (13/19). The majority binged monthly or less (11/19), while pints were their most popular usual drinking measure (12/19).

Drinking consequences and alcohol-related problems

For the whole sample, over one-third (38%; 23) indicated that, on at least one occasion in the last year, they had been unable to stop drinking once started. Comparisons across groups showed an upward trend with increasing age: 17% reported this consequence at 16–17 years, 42% did so at 18–21 years, and 53% at 22–24 years. Almost half of the total sample (47%; 28) said that they had ‘failed to do what was normally expected’ of them, because of their drinking. This percentage was consistent over all groups. Breakdown of the 22% (13) who reported that they had needed a first drink in the morning to get themselves going after a heavy drinking session, showed that this problem occurred more often in the 18–21-year-old (33%) group, compared with 12% at 16–17 years of age, and 16% at 22–24 years of age. One-third of all young men (32%; 19) said that they had experienced ‘a feeling of guilt or remorse after drinking’. There were little differences between groups. Over one-half of the sample (57%; 34) claimed that, at some time during the last year, they had been unable to remember what had happened the night before due to their drinking. The highest proportion was found in the 18–21-year-old group (75%), closely followed by 22–24-year-olds (68%), while 17% of 16–17-year-olds reported this consequence. Over one-third (37%; 22) reported either sustaining or causing injury because of drinking. Again, the highest level was found in the 18–21-year-old group (58%), compared with 24% of 16–17- and 21% of 22–24-year-olds. Finally, 17% (10) said that a friend, doctor or health worker had been concerned about their drinking and advised them to cut down. More 18–21-year-olds (21%) reported this problem than 16–17- (12%), and 22–24-year-olds (16%).

Compared with 16–17-year-olds, the proportion of 18–21-year-olds with ‘high’ AUDIT scores increased from below 47 to 75%, three times that of those from the same age group recording ‘low’ scores. This remains approximately the same at 22–24 years old, indicating that between 18 and 24 years old, alcohol consumption among three-quarters of these young men was at a hazardous and harmful level. Comparisons of mean AUDIT scores across age groups produced a different pattern: mean ‘high’ AUDIT scores increased between 16–17 and 18–21 years, and fell again at 22–24 years. In both categories of mean AUDIT score, the 16–17-year-old and 22–24-year-old groups were roughly equal. In all groups, the total mean AUDIT scores were at or above the critical figure. Many of these trends are also mirrored in the levels and patterning of drinking consequences and alcohol-related problems reported by the sample.

In most of the measures of problems relating to drinking, it was the 18–21-year-olds who reported the highest levels of negative consequences. These findings support the alcohol consumption trends and the levels of hazardous and harmful drinking recorded by AUDIT in the study. They also accord with previous research which found this age group to be the heaviest consumers (Plant et al., 1990; Health Education Authority/MORI, 1992). According to the daily recommended limit and AUDIT, a large proportion of the young men were drinking at hazardous and harmful levels and, resulting from their drinking, many also experienced a range of negative consequences and problems. The data also indicate that, among this age group, consumption patterns and the context of drinking change over time.

At 16 or 17 years, drinking was at most a weekly Saturday night activity in which a mixture of lager, ciders and alcopops were most frequently drunk in mixed sex groups at friends' houses. On average, compared with the other groups, they were consuming the least alcohol and on the whole did not binge drink often. However, according to the revised recommended sensible drinking message, most were still drinking harmfully. At 18–21 years, drinking was different. These young men were drinking more measures of alcohol on more occasions, they were binge drinking regularly on both Friday and Saturday nights and, as a result, they were on average consuming over twice the previous recommended weekly benchmarks and four times more alcohol than the younger age group. They also preferred to drink ‘in the pub’, rather than in private. At 22 years, alcohol consumption decreased again. While drinking pints ‘in the pub’ remained popular on Friday nights, alcohol consumption among the 22–24-year-old group was less frequent, the amount they drank was less than that of 18–21-year-olds and the frequency of bingeing decreased from weekly or twice-weekly to a monthly pattern.


In this paper, we have reported the results of a survey of young white male drinkers living in one relatively deprived locality in east London. We fully acknowledge the limitations of our small sample size and do not claim these can be generalized to all young men. It should also be made clear that, due to the sampling procedure, the study cannot claim to be representative of 16–24-year-old males living in the ward. The results are inevitably biased towards young men in public settings and those likely to drink. However, what is interesting about this research is that the sample consisted of young drinkers, who were recruited due to their presence in leisure settings. This highlights two issues. First, with regard to the use of AUDIT, unlike similar studies, which have used AUDIT in a variety of institutional settings, such as universities and hospitals (Fleming et al., 1991; Barry and Fleming, 1993; MacKenzie et al., 1996; Piccinelli et al., 1997; Thom et al., 1997), this study was situated in the community. As a tool for monitoring alcohol consumption and providing an evaluation of the potential levels of harm within the wider community, we have not only shown that AUDIT would prove useful, but also that it is possible to administer it successfully. In addition, AUDIT was used in conjunction with other questions dealing with the contextual nature of drinking. This allowed for presentation of AUDIT scores and of the usual alcohol consumption practices of respondents, and, as a consequence of this design, a description of the way in which alcohol consumption practices differed across the age range. Second, in terms of how young men consume alcohol, this sample indicated that drinking above recommended daily benchmarks may be normal drinking practice. If this is so, problems related to drinking may be becoming more widespread among Britain's youth, with the risks being particularly acute among the 18–21-year-old age group.

Hopefully, our work will provide a useful reference point from which to conduct further research into the consumption patterns, alcohol-related consequences, and the level of harmful and hazardous drinking of other population groups in Britain. This sort of approach might prove useful in identifying and comparing differences in the population's use of alcohol which, in turn, would direct the most effective harm reduction strategies and ‘safe’ drinking messages at the most effective target groups.


The authors wish to thank the Health Education Authority for funding the research project. We are indebted to the young men who gave up their time to fill in the questionnaire and to Steve Fitch for helping to administer it.


  • *Author to whom correspondence should be addressed.


View Abstract