OUP user menu


Jan S. Gill
DOI: http://dx.doi.org/10.1093/alcalc/37.2.109 109-120 First published online: 1 March 2002


Results of a literature review of 18 studies investigating the drinking behaviour of undergraduate students at UK universities over a period of 25 years are presented. While comparison between studies is complicated by inconsistencies in the terms employed to describe drinking behaviour, it is concluded that significant numbers of both male and female students are reported to exceed sensible weekly consumption guidelines. Recorded levels of binge drinking among both female and male students are extremely variable between studies. Further research is needed to clarify this position. However, if the most recent research evidence is substantiated, female and male binge drinking levels may exceed those of their peers in the general population and their US counterparts. The reported ramifications of harmful drinking for the health and well-being of students are reviewed. A possible link between poor academic performance and alcohol consumption appears tenuous and merits further investigation. Evidence relevant to the view that the drinking behaviour of female students is changing is considered.


Several studies reported within the last decade have examined the drinking behaviour of adolescents and young adults. UK teenagers are characterized by high levels of intoxication and binge drinking (identified as more than five drinks consumed in a row), when compared with their European counterparts (Hibell et al., 2001). Recent evidence suggests that these high levels are being maintained but, significantly, have not increased further (Miller and Plant, 2001).

Drinking behaviour among another group of young adults, namely college students, has been studied extensively within the USA. Wechsler et al. (1998) stated that ‘binge drinking is by far the single most serious public health problem confronting American colleges’ (p. 57). It is the purpose of this review to consider studies which have monitored the drinking behaviour of undergraduate students within the UK. Particular focus will be given to reported levels of drinking in excess of health guidelines, ‘binge’ drinking, and its consequence, ‘heavy’ or high risk drinking, and female drinking behaviour.


Sample populations

The sample population and methodology employed in 18 studies identified by literature search (using MEDLINE, CINAHL and ArticleFirst databases) are summarized in Table 1. In all but one study, drinks consumed are presented in terms of the UK unit, one unit being equivalent to 8 g of pure ethanol (alcohol). All studies have employed self-report questionnaires to assess alcohol consumption retrospectively and almost 50% have targeted medical or dental students. The quality of the method of sample selection has varied greatly. For example, Webb et al. (1996) surveyed students at ten UK universities and then randomly selected students to reflect the distribution characteristics of the faculties of UK universities, whereas Norman et al. (1998) recruited students following a personal approach by a researcher. However, although undergraduate students in all 5 years of study have been investigated, there was generally greater representation of the earlier years.

View this table:
Table 1.

Summary of research studies investigating drinking behaviour of UK undergraduate students

Author(s)Study populationAge (years; range or mean ± SD)Academic year of studyWhen administeredMethodResponse rateDefinitions of ‘drink unit’ and ‘binge drinking’
Orford et al. (1974)1323 students (80% male) (Faculties of Art, Engineering, Law, Medicine, and Science) (2 colleges of one University)17–191st1st week of first term44-page questionnaire distributed via internal mail. Drinking quantity and frequency over preceding 12 months and attitude to drinking examined76%1 drink = 8 g alcohol.
 (a) >10 drinks within 21/2 h;
 (b) >10 drinks within 10 h;
 (c) 6 drinks with 1 h;
 (d) 6 drinks within 4 h
3rd1st week of first term
Brown and Gunn (1977)321 students in halls of residence (57% male), 10% of student population targetedNot reportedNot reportedNot reportedQuestionnaire relating to type and frequency of alcohol consumption distributed to students73%1 drink = 1 UK unit. ‘spree’ drinking
Anderson (1984)1083 students (48% male). Random sample stratified by sex; 75% in hall of residence20.2 ± 2.2 (female), 20.3 ± 2.1 (male)1–34th week of second termQuestionnaire relating to actual drinking in last 4 days (questionnaires despatched either Friday or Monday). Problems associated with drinking in previous month recorded. Delivered via internal mail76% female, 69% male1 drink = 1 UK unit. ‘Harmful daily limit’; 8 units male, 6 units female
Collier and Beales (1989)260 medical students (52% male)Not reported5 years 29% in year 1, 15% in year 5Not reportedQuestionnaire (quantity and frequency) relating to weekly alcohol consumption and 7 day retrospective diary. (Also investigated knowledge relating to alcohol consumption and administered CAGE questionnaire for alcohol abuse.) Questionnaire distributed to students, ‘freely available’1 drink = 1 UK unit
West et al. (1990)270 students (46% male)211–3Not reportedQuestionnaire and shortened form of Michigan Alcoholism Screening Test employed. Amount consumed in average week investigated as was drinking behaviour. Questionnaire sent to students69%1 drink = 1 UK unit
Leavy and Alexander (1992)185 studentsNot reported1Spring termQuestionnaire. Quantity–frequency index of consumption. Self-report perceptions of problem drinking. How administered not reportedNot reported1 drink = 13 g alcohol
File et al. (1994)774 medical and dental students; 2 medical schools (58% male). Group 1: 1st year; Group 2: years 2,4,5; Group 3: 1st yearNot reported1, 2, 3 and 5Group 1 1st term; Group 2 3rd term; Group 3 Freshers' week 2nd and 3rd termQuestionnaire relating to diet, exercise, health, smoking and alcohol intake. Alcohol intake assessed by ‘typical week’ and ‘last week’. Administered at lectures38–100%1 drink = 1 UK unit. >10 units on more than one occasion per week
Ghodse and Howse (1994)442 medical students (53% male) at one institution, 835 medical students (52% male) at 12 institutions; total sample 1278 students21.4 ± 2.31–5Not reportedQuestionnaire monitoring alcohol use last week and typical week. Postal questionnaire48–92%1 drink = 1 UK unit
Ashton and Kamali (1995)186 medical students (41% male). Data pooled over 2 years20.4 ± 1.82Last week of spring summerDrinking ‘habits’ studied. Eysenck personality questionnaire and hospital Anxiety Depression scale plus questions relating to smoking, exercise and illicit drug use also employed. Administered at lecturesAttendance at class = 65%. Response rate = 99%.1 drink = 1 UK unit
Delk and Meilman (1996)456 students (44% male). Randomly selected; 58% in hall of residence14.9% aged <19 years; 3.6% aged; >25 years;1–4Not reportedPostal questionnaire (Core Alcohol and Drug Survey). Qualitative data obtained via focus groups. No gender specidic data presented65%Drink not defined. ≥5 drinks in one sitting
Webb et al. (1996)3075 students (53% male) from 10 universities. Randomly selected to represent interfaculty distribution of UK University20.9 (188 mature students)2 (Scotland 3)Spring termsSee Ashton and Kamali (1995). How drinking behaviour recorded not reported. Administered at lectures‘Nearly’ 100% of those present. Lecture attendance rate not given1 drink = 1 UK unit (modification for strong lager). Drinking more than half the sensible weekly limit (>7 for a female and >10 for a male)
Hannay (1998)3 separate studies involving random samples: (A) 46 medical students; (B) 100 students in halls of residence; (C) 88 students also in halls of residence. ‘Approximately equal numbers of men and women’18–24Not reportedNot reportedQuestionnaire. How drinking behaviour recorded not reportedNot reported1 drink = 1 UK unit
Webb et al. (1998)754 students (44% male) from 7 medical schools18–242Spring and autumn termsSee Ashton and Kamali (1995). Administered at lecturesVirtually 100% attendance at lectures. Varied from 52 to 90%1 drink = 1 UK unit. Binge drinking: see Webb et al. (1996)
Williams and Clark (1998)300 students (54% male)22.1 ± 6.42Not reportedNot reportedQuantity and frequency within the last year and largest amount of drinks consumed on a single occasion and frequency in last yearNot reported1 drink = 1 UK unit. Largest number of drinks ever consumed in a single occasion, and frequency
Norman et al. (1998)136 students (43% male). Individuals approached personally by researcher18–27Not reportedNot reportedQuestionnaire relating to typical weekly consumptionNot reported1 drink = 1 UK unit. Binge drinking: see Webb et al. (1996)
Pickard et al. (2000)136 medical students (34% male)Not reported2Not reportedQuestionnaire relating to alcohol and illicit drug use. Details of how drinking behaviour assessed not given. Administered at lectures81% of those present (65% of year)1 drink = 1 UK unit. Binge drinking: see Webb et al. (1996)
Underwood and Fox (2000)199 dental students (45% male)Not reported1–53rd termQuestionnaire consumption ‘last week’ and ‘typical week’ monitored. Administered at lectures and absentees contacted via internet mail75%1 drink = 1 UK unit. Binge drinking: see Webb et al. (1996)
Newbury-Birch et al. (2000)194 medical students (33% male)18.8 ± 2.111st week 1st termQuestionnaire relative to lifestyle variables (including alcohol use) stress anxiety and personality. Typical weekly consumption of alcohol before entering university recorded. Administered at lectures95%1 drink = 1 UK unit. Binge drinking: see Webb et al. (1996)

Methods and procedures of assessment of drinking behaviour

Drinking behaviour has been recorded by a variety of methods. Delk and Meilman (1996) employed a validated questionnaire; the Core Alcohol and Drug Survey. Orford et al. (1974), Brown and Gunn (1977), Leavy and Alexander (1992) and Williams and Clark (1998) favoured quantity–frequency questionnaires. The time frame selected for investigation differs; Orford et al. (1974) favoured 1 year, whereas Brown and Gunn (1977) studied 1 week. Norman et al. (1998) and West et al. (1990) reported ‘average weekly’ consumption. Retrospective diaries were analysed by Anderson (1984) for a 4-day period (but these were distributed either on a Friday or a Monday). Collier and Beales (1989), Ghodse and Howse (1994) and Underwood and Fox (2000) investigated consumption ‘last week’ and on a ‘typical week’. File et al. (1994) adopted a similar format but questions relating to alcohol consumption were within a questionnaire investigating diet, health etc. — a format which may yield more reliable estimates of alcohol consumption (King, 1994).

Ashton and Kamali (1995) investigated ‘drinking habits’ and estimated weekly consumption figures from a questionnaire that also examined personality characteristics and lifestyle variables. Webb et al. (1996, 1998) cited this paper in their methodology sections. Hannay (1998) and Pickard et al. (2000) did not provide details of how drinking behaviour was recorded.

Several researchers distributed questionnaires at lecture times and obtained relatively high completion rates as a result (e.g. File et al., 1994; Ashton and Kamali, 1995; Webb et al., 1996, 1998; Newbury-Birch et al., 2000; Pickard et al., 2000). However, often no information is available to characterize absent students. Ashton and Kamali (1995) recorded a 99% response rate but a 65% attendance rate. Webb et al. (1996) suggested non-attendees might be heavy users of alcohol.

Time of study

The time point during the academic year selected for study is important. Nine studies provided no such information. ‘Freshers’ week, the pre- and post-examination periods, etc., may be influential factors for drinking behaviour. Orford et al. (1974) and Newbury-Birch et al. (2000) selected the first week of the first term for study. Similarly, the academic year of study merits attention. Freedom from parental supervision, peer pressure and stress associated with living away from home may influence first year drinking behaviour in particular. File et al. (1994) attempted to report drinking behaviour for different year groups separately while other studies made no such distinction (e.g. Delk and Meilman, 1996; Hannay, 1998).

Residence of study subjects

Details of place of residence were provided by very few studies. Brown and Gunn (1977) studied students in halls of residence exclusively. In the study of Anderson (1984), 75% of students lived in halls while for the Delk and Meilman (1996) study group the figure was 59%.

Age of study subjects

Ten of the studies cited in Table 1 did not report the age of their subjects adequately. Mature students may have greater family and economic commitments, which in turn may influence drinking behaviour. Large population studies within the USA have suggested that students aged 17–23 years have much higher binge drinking rates than older students (Wechsler et al., 1994).

Beverage and units of measurement

There is little evidence that researchers have considered brand name of alcohol consumed, an important precaution when quantifying consumption (see Stockwell and Single, 1997). Webb et al. (1996) did make an allowance for the different alcoholic strengths of commercially available lager, and Leavy and Alexander (1992) considered the different alcohol content of Scottish and US beer. However, a considerable underestimation of alcohol consumption is likely to result when drinking behaviour is considered simply in terms of UK ‘standard drinks’ (one half pint of beer, lager or cider, one measure of spirits or one glass of wine) — a policy adopted by the majority of studies. Lager in the UK can vary from 4% alcohol by volume (ABV) to 8.4% ABV. One half pint will contain, not one, but 1.1–2.4 units of alcohol. Similarly the typical ABV of wine is 11–12%. A 125 ml glass will contain 1.5 UK units, a 175 ml glass ∼2 units. Further inaccuracies in the estimation of alcohol consumption may be evident when the report by Anderson (1984) is considered; 25% of male and 45% of female students living outside college reported drinking either in their own or someone else's room (for living within college the figures were 27% and 38% respectively). Drinking measures may well be more generous in these situations. Both the above factors would act to underestimate alcohol consumption.

In addition, it must be stressed that all measures of drinking behaviour have resulted from self-reported data. Problems of under-reporting or over-reporting of consumption may arise.

Other definitions

Different definitions of ‘safe’, ‘heavy’ and ‘binge’ drinking have been employed. In the case of binge drinking, some studies have made no gender distinction (File et al., 1994; Delk and Meilman, 1996). Valid comparison between studies of reported incidences of different types of drinking behaviour in some instances is therefore problematic.

Some authors have reported drinking behaviour in terms of a percentage of ‘those students who drank’, whereas others have reported their results in terms of ‘all students sampled’. Where possible findings have been altered to the latter form for the purposes of this review.


Not all studies reported abstention rates (see Table 2) and different forms of definition were employed. Anderson (1984) defined an abstainer as someone who never drinks, whereas Orford et al. (1974) recorded those who drank on three or fewer occasions in the last year. The majority of studies have reported those who drank no alcohol in the typical week as abstainers. Gender-specific data were not always provided. In the case of the study by Underwood and Fox (2000), abstention rates have been calculated from available data.

View this table:
Table 2.

Reported levels of abstinence and drinking exceeding sensible weekly limits of 14 units for females and 21 units for males

Author(s)AbstainersExceeding 14 units for females, 21 units for males
Orford et al. (1974)5% males. 3.5% females reported drinking on ≤3 occasions in preceding 12 monthsNot reported
Brown and Gunn (1977)3.8% males, 6.5% femalesNot reported
Anderson (1984)3% malesNot reported
4% females ‘never drank’21% of females exceeded 6–10 units in the 4 day study period
45% of males exceeded 6–10 units in the 4 day study period
Collier and Beales (1989)6% males34% of males
8% females43% of females exceeded 11 units per week
West et al. (1990)12.0% males, 13.8% females37.5% of males
(zero consumption)20.7% of females
Leavy and Alexander (1992)Not reportedNot reported
File et al. (1994)Of Asian students n = 271 54% males,‘about one third’ of non-Asian males (58% in year 5)
65.5% females12–22% of non-Asian females (26% in year 5)
Not reported for non-Asian students
Ghodse and Howse (1994)11% of males17.9% of males
8% females, single institution (n = 442)3.9% of females in single institution
13% males25.2% of males
14% females, 12 universities (n = 835)13% of females in 12 universities
Ashton and Kamali (1995)18.4% males26.3% of male drinkers
12.8% females18.3% of female drinkers
Delk and Meilman (1996)5.8%: no gender-specific data21.5% of drinking students exceeded 15 units per week.
No gender-specific data
Webb et al. (1996)10.9% males, 10.6% females61% of male drinkers (54% of males)
(6% white students, 52% non-white students)48% of female drinkers (43% of females)
Hannay (1998)Not reported63% of males, 58% of females
Webb et al. (1998)12.9% males, 16.3% females48.3% males
(5.8% white students, 43.3% non-white students)38.3% females
Williams and Clark (1998)Not reported17.0% of males and 4.6% of females exceeded 21/28 units per week
Norman et al. (1998)Not reported50.8% of males
16.9% of females
Pickard et al. (2000)17% male41% of males
12% female41% of females
Underwood and Fox (2000)18% male51% of males
10% female38% of females
Newbury-Birch et al. (2000)8% males, 12% females49% of male drinkers, 43% of female drinkers
(52% non-white students, 4% white students)(45% of all males, 39% of all females)

Generally speaking, the pre-1989 studies recorded lower levels of abstention than the later studies. The General Household Survey (GHS) conducted in 1998 (Bridgwood et al., 2000) reported levels of abstainers within the 16–24-year age group as 7.7% for males and 12.2% for females. Mean data from studies in Table 2 reported from 1998 to date (n = 4) would suggest a higher abstention rate for males of 14%, but for females, a similar value (12.6%). Interestingly, the two studies by Pickard et al. (2000) and Underwood and Fox (2000) reported higher abstention rates in males than females.

However, any attempt to make definite conclusions regarding the extent of abstention among the UK undergraduate population from present data would seem ill-advised. The ethnic origin of the students studied is a critical factor and this is not always reported. File et al. (1994) noted high abstention rates in Asian students (54% of males and 65% of females), whereas Webb et al. (1996) reported rates for ‘all’ students of 11%, but 6% for white students and 52% for non-whites. Newbury-Birch et al. (2000) reached similar conclusions.

This evidence may explain the relatively low abstention rates reported by one recent study (Delk and Meilman, 1996). This population was 95% white. Another relevant factor is the influence of university life on drinking behaviour. Thus, the time of survey may be critical. File et al. (1994) reported that eleven of 66 medical students who were teetotal at matriculation were regular drinkers by the end of the first year.


While current advice on sensible drinking, within the UK, advocates consideration of daily consumption levels and the number of alcohol-free days (Department of Health, 1995), many of the reviewed studies have used weekly benchmarks of 14 units for women and 21 units for men to monitor safe or sensible drinking behaviour. In the GHS survey (Bridgwood et al., 2000), 36% of men and 13% of women within the 16–24-year age group exceeded these weekly consumption guidelines.

The study of West et al. (1990) provides similar data for male students (see Table 2), but the later studies of Webb et al. (1996, 1998), Norman et al. (1998), Hannay (1998), Pickard et al. (2000), Underwood and Fox (2000) and Newbury-Birch et al. (2000) all reported higher percentages of 54, 50.8, 63, 48, 41, 51 and 45%, respectively (mean ± SD: 52 ± 8%). This most recent evidence would suggest that male students are more likely than males in the general population to exceed 21 units of alcohol per week, and more importantly, that approximately one in two male students is doing so.

For females, File et al. (1994) reported that between 12 and 22% of non-Asian females exceeded 14 units per week in the first three undergraduate years. This figure rose to 26% in year 5. Ashton and Kamali (1995) reported a figure of 18.3% while Norman et al. (1998) recorded 16.9%. Again the later studies (Webb et al., 1996, 1998; Hannay, 1998; Newbury-Birch et al., 2000; Pickard et al., 2000; Underwood and Fox, 2000) present higher values; 43, 58, 38, 41, 38 and 39%, respectively; mean ± SD: 43 ± 7%.

Comparison with GHS data would suggest that female students are at least three times more likely to exceed weekly guidelines than their counterparts in the general population.

Significantly, the difference between the genders clearly evident in the GHS data is less obvious in the UK undergraduate population. GHS figures suggest a male:female ratio of 2.8:1 for those exceeding weekly guidelines. Studies cited above might suggest a male:female ratio of 52:43, i.e. 1.2:1.

According to several investigators, the main reason given by students for drinking was pleasure (Ashton and Kamali, 1995; Webb et al., 1996, 1998; Newbury-Birch et al., 2000) or ‘enjoying the taste’ and ‘being sociable’ (West et al., 1990). Less than 10% felt that examinations were the motivation (Ashton and Kamali, 1995). Other factors which might influence drinking behaviour, e.g. student demographics, require further investigation.


As with many terms employed within the alcohol research field, ‘binge drinking’ is problematic — both in definition and semantics. It has nevertheless been adopted by many of the studies reviewed here (although seven different definitions are employed). ‘Binging’ may imply drinking over a prolonged period (at least 2 days) and the term ‘heavy episodic’ drinking may be more appropriate (Schuckit, 1998). Murgraff et al. (1999) prefer the term ‘risky single occasion drinking’ (RSOD), whereas Measham (1996) refers to ‘heavy sessional drinking’.

While there is also a lack of consensus regarding the critical amount of alcohol drunk at one session [Wright and Cameron (1997) defined it as >8 units of alcohol in one day, Measham (1996) as >11 units on one occasion and Davey (1997, p. 11) adopted a more qualitative definition: ‘a plan or intent to drink more than the normal or to get drunk’] it is becoming clear that any definition must consider three important factors: gender [to accommodate sex-differences in alcohol pharmacology (Baraona et al., 2001)], duration of consumption and drinking frequency.

When considering ‘potential for harm’, ‘time to consume’ is an important factor. Given a rate of alcohol metabolism of ∼8 g (1 UK unit) per hour, 6 units of alcohol will pose a much greater risk for an individual when consumed within 1 h rather than at equal periods from 18:00 until 24:00. However, some researchers might classify both patterns as binge drinking. The study reported by Orford et al. (1974) (see Table 1) is quite unique in its attempt to monitor drinking speed. For this type of drinking pattern, the term ‘concentrated drinking episode’ (CDE) may be appropriate.

It is also essential that the frequency of binge drinking be considered. Some studies did not specify a time frame. Binge drinking at least once is likely to be undertaken by many young people within the 18–24-year age group. It is much more important from a health perspective to know how regularly this form of drinking occurs. Additionally, Gladstone et al. (1996) cited evidence that the volume of alcohol consumed at binge sessions by women in the US is increasing. Thus, to better inform health education planning, it would be useful in future UK studies to monitor the actual amount of alcohol consumed at one session in addition to the frequency of sessions.

Study findings are summarized in Table 3. Brown and Gunn (1977) reported that 24% of males and 10% of females took part in ‘spree drinking’, but did not define this term. In sharp contrast, as noted, Orford et al. (1974) provided a very detailed description of how they defined binge drinking. Almost two in five male students reported drinking >10 units of alcohol within 2½ h and one in ten males had done so more than ten times in the previous year. For females, very similar figures were recorded for the consumption of six units within 4 h (see Table 3). File et al. (1994) reported that among non-Asian males, 27% had taken part in binge drinking during the previous week. Delk and Meilman (1996) noted that 62.6% of students had binged in the last fortnight (11% more than five times). A binge was defined as five US drinks per session and no gender-specific data were provided.

View this table:
Table 3.

Reported levels of binge drinking and heavy or hazardous drinking

AuthorBinge drinkingHeavy or hazardous drinking
Orford et al. (1974)38.8% of males and 11.5% of females reported drinking >10 drinks within 2½ h, ‘ever’ in previous 12 months (10.9% of males and 1.7% of females had done so more than 10 times). For 6 drinks within 1 h the figures were 34.2% (males) and 5.2% (females). For 6 drinks within 4 h, the figures were 50.2% (male), 37.4% (female)For 6.1% of females and 31.6% of males, the usual quantity of alcohol consumed was 4–10 drinks per session. For 0.4% of females and 0.8% of males, the usual quantity was 11+ drinks. 59.2% of males reported being drunk, 12.1% >10 times in 12 months. For females, the figures were 34.8 and 3.9% respectively. 50.4% of males and 27.4% of females reported ever having a hangover (8.3% and 3.5% >10 times)
Brown and Gunn (1977)24% of males and 10% of females took part in ‘spree’ drinking bouts3% of males reported that they got drunk often while 10% reported that they often had a hangover. For females, the comparable figures were 6% and 3%
Anderson (1984)7% of men and 2% of women exceeded the harmful daily limit of 8 units (male) 6 units (female)19% of all males and 9% of all women had been drunk three or more times in the last month. Heavy drinking (more than 20 units in 4 days) was reported by 19% of all men and 4% of all women
Collier and Beales (1989)Not investigated23% of all males exceeded 35 units per week. 22% of all females exceeded 21 units per week; 12%, 35 units. 9% of all males and 6% of all females exceeded 50 units per week
West et al. (1990)Not investigated25.6% of males exceeded 35 units per week. 14.5% of females exceeded 21 units per week; 3.4% exceeded 35 units. 6% males, 1% females rated as problem drinkers on MAST scale
Leavy and Alexander (1992)Not investigatedAverage consumption per week for males = 34.66 ± 31.79 equivalent to 56 ± 52 UK units. For females: 11.72 ± 13.07 equivalent to 19 ± 21 UK units
File et al. (1994)Of non-Asian male students all those drinking >35 units per week (n = 40) drank >10 units on one or more occasions per week. Of those classed as safe drinkers (<21 units per week), 23.7% binge drank in this way. Binge drinking rates for male high drinkers (21–35 units per week) and for females not reportedOf non-Asian males 16% were drinking >35 units per week. For females drinking >21 units, the figure was 9.8%
Ghodse and Howse (1994)Not investigated11% of all males exceeded 35 units per week; 4% exceeded 51 units. 4% of all females exceeded 26 units per week with 1% exceeding 35 units per week
Ashton and Kamali (1995)Not investigated7.9% of all males exceeded 35 units per week. 8.3% of all females exceeded 21 units per week; 1.1% exceeded 35 units
Delk and Meilman (1996)62.6% of all students had binged in the last fortnight (11% >5 times). No gender-specific data reported88% of students reported a hangover in the last year (32.4% ≥10 times). 3.3% consumed alcohol every day. 20.7% 5 times per week. 10% of drinkers drank 21 or more ‘drinks’ per week
Webb et al. (1996)31% of male drinkers and 24% of female drinkers (28% of all males, 21% of all females)31.5% of all males exceeded 35 units per week. 26% of all females exceeded 21 units per week. 15.6% of all males exceeded 50 units per week, whereas 9.4% of all females exceeded 25 units
Hannay (1998)Not investigatedTwo sub-groups, B (26% of males and 15% of females); C (23% of males and 13% of females) identified as problem drinkers using CAGE score
Webb et al. (1998)23.9% males and 14% females27% of all males exceeded 35 units per week. 12.3% exceeded 50 units. 22.5% of all females exceeded 21 units per week; 6.7% exceeded 35 units. 42.6% of men and 36.4% of women reported that, within the last 6 months, they had been seriously affected by alcohol consumption. For 17.6% of males and 12.2% of females, this has occurred 6, to >21, times
Williams and Clark (1998)Not reported17.0% of males consumed more than 28 units per week. 4.6% of females consumed more than 21 units per week
Norman et al. (1998)46.3% of all students reported binge drinking at least once per week. Males reported an average of 4.67 binge sessions per month, females 2.46. 64% of males and 32.5% of females binge drank at least once per weekNot reported
Pickard et al. (2000)50% of all males and 63% of all females23.9% of all males exceeded 35 units per week, 2.2% exceeded 50 units. 22.1% of all females exceeded 21 units per week, 4.4% exceeded 35 units
Underwood and Fox (2000)46% of all males and 53% of all females reported binge drinking in a typical week10% of all male students exceeded 50 units per week. 63% of all female students exceeded 35 units per week
Newbury-Birch et al. (2000)27% of males, 14% of females38% of all males exceeded 28 units per week; 16%, 50 units. 17% of all females exceeded 21 units per week, 5%, 35 units

Webb et al. (1996, 1998), Norman et al. (1998), Pickard et al. (2000), Underwood and Fox (2000) and Newbury-Birch et al. (2000) all defined binge drinking as >7 units for females, 10 units for males (half the weekly recommended units) in one session (Moore et al., 1994). These authors recorded binge drinking levels as 31, 23.9, 64, 50, 46, 27%, respectively, for males and 24, 14, 32.5, 63, 53 and 14%, respectively, for females.

The considerable variation in results may be partially explained by differences in the time frame of measurement. Norman et al. (1998), Underwood and Fox (2000) and Newbury-Birch et al. (2000) report levels occurring in a typical week. It is not clear if Webb et al. (1996, 1998) and Pickard et al. (2000) used the same reference period. Newbury-Birch et al. (2000) reported low values but they studied binge drinking before university admission. The lowest values for both sexes were recorded by Webb et al. (1998) who studied seven institutions, but as many as 42.6% of males and 36.4% of females had been debilitated by alcohol one to five times in the last 6 months. Two studies (Pickard et al., 2000; Underwood and Fox, 2000) reported higher levels of binge drinking in females than males.

In 1998, for the first time the GHS in the UK (Bridgwood et al., 2000) investigated daily alcohol consumption. Within the 16–24-year age group (n = 698), 37% of men reported drinking ≥8 units on 1 day in the last week. For women, the corresponding value for drinking ≥6 units was 23% (n = 809). While the values reported by the student research are clearly variable (particularly for female students), three groups of authors, Norman et al. (1998), Pickard et al. (2000) and Underwood and Fox (2000), reported values for both genders in excess of those in the GHS survey which used less demanding criteria to define binge drinking. As with ‘sensible drinking’, the UK student population may be binge drinking more often than their peers in the general population. A suggestion (Pickard et al., 2000) that at least three in five female undergraduates are binge drinking merits investigation. Alternatively, the lower student binge drinking levels reported by Webb et al. (1996, 1998) from the two largest study populations and involving more than a single institution, may be the most reliable. However, they too show some disparity (females 24 and 14%; males 31 and 23.9%).

Within the USA, several national studies (using a slightly less demanding definition of binge drinking) have reported that ∼44% of all students binge (see, e.g., Wechsler et al. 1998). Wechsler et al. (1995a) recorded 50% of male students to be binge drinking (around 8 UK units per session) and 39% of females (∼6.5 units) at least once in the preceding fortnight. One study cited in Table 3 (Delk and Meilman, 1996) did make a direct comparison between UK and US student populations (these results are not shown in Table 3). These latter authors did not present gender-specific data, but defined binging as five or more US drinks in one sitting in both student samples. More Scottish students (62.6%) reported binge sessions than US students (40.4%) and more had binged three or more times in the previous fortnight (31.4 versus 16.3%).

In summary, binge drinking levels have not been reported in all student-drinking surveys. While definitions of this term vary, reported levels of binge drinking differed even among those research reports adhering to the same format. There is thus an urgent need to establish whether evidence of extremely high levels of binge drinking in both sexes, but particularly in women, is real. However, future studies must adopt a precise definition of this drinking term. Amount and speed of alcohol consumption must be monitored, and time frame of measurement must be stated.


Not all authors cited in Table 1 defined heavy drinking in the same terms, nor did all specifically quantify consumption exceeding 35 units per week. The following discussion will relate to reports of female drinkers exceeding 35 units per week, and males exceeding 50 units per week.

Within the GHS survey (Bridgwood et al., 2000) the extent to which males within the 16–24-year-old age group were recorded in 1988 to be drinking >50 units per week was 10% (n = 1356), increasing to 13% (n = 701) in 1998. Collier and Beales (1989) reported a figure of 9%, Ghodse and Howse (1994) 4%, Webb et al. (1996) 15.6%, Webb et al. (1998) 12.3%, Underwood and Fox (2000) 10%, Pickard et al. (2000) 2% and Newbury-Birch et al. (2000) 16%. Again disparity between results is evident. All studies, apart from those of Webb et al. (1996) and Newbury-Birch et al. (2000) reported values below that of the GHS data. Webb et al. (1996) studied the greatest number of students from across the faculties, whereas the remaining authors investigated students from the faculties of medicine or dentistry. Newbury-Birch et al. (2000) asked students about their drinking behaviour before coming to university and reported the highest values of all. The very low value reported by Pickard et al. (2000) may reflect the small sample size (n = 46).

There is some tentative evidence that around one in ten male medical students may be consuming >50 units of alcohol per week. While this level of consumption may be mirrored in the general population, its adoption by so many future health care workers is surely cause for concern. Collier and Beales (1989) identified 36% of male medical students giving positive responses to the CAGE questionnaire for alcohol abuse, whereas the use of the Michigan Alcoholism Screening test (MAST) identified 28% of males with some degree of problem (West et al., 1990).

For female students, Collier and Beales (1989) reported that as many as 12% were exceeding 35 units per week. Generally, the figures reported by other authors are much lower: 3.4% (West et al. 1990), 1% (Ghodse and Howse, 1994), 1% (Ashton and Kamali, 1995), 9.4% (Webb et al., 1996), 6.7% (Webb et al., 1998), 6.3% (Underwood and Fox, 2000), 4.4% (Pickard et al., 2000) and 5% (Newbury-Birch et al., 2000). The latter four most recent studies provide values very similar to that (6%) recorded in 1998 by the GHS survey. The CAGE questionnaire identified 29% of female medical students providing positive responses (Collier and Beales, 1989) whereas West et al. (1990), using MAST, reported that 17% of females had some degree of problem. It is interesting that two studies involving medical students recorded female heavy drinking levels exceeding those of males (Collier and Beales, 1989; Pickard et al., 2000).


Several investigations considered the repercussions of excessive student drinking. These will be described in terms of ‘secondary’ binge effects, influences on sexual health, behavioural consequences, and finally, evidence that alcohol impairs academic performance.

Secondary binge effects

Non-drinking associates of a binge-drinking student (Wechsler et al., 1998) feel secondary binge effects. Orford et al. (1974) reported that 8.5% of male students and 3.5% of female students had broken or damaged friendships through drink. In another study, 16.8% of males and 5.5% of females had ‘neglected their obligations’, due to drinking [a finding consistent with the Schuckit (1998) view of how binge drinking should be defined], 3.2% of males and 0.7% of females admitted losing friends, while 19% of males and 10% of females had been assaulted as a result of alcohol use (West et al., 1990). File et al. (1994) reported that, while 50% of males drinking >35 units of alcohol per week admitted to being physically hurt, 15% admitted hurting others as a result of their drinking behaviour. In their study of Scottish students, Delk and Meilman (1996) found that 53.2% of students, due to alcohol use, had done something they later regretted.

Sexual health

The ramifications of heavy drinking for the health of an individual have been the subject of much research. An association with unplanned pregnancy, risk of HIV infection, etc., have been described (Robertson and Plant, 1988). Delk and Meilman (1996) reported that 15.8% of students had been taken advantage of sexually, whereas 7.8% admitted that they had taken advantage of another sexually within the last year as a consequence of alcohol use.

Behavioural consequences

Orford et al. (1974) reported that 8.3% of males and 4.3% of females had been involved in an accident in which someone was hurt, 30.3% of males and 13.8% of females had broken the law with no police involvement, and 18.6% of males and 6.9% of females concealed the amount of alcohol they were drinking. West et al. (1990) found that 20% of male and 6% of female students had caused damage to property. Delk and Meilman (1996) recorded that 88% of students had had a hangover in the last year (32.4% ten or more times), 14.5% had taken part in acts of violence, 20.9% had been hurt or injured and 27.9% had been in arguments or fights.

While a range of assumed consequences of alcohol misuse have been reported to date, further research is needed to evaluate the strength of the link between any one behaviour and a particular drinking profile. In addition, the behaviour of different subgroups of drinkers should be monitored prospectively.

Within the USA, Wechsler et al. (1994) reported from a large study involving 140 colleges that frequent binge drinkers (i.e. binging three or more times in a 2-week period) were seven to ten times more likely than non binge drinkers not to use protection during intercourse, engage in unplanned sexual activity, get injured, and finally damage property.

Academic performance

Several studies present findings that imply a relationship between drinking behaviour and studying ability, and by implication, poor academic performance. Orford et al. (1974) noted that 11.9% of male and 8.9% of female students had missed a class due to their drinking behaviour. For 6.8% of males and 3.1% of females, this had occurred more than once in the last 12 months. Poor working over a period of some days or longer was reported by 8.1% of males and 5.0% of females. Fletcher (1974), as cited in Leavy and Alexander (1992), recorded that 49% of medical students had missed morning lectures due to drinking. In another study (Anderson et al., 1984), 10% of female and 11% of male students admitted that alcohol had interfered with their studying in the previous month. As many as 53% of students in the Collier and Beales (1989) study claimed a negative effect on academic performance (6.5% frequently). In the Delk and Meilman (1996) study, 19.2% of students maintained that they had performed badly in academic work and 56.2% had missed a class as a consequence of their drinking. In the Webb et al. (1998) study, 42.6% of male and 36.4% of female medical students claimed that alcohol consumption had affected their ability to work the next day or had caused serious debilitation from one to five times within the previous 6 months. In another study involving medical students, 36.8% of males and 58.2% of females considered that their performance had been affected by alcohol consumption on a least one day in the last month (Pickard et al., 2000).

It must be borne in mind that many of these claims, while initially worrying given that many are made by those destined to become health care workers, are subjective. It is possible that alcohol is a convenient scapegoat. While speculative, it could be suggested that questionnaires which ask about drinking behaviour, and then about attendance and academic performance, may unwittingly provide a link which the poorly performing student is willing to follow. Other factors may influence academic performance. There is little evidence of objective assessment of the putative link between academic performance and alcohol intake. However, File et al. (1994) studied the drinking behaviour of 11 students from the first and second year medical courses who failed re-sit examinations. None belonged to the dangerous drinker group. From a very small population of medical students (n = 46), Hannay (1998) considered that, for female students, there was evidence of an association between drinking >14 units per week and poor examination grades. This relationship was not found with male students. Within the USA, Clark et al. (1987) have suggested, ironically, that in fact alcohol consumption may be associated with improved academic performance.

More research evidence is required before the question of whether or not alcohol consumption has a detrimental effect on academic performance can be answered. In addition, it is essential that the ‘pattern’ of consumption be considered. For example, binge drinking once a week on Friday might have very little detrimental effect on academic performance, but drinking 2–3 units regularly (yet still staying within ‘sensible’ weekly guidelines) at lunchtime before afternoon classes may not be without effect.

It is interesting that Newbury-Birch et al. (2000) from their study of students’ drinking behaviour prior to coming to university, reported that 42% of men and 36% of women had missed half a day or more of study as a result of alcohol intake. In addition, 48% of men and 43% of women ‘had become more sexually involved with someone than they would normally have wanted’ (p. 126), with 10% of men and 13% of women having failed to use contraception during intercourse.

Thus many of the behaviours associated with alcohol misuse highlighted in this section may, as suggested by Newbury-Birch et al. (2000), not be unique to university life.

Finally, an important point must be emphasized. While behavioural consequences have been highlighted, the immediate physiological consequences of hazardous drinking may be less obvious, but just as important. Adverse effects of binge drinking on blood homeostasis and its circadian variation (Numminen et al., 2000), cardiac rhythm (Fuenmayor and Fuenmayor, 1997), ischaemic heart disease (Kauhanen et al., 1997), blood pressure (Seppa and Sillanaukee, 1999), white blood cell activity (Vinson et al., 1998), female reproductive hormone levels (for review, see Gill, 2000) and the fetus (reviewed by Gladstone et al., 1996) have been reported. Young adult rats seem particularly prone to brain damage following exposure to binge ethanol levels (Crews et al., 2000).


As noted, there is tentative evidence from the data reviewed in Table 2 that the proportion of women in the undergraduate population exceeding ‘sensible’ weekly intakes of 14 units of alcohol may be as high as 40%. If correct, this figure is three times that recorded in the GHS study and these women appear more similar to their male cohort than do women in the general population.

For binge drinking, again there is a suggestion in the most recent surveys of particularly high levels of women students (more than one in two) indulging in this form of consumption. In fact two studies (Pickard et al., 2000; Underwood and Fox, 2000) recorded even higher levels of binge drinking for females than for males. In the same two studies female students recorded lower abstention rates than their male counterparts. It has also been noted that there was evidence of more women than men being classified as heavy drinkers in two studies while four studies reported female heavy drinking levels exceeding that quoted in the GHS.

While it must be acknowledged that various criticisms of study methodology are warranted, these findings are sufficiently important to merit attention; further research is required. The contention made over 25 years ago (Orford et al., 1974) that female drinking behaviour is constrained by the company of males may no longer hold true. Two studies referred to above which reported particularly high levels of binge drinking in females (Pickard et al., 2000; Underwood and Fox, 2000) involved medical and dental students. Mangus et al. (1998), from a study of fourth year medical students at eight US medical schools, noted that 18% of women and 11% of men (P = 0.02) claimed that their intake of alcohol had increased during their medical training. This, the authors maintained, is consistent with the view of Flaherty and Richman (1991) that, among medical students, a ‘gender convergence’ of drinking behaviour is evident. Increased alcohol consumption may be a side-effect of medical training, especially for women.

Wechsler et al. (1995b) suggested that women who binge drink are less willing to recognize their alcohol problem. In a large study of student drinking habits (n = 17592), 22% of men classified as binge drinkers described themselves as heavy or problem drinkers; but only 8% of women did so. This proposal needs to be investigated within a UK undergraduate population, as it has important implications for health promotion initiatives. The vulnerability of female students to the consequences of alcohol misuse have been stressed by Collier and Beales (1989), who suggested that female students showed a disparity between their awareness of drinking problems on the one hand, and their own behaviour on the other.

Anderson et al. (1984) reported that 45% of women living outside college drank in their own or someone else's room. For those living in college, the figure was 38%. As previously noted, several factors may combine to suggest that female consumption levels reported by the present studies are underestimated.


How likely is it that the study findings reviewed here present an accurate picture of UK undergraduate drinking behaviour at the end of the millennium? Preliminary comparison of the data with one UK national population study (GHS), would suggest that more male students exceed sensible weekly drinking guidelines and, curiously, are classed as abstainers, than their age cohort as a whole. As many as one in two male students may exceed sensible weekly guidelines and take part in binge drinking. However, we must first investigate the disparity between study results. (This is particularly true for binge drinking where future studies must clearly define the time frame of measurement.) The possibility that as many as one in ten male healthcare students destined to influence and administer sensible drinking advice is regularly drinking >50 units per week merits attention. The evidence that 34% of male medical students exceeded their own sensible guidelines (Collier and Beales, 1989) and that 22% drank alone ‘regularly’ or ‘sometimes’ (West et al., 1990) reinforces this concern.

It is interesting to note that the investigation of pre-university drinking behaviour reported by Newbury-Birch et al. (2000) recorded one of the lowest binge drinking values for males (27%) but curiously the highest percentage of heavy drinkers (16%). Does the pattern of drinking, but not the amount, change during the transition from school to university?

Several aspects of female student drinking behaviour have been highlighted. Three times as many female students exceed sensible guidelines as women in the general population. For binge drinking, as for males, there is an urgent need to clarify the disparity between several study reports. A report of particularly high levels of binge drinking (almost three times that of females in the general population) demands investigation (Pickard et al., 2000). Students may be drinking more like their male colleagues than females in the general population.

Various initiatives within the UK led to the establishment and promotion of sensible drinking guidelines. The evidence from studies reviewed here would suggest that, for a large number of young adults, this message in its present form is to a large extent unheeded. Furthermore binge drinking for many within this age group may be viewed as a normal pattern of alcohol consumption.


View Abstract