Illicit drug use in university students in the UK and Ireland: a PRISMA-guided scoping review

The search produced 1583 potential papers for inclusion. The inclusion and exclusion criteria were formulated as an iterative process once the breadth of the literature was understood. Systematic reviews and literature reviews were excluded but reference lists were searched. Primary studies were not excluded based on their design, and both quantitative and qualitative research was included. Theses and student dissertations were included, but other unpublished literature was assessed separately. Studies that also included non-student participants or non-UK/Ireland-based universities were only included if the results were separated by population and the relevant data could be extracted. Papers discussing drug education in the university curriculum of healthcare professionals were excluded. The full search included papers that focussed on the use of drugs prescribed by a healthcare professional (even if used illicitly e.g. those used as ‘cognitive enhancers’). However, this report focuses on illicit drugs only (see data supplement 2 for excluded papers on the latter topic). If more than one report used duplicate data, the most comprehensive or relevant paper was included.

Two independent reviewers (MB and ED) undertook the study selection process. The titles of the records found in the search were screened and the relevant abstracts independently assessed, with any disagreements between reviewers resolved through discussion. Full-text papers were then obtained and reviewed. The reviewers met frequently to discuss challenges surrounding study selection and to ensure the search strategy was suitable. Papers that assessed any aspect of the use of illicit drugs conducted in the UK or Ireland were included. Figure 1 shows the process of study selection using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) framework [24]. The excluded papers are listed in Supplementary file 2, along with the reasons for exclusion.

An iterative model was used to determine the study characteristics extracted [20]. Both researchers independently used a data-charting tool to extract study characteristics from the first five papers included, before meeting to discuss any difficulties and refine the variables to extract. The final characteristics recorded were [1] author, [2] year of publication, [3] journal, [4] location of data collection, [5] study design, [6] delivery method (e.g., online survey, in-person, postal survey), [7] participant number, [8] response rate, [9] participant course of study, [10] participant year of study, [11] participant degree level (i.e., undergraduate, postgraduate), [12] participant gender, [13] participant age, [14] time-period assessed (e.g., lifetime, current use, past 12 months), [15] primary aim [16] primary outcome (categorized into 6 groups: prevalence and patterns of drug use, risk and protective factors, consequences of drug use, attitudes and knowledge, motivations for drug use, source of drugs), [17] ethical approval, and [18] funding source. One researcher (MB) then extracted the 18 study characteristics from each included study and the other researcher independently reviewed the completed data-charting form, with any disputes resolved with discussion. Quality was not formally assessed but ethical approval and funding were used as crude proxies in line with scoping review guidance [25].

A review of the grey literature found eleven relevant papers, all of which focussed on the prevalence of drug use among university students. One was a survey conducted by the UK National Union of Students and a national charity providing expertise on drugs and the law (Release) [11], and a second was commissioned by the UK Higher Education Policy Institute from a professional surveying organisation (YouthSight) [26]. A third reported on a national student survey in Ireland, and the remainder were student newspaper reports. The largest survey was carried out by The Tab, an online magazine covering youth culture and student issues. This had 16,000 responses from students across the UK, but no information was provided on the sampling methodology used. None of these papers reported how they recruited participants or their demographic characteristics, and so these studies are not considered further in this review but are detailed in Supplementary file 2.

Table 1 summarises the study characteristics. Since the 1980s the number of published papers has increased with each full decade, and almost half of the papers included in this review were published within the last 10 years. All four nations of the UK (England, Wales, Scotland, Northern Ireland) and the Republic of Ireland were represented. Five papers did not report their location within the UK, and only seven papers included universities in more than one country. English (44%) and Irish (19%) universities were the best represented. The majority (63%) of papers recruited participants from just one university, and only 3 (6%) recruited participants from 10 or more universities.

The subject that participants were studying varied, with 16 (30%) papers limiting the sample to a particular course or courses (usually medical, dental or nursing studies). The academic year of study of participants was recorded in 41 (76%) of the papers, and the choice of year group was usually decided pragmatically with no particular focus on students early or late in their course of study. The gender breakdown in the 43 (80%) papers that reported it was between 30 and 89% female. The mean age of the sample was reported in 21 (39%) papers and ranged between 18.8 and 24.9 years.

A clear majority of the studies had a cross-sectional design, with only 4 reporting data from more than one time point. Forty-seven papers (87%) reported a quantitative analysis, 4 (7%) reported a qualitative analysis and 3 (6%) used mixed methods. In those that employed mixed methods, cross-sectional data was used to inform later semi-structured interviews and open questionnaires. In the 51 papers where numerical data were collected, 36 (67%) used in-person interviews, 9 (17%) on-line data collection, 6 (11%) postal data collection and 1 analysed secondary data.

Under half (23, 43%) of the papers reported that they had obtained ethical approval and the remainder did not report whether approval had been sought or not. Twelve of the papers reported funding from various sources including university funds, the British Medical Association, the Wellcome Trust and the Northern Regional Health Authority. The remainder either did not report whether funding was received or specified that there was no external funding for the project.

The drugs assessed in the papers were grouped into categories: a broad definition of illicit drugs (sometimes including prescription drugs used in an illicit way) (43, 80% of papers), cannabis only (8, 15%), and ecstasy only (3, 6%). Most papers included questions about drugs alongside alcohol and tobacco. Cannabis was the most reported drug under study, but the range of substances reported increased over time.

The outcomes of the papers were coded into five categories, with papers that reported more than one outcome coded multiple times. Forty-one of the fifty-four papers (76%) reported the prevalence or pattern of drug use, 28 (52%) factors associated with drug use, 14 (26%) student attitudes towards, knowledge about, and motivations to use drugs, 6 (11%) information about the source and supply of drugs, and 7 (13%) the consequences of drug use. The included papers are categorised by outcome in Table 2.

The 41 papers examining prevalence or patterns of illicit drug use collected data through either an in-person interview (29, 71%), a postal response (4, 10%), an online survey (7, 17%) or both in-person and online methods (1, 2%). The number of participants ranged from 47 to 7855, and 39 of the 46 (85%) samples reported included a percentage response rate ranging from 6 to 100%. The response rate varied by the method used to collect data (in-person 60–100%, postal 33–97%, online 6–33%, in-person and online 41%). In terms of participants recruited the online surveys had the most participants (mean of 3382 compared with mean of 765 in the in-person surveys and 430 in the postal surveys) but the lowest response rates (mean of 15.5% compared with 82.2% in-person and 68.5% postal). A range of time periods of drug use were assessed, including lifetime (35 papers), past year [13], past 6 months [2], past 3 months [1], past 3 months [1], past 30 days [1], past month [1], past 4 weeks [1], past week [3], ‘since starting degree’ [1], ‘current academic year’ [1] and ‘current’ [12]. There was a broad trend towards students reporting more experience of a wider range of illicit substances. However, the variability in participant samples, the methods used to collect the data, and the time periods of drug use considered meant that it was not possible to formally assess trends in drug use over time.

Twenty-eight papers (52%) assessed factors associated with drug use, and these are summarised in Table 3. Many papers reported more than one associated factor. Twelve papers (22%) explored demographic variables, including age, gender, ethnicity, socioeconomic status, living circumstances and international student status. Eleven (20%) measured personality factors or mental health, including instruments measuring sensation seeking and anxiety. The link between health-related behaviours, including tobacco smoking, alcohol consumption and physical activity, and drug use was assessed in 11 (20%) papers. Other associations included academic course or year of study (8, 15%), attitudes to drug use and health awareness (3, 6%), normative beliefs (2, 4%), academic performance (2, 4%) and religious beliefs (2, 4%). No clear and consistent patterns of association could be drawn from the data.

Fourteen papers assessed student attitudes towards drugs and/or their knowledge about their effects. Issues covered included the morality or ethics of drug use, safety beliefs, the perceived effect of drug use and perceived motivations for use. Early studies noted that attitudes to cannabis were markedly different to those towards tobacco or alcohol [27]. People that drank alcohol perceived that people that used cannabis were ‘definitely emotionally unstable’ and ‘definitely less able to cope with life’, whereas the latter group perceived that people that used illicit drugs in general were ‘more interested’ and ‘vested with more friends’. This attitudinal separation was hypothesised to be an effective ‘barrier’ to starting cannabis use.

By the late 1990s, papers were reporting that students that regularly used illicit drugs were similar to the general population of students both in their views about the causes of drug use and their personal and social characteristics [28]. Both students that used drugs and those that did not agreed that youth culture influences and sensation-seeking were the most endorsed reasons for drug taking. The illegality of drugs had little influence on levels of consumption [29]. Some support was found to support the idea that increasingly liberal views towards drugs would appear across a student’s time at university [30]. Although attitudes towards tobacco became less positive in year 3 when compared to starting university, there was no such change for illicit drugs. By the 2010s, students were shown to rate tobacco as most harmful to physical health, alcohol most harmful with regard to injuries and social consequences, and cannabis as most harmful with regard to mental health [31]. As the legal substance alcohol was rated as more harmful than the illegal substance cannabis, the authors hypothesised that young people in the years to come may be less supportive of a traditional drug policy based on criminalization [31].

By 2018, Patton’s survey found that the top three reasons for drug consumption were for fun or pleasure, for relaxation, and to enhance an activity [16]. These reasons were thought to fit with the ‘normalization’ hypothesis [32]. Depictions of drug use in the media were widespread, and 78% felt comfortable consuming media that featured drug use. There was also further evidence of the shift from drug use as a deviant activity into mainstream cultural arrangements (59% of abstainers had one or more close friends who use drugs) [16]. These results suggest that attitudes towards drugs may have changed over time amongst the student population, but the level of acceptance is not uniform or consistent between different substances in different populations.

Six papers were concerned with the supply of drugs. An early study in the 1970s in Ireland found that most students were approached to buy drugs at parties, in pubs or hotels, or at clubs. Roughly half of students obtained their first drugs from friends [33]. A study in England in the 1990s also found that drugs were usually bought from friends and were most commonly consumed in other people’s rooms or at parties [29]. The authors contrasted this with alcohol which was consumed in bars or public places. More recently, a larger survey of 7 of the 9 universities in Wales found that half of the students that used drugs obtained them solely from friends and associates, and another 25% used friends and external markets [34]. In many cases supplying drugs amounted to sharing them or giving them away, but over a third said they had sold drugs. Drugs like nitrous oxide, cannabis, synthetic cannabinoids, ecstasy and magic mushrooms were usually sourced from friends, whereas other drugs (khat, crack, steroids, heroin) more likely to be bought. Male students were more likely to buy from dealers. The authors concluded that methods used by university students to obtain and supply drugs shared features of both ‘social supply’ and ‘traditional drug markets’ [34].

Moyle and Coomber also considered the nature of the supply of drugs in students. They conducted semi-structured interviews with 60 social suppliers of recreational drugs in two studies involving both a student population (n = 30) and a general population sample (n = 30) [9, 35]. Both samples provided evidence that supplying drugs to, and receiving them from, friends and social contacts had become increasingly normalised and seen as less than ‘real dealing’ and more like gift-giving [35]. Early experiences of social supply occurred pre-university and usually involved a ‘one off’ act of sharing cannabis. However, once at university this had increased to ‘buying cannabis in bulk and selling excess amounts to friends, and/or purchasing ‘standard’ 3.5 g bags of powders like cocaine and MDMA on behalf of a group and retaining a quantity of the substance as payment’. This behaviour rarely continued when they returned home [9].

Seven papers reported the consequences of drug use. Two studies surveyed consequences of any illicit drug [36, 37], four focussed on cannabis [38‐41] and one on ecstasy [42]. Two large online surveys collected data on physical and psychological effects of drugs and drug-related crime, providing a broad overview of a range of issues. In contrast studies utilising in-depth structured interviews with qualitative analysis explored the positive and negative effects of cannabis [39] and the effects of cannabis on driving [41]. One lab-based study examined cannabis-related impairments in prospective memory by comparing people that used cannabis and people that did not on both self-reported prospective memory failures and on an objective video-based prospective memory task [38]. Two studies used objective measures to quantify the consequences of drug use on mental health e.g., the Hospital Anxiety and Depression scale [40, 42]. Overall, the positive effects of relaxation, mood elevation and enhanced creativity were balanced by negative effects such as forgetfulness, poor concentration, and reduced productivity. Impaired mental health was a common theme, including paranoia, moodiness, anxiety, irritability, confusion and dependence. Crime-related consequences included driving under the influence, antisocial behaviour and selling drugs [36, 41].

Monitoring student drug (and alcohol) use over time should be critical to the development of effective evidence-based policy and intervention strategies. High quality estimates can be used to identify trends and patterns, understand the direct and indirect harms of drug use, and guide further research to understand risk and protective factors for student drug use and the effectiveness of policy or treatment interventions. This review found that the prevalence of drug use was the most studied area in terms of number of published papers. However, as has been noted by other researchers [16], the existing UK/Irish research is methodologically limited. Most published studies reported prevalence of use, but the time window of assessment and the instruments used were variable. Studies including students from more than one university were the exception, and the population under study was often drawn from a single university department. It was rare for a paper to distinguish between single use (e.g., tried once in a lifetime) and regular use (e.g., several days in the past week), and validated clinical diagnoses were never reported. Most of the published papers described cross-sectional studies, with no attempt to follow up participants. One research group based in the north-east of England has repeated studies over time [43‐51], but not used the same methodology or followed through one cohort for significant length of time. Little of the research moved beyond simple descriptions and correlations, and it was rare that any under-pinning theory or conceptual model was described. Finally, there was very little evidence about drug use in specific high-risk populations e.g., LGBTQ or non-white students, or students with co-existing mental health problems.

The Advisory Council on the Misuse of Drugs (ACMD) has noted that prevalence data on young people’s drug use within the UK are ‘generally limited, highly variable and of low quality’ [52]. They recommend reviewing the scope and detail of the current approaches to monitor prevalence, ensuring that outcome measures used are fit for purpose. Expertise and experience in this area can be drawn from North America, where regular national surveys using sound methodology include university-age students [12]. The authors of a Canadian report on quality standards for measuring drug use in high schools [53] noted that surveying students in school is an efficient and cost-effective means of collecting data from young people, and similar reasoning could be applied to universities. However, there is a need for standardised methodology that can be replicated over time in representative samples.

Recommendation: A national survey of student drug use that covers the whole of the UK and Northern Ireland would be helpful, particularly if it was repeated to monitor trends over time. Standardised data collection instruments tailored to young adults must be developed and tested (e.g. see [53]), and the first step towards this approach has already been taken in Ireland under government direction [54]. Alternatively, the creation of representative student ‘consumer panels’ may allow researchers to explore patterns and frequencies of drug use in nationally representative samples.

Although several studies described the association of a range of demographic, psychological and social factors with drug use, methodological limitations limited the utility of these findings. Some important areas of potential study were completely absent in our findings. For example, a developing evidence base suggests that there is a positive association between adverse childhood experiences (ACEs) and the development of substance use disorder in adolescence and adulthood [55]. One potential framework for study is the life course model of substance use [56], with a focus on the role of illicit drug use in developmental role transitions. Many authors have argued that emerging adulthood (i.e. the traditional university years) is developmentally different from adolescence (school) or full adulthood [10]. Independence from parents, new social and romantic relationships with peers, increased access to drugs and alcohol, and the need for self-directed study all contribute to a unique social milieu at a developmental stage already characterised by peak levels of risk-taking and high levels of mental health problems [57‐59]. Understanding the factors that increase or decrease substance use at university is important to develop effective responses.

Recommendation: Prospective cohort study designs are needed that include student populations to identify which factors play a role in the initiation, development and cessation of drug use. Risk and protective factors for illicit drug use in young adults may be conceptualised as contextual (e.g. availability of the drug, or social norms that are tolerant of illicit drug use), fixed markers of risk (e.g., sex, parental and sibling substance use, poverty or potential genetic factors), and individual and interpersonal risk factors (e.g., novelty and sensation seeking, conduct disorder in childhood, parenting styles, or poor quality of parent–child interaction) [2, 60]. Affiliation with peers that use drugs is one of the strongest predictors of illicit drug use in young adults and of crucial importance in the transitional social milieu of a university campus [61]. The impact of an increasing awareness of neurodiversity in young adults should also be investigated.

The consequences of illicit drug use were only reported in seven of the papers included in this review, with some limited focus on mental health and criminal justice issues. There were no studies exploring the impact on university-specific outcomes such as completion of a course of study, academic achievement and progression to further study or employment. This is surprising, as these outcomes are important markers of university quality used in national league tables. Understanding the specific harms that relate to university students will help to tailor prevention and treatment responses to this population.

Recommendation: The potential harms of illicit drug use in student populations occur across several domains, including academic performance (attendance and grades), other high-risk behaviours (unprotected sex, violence, driving under the influence), exacerbation of mental health problems, or legal issues (prosecution for possession or dealing) [62]. These consequences could potentially reshape the entire trajectory of the student’s life course [63]. Therefore, it would be useful to collect, collate and track standardised data nationally from a range of university departments (e.g., student welfare, registry) as a marker of illicit drug-related harm. Such quantitative data could usefully be supplemented by detailed qualitative studies of each aspect of harm (e.g., academic, physical, mental, social, or legal).

Levels of objective knowledge about drug use were rarely studied, and yet may form the bedrock of a harm reduction approach [64]. Likewise attitudes and motivations to use drugs were not often reported or studied, despite the existence of potentially useful underpinning theories such as the theory of planned behaviour [65]. The use of conceptual models to guide findings is important to build an effective evidence base for prevention and intervention. The 2021 UK Government Drug Strategy aims to achieve a ‘generational shift in the demand for drugs’, and one proposed strategy for achieving this is ‘research and testing messaging through an evidence-based, targeted behaviour change initiative, initially aimed at students in further and higher education’ [66] (p49). The document further notes that communications campaigns work best when they are tailored and targeted to the audience, and a recent ACMD report on prevention of illicit drug use notes that some activities have been ineffective, such as fear arousal approaches (including ‘scared straight’ approaches) or stand-alone mass media campaigns [67].

Our results suggest that a national trend towards ‘normalisation’ of recreational drug use has been replicated on university campuses. Furthermore, this has merged with the social supply of drugs, leading to a perception amongst some students that supplying (sometimes) large amounts of drugs is routine [35]. However, the overall picture is complicated, and not all studies included students who didn’t use drugs and so their voice was often not heard. Survey work using student panels rather than open online questionnaires shows that some students believe that their university should take a tougher stance on drugs [26]. The ‘social norms’ approach is based on challenging misperceptions individuals hold about their peers. Research at eight Further Education Colleges in the UK reported a perceived norm of frequency of substance use that was higher than the reported norm, and the majority of respondents did not actively approve of tobacco, cannabis or other drug use [68]. This reflects similar findings in the university system in Canada [69]. The social norms approach may be a viable method of developing effective methods of behaviour change in UK students.

Our scoping review found little evidence of published work in UK/Irish universities on prevention of, or intervention to reduce, drug-related harms. Reviews of the North American literature on prevention and treatment have also noted a lack of published studies, but parent-based and in-person brief motivational interventions appear to be promising [18, 70]. Our review found no such interventions published in UK or Irish student populations. This may reflect the slow response of universities to consider drug use and tackle its potential harms, and the impact of stigma and illegality on students’ help-seeking attempts. A review of psychological interventions for prevention and treatment of mental health disorders in university students [71] was also limited by the poor quality of the literature and exclusion of non-published data. It noted considerable uncertainty about the best way to provide interventions for students, and relatively few trials adapted intervention delivery to student-specific concerns. It called for further work to better understand the mechanisms underlying students’ mental health problems, perhaps using transdiagnostic, stepped care approaches. Research on both mental health and drug use should involve students in the design of interventions to increase their acceptability to this population.

The prevention of alcohol-related harm has been well studied in university populations, and several existing interventions for student drinking share theoretical and methodological underpinnings with effective interventions in drug prevention and treatment in other populations (i.e., school-based prevention, adolescent and adult drug treatment) [6, 18]. These interventions could be adapted to target drug prevention on university campuses. As is the case in the wider community, a recovery orientated system of care is required, with a full continuum of care encompassing harm reduction through to abstinence [72]. The development of the first Collegiate Recovery Programs [21] to support abstinent students at Teesside University and the University of Birmingham [73] represents the first part of such a continuum. However, despite big strides in the development of campus-based mental health support in the past decade, the issue of addiction to drugs, alcohol or other behaviours has been largely ignored. Tackling this deficit will be important to ensure that students are able to maximise the potential benefits of a university education.

Recommendation: Several commentators have noted that the stigma of psychoactive drug use appears to be particularly prominent in universities in the UK and Ireland, with ‘zero tolerance’ approaches often limiting informed debate [64]. Working collaboratively across the Higher Education sector may be helpful in supporting universities to provide education and prevent harm whilst respecting the illegality of illicit drug use. There is also a need to develop interventions tailored to the unique needs of students who have developed a drug use disorder, and to evaluate abstinence-based recovery programs on campus.