Use and abuse of alcohol in UK university sport

4 October 2010


Dr Sarah Partington, Dr Elizabeth Partington, Professor Nick Heather, Professor Alan St. Clair Gibson, Fran Longstaff, Susan Allsop, Helen Wareham, Mark Jankowski.

Northumbria University.


Previous evidence on the relationship between sport participation and alcohol consumption among students is ambiguous, with some studies reporting that students who take part in university sport drink less than their peers and other studies that they drink more. There has been a suggestion that involvement in sport can be a protection against hazardous drinking among students, but further information is required before this can be fully endorsed.

To investigate these issues, a survey was carried out in a purposive sample of 826 students from 7 HE and 2 FE institutions, with a range of geographical locations in England. Institutions also differed in their commitment to sport. Although the sample was not intended to be representative of the student population in the UK, it was roughly comparable in age and gender breakdown but contained a slightly higher proportion of students from a Black ethnic background.

Students were selected from a range of degree courses and completed a survey questionnaire at the beginning or end of lectures. The instrument used to measure alcohol consumption and related behaviour was the Alcohol Use Disorders Identification Test (AUDIT). In addition to socio-demographic information, other instruments were included in the questionnaire to throw light on the nature of the relationship between sport participation and alcohol consumption. These instruments were the Athletic Identity Measurement Scale (AIMS), the Readiness to Change Questionnaire (RCQ) and the Drinking Expectancy Questionnaire (DEQ).


Mean score on the AUDIT in the overall sample was 9.5. This is higher than the cut-point for the designation of an alcohol use disorder, which is 8 for males and 7 for females. 61% of the sample was classified as having an alcohol use disorder by the AUDIT, rising to 68% when students who had never drunk or tasted alcohol, mainly for religious reasons, were excluded from the analysis. According to the AUDIT, 39% of the overall sample was classified as “low-risk drinkers”, 41% as “hazardous drinkers”, 11% as “harmful drinkers” and 9% as showing “probable dependence”.

In terms of how often students drank, the largest proportion (39%) reported doing so 2-3 times per week. On a typical drinking occasion, 26% said they drank only 1-2 drinks but 21% drank 7-9 drinks and 20% drank 10 drinks or more. Over a third of students (35%) engaged in “binge drinking” (6 or more drinks on one occasion) on a weekly basis. Although comparisons with previous British studies are difficult, this suggests heavier drinking among students now than in the past. It is not clear whether students in England drink more than their age-group peers in the general population.

There were large and highly significant differences in AUDIT scores between the institutions taking part in the survey, with a range of means from 4 to 14. In addition to this there was no significant difference between the mean AUDIT scores of men and women and no difference between men and women in the proportions falling into AUDIT risk categories, although men drank significantly larger quantities on single occasions. Nevertheless, this evidence suggests that the “gender convergence hypothesis” regarding alcohol consumption applies to students in England.

Compared with students who did not play university sport, those who did showed a significantly higher mean AUDIT score, and a higher proportion were classified as having an alcohol use disorder. They also reported drinking more frequently and in larger typical quantities, and binge-drank more often. Contrary to expectations, there were no significant differences in AUDIT scores according to the competitive level of sport played. Students who took part in team sports showed a higher mean AUDIT score than those who took part in individual sports, with the highest scores associated with traditional team sports involving 11 or more per side (e.g., rugby, football and hockey). 84.5% of students who played team sports were classified as having an alcohol use disorder.

1st year students showed a significantly higher mean AUDIT score than 3rd year students, who in turn showed a significantly higher mean score than 2nd year students. When “never drinkers” were excluded from the analysis, there was no longer a significant difference between 2nd and 3rd year students.

Mean AUDIT for students on sport-based courses was significantly higher than for those on non-sports-based courses, a difference that was maintained when “never drinkers” were excluded from the analysis. Students living on-campus had a higher mean AUDIT score than those living off-campus, who in turn had a higher mean score than those living with family. In the overall sample, there was a modest but significant negative correlation between AUDIT total score and student age, with older students tending to show lower AUDIT scores.

In an initial logistic regression analysis, sport participation was not an independent predictor of an alcohol use disorder when other background variables were taken into account. The strongest predictors of an alcohol use disorder were the institution attended, age and term-time living arrangements. Following further logistic regression analysis, significant predictors of sport participation were the institution attended, age, type of degree course, year of study and term-time living arrangements.

Students playing sports were significantly more likely to identify with the athlete role as measured by AIMS than students not playing sports, but the correlation between AIMS and AUDIT total score was low and not significant.

The sample as a whole was low on readiness to change drinking behaviour as measured by the RCQ and students playing sport were not more ready to change than those not playing sport. Over half of students in the hazardous drinking category (59%) were in the Precontemplation stage and therefore apparently not concerned about their drinking. Even at the highest level of risk (probable dependence), nearly 30% were unconcerned about their drinking and over 50% were still contemplating change. Among those classified as having an alcohol use disorder, 51% were in the Precontemplation stage, 27% were in the Contemplation stage and 22% were in the Action stage.

In terms of alcohol outcome expectancies, as measured by the DEQ, in the overall sample the highest subscale scores were for Sexual Enhancement, Assertion and Tension Reduction. Correlations between AUDIT total scores and DEQ total and subscale scores were significant, with large effect sizes for the DEQ total score and for the Assertion and Dependence subscales. However, there was very little difference between sport and non-sport participants on subscales or total score from the DEQ.

When degrees of misperception of drinking norms were calculated, misperceptions of quantity of drinking and frequency of drunkenness tended to decrease with age. There were low but significant correlations between AUDIT total score and degree of misperception of frequency of drinking and for quantity of drinking but these were probably due to chance.


Alcohol consumption and alcohol use disorders in this sample were very high. Assuming that the figures recorded here reflect drinking behaviour in the student population in England, this is a cause for grave concern. Central government, local authorities and university and college authorities should take urgent measures to restrict the easy availability of cheap alcohol to students.

The lack of difference between genders in proportions showing alcohol use disorders is especially alarming in view of women’s greater vulnerability to the damaging effects of heavy drinking. Particular attention should be paid to measures aimed at reducing alcohol consumption among female students.

A particular risk factor for the development of alcohol use disorders seems to be the 1st year of student life combined with on-campus accommodation. University and college authorities should develop policies aimed at reducing drinking among 1st year students living in halls of residence.

There is an urgent need for the wide implementation of brief interventions, treatment and treatment referral mechanisms among students with alcohol use disorders and who are ready to change drinking behaviour. Internet-based brief interventions should be targeted at students drinking at hazardous levels, face-to-face brief motivational counselling at students showing harmful drinking and the offer of treatment on-campus or referral to treatment off-campus for those showing probable dependence.

In the present sample, students playing sport showed clearly higher levels of alcohol consumption and alcohol use disorders than those not playing sport and there was no evidence to support the suggestion that sport participation by students can protect against hazardous drinking. Instead, sport participation may be a risk factor for the development of alcohol use disorders. Research is needed to find effective ways of breaking the link between sport participation and heavy drinking by students.

A hypothesis based on the present findings is that heavier drinking among students who take part in sport is not the result of sport participation per se but is related to other demographic and background variables associated with both heavier drinking and sport participation (e.g. attendance at heavier drinking universities, younger age and on-campus term-time accommodation). This hypothesis should be tested in further research.