Prevalence of alcohol use amongst sexual and gender minority (LGBTQ+) communities in the UK: a systematic scoping review

Abstract

Background: globally, alcohol use rates vary by sexual orientation and gender identity (SOGI), but UK government statistics on alcohol use in the LGBTQ+ population are missing. Aim: this systematic scoping review determined the prevalence of alcohol use amongst gender and sexual minority people in the UK. Methods: empirical UK studies from 2010 onwards reporting the prevalence of alcohol use in SOGI compared with heterosexual/cisgender people were included. Searches in MEDLINE, Embase, Web of Science, PsycINFO, CINAHL, Cochrane Library, Google Scholar, Google, charity websites and systematic reviews were conducted in October 2021, using SOGI, alcohol and prevalence terms. Citation checking was done by two authors, with disagreements resolved through discussion. Data extraction was done by one author (CM) and checked by another (LZ). Quality assessment was performed by study design, sample type and statistical analysis of results. A narrative synthesis was qualitatively combined with a tabular presentation of results. Results: database and website searches found 6607 potentially relevant citations, and 505 full texts were reviewed with 20 studies included, found in 21 publications and grey literature reports. Most were on sexual orientation, including 12 from large cohort studies. Harmful alcohol use is higher in LGBTQ+ people than heterosexual people in the UK, a result similar to that found in other countries. Qualitative data reflected alcohol’s role as emotional support. Fewer asexual people drank alcohol compared with allosexual people, and there were no data available regarding intersex people. Conclusion: funded cohort studies and service providers should routinely collect SOGI data. Standardized reporting of SOGI and alcohol use would improve comparability across studies.

alcohol, hazardous alcohol use, harmful alcohol use, LGBT+, LGBTQ+, sexual and gender minority, prevalence, UK

Introduction

Hazardous alcohol drinking is a pattern of alcohol consumption that increases someone's risk of harm (MacKillop et al., 2022). The term ‘harmful drinking’ is used when alcohol use causes mental or physical damage (NICE 2010). The World Health Organization (WHO) considers that the harmful use of alcohol is the third most prevalent risk factor for poor health across the world (WHO 2010). Globally, the rate of premature death from alcohol use has declined from 43 deaths per 100 000 people in the early 1990s to 35 deaths in 2017 (Ritchie 2018).

People of minority sexual orientation (SO) can identify as lesbian, gay, bisexual or queer (LGBQ). People of minority gender identity (GI) can identify as trans, non-binary or queer. Non-binary is a term used by people where their gender doesn’t fit within the margins of a male/female gender binary and, instead, gender is understood as moving beyond identifying as either a man or a woman. Queer can be defined as a person who seeks to deconstruct, or re-imagine, dominant binaries of gender, sex and sexualities by questioning or destabilizing normative ideals (Zeeman et al., 2014). Asexual people are those who have a lack of interest in sexual activity or desire (Bauer et al., 2020). Grey-asexual people are those intermediate between asexual and allosexual (i.e. not asexual) people and include those who rarely experience sexual attraction. Intersex variance can be understood as physical, hormonal or genetic features that people are born with, or develop later in life. Variations relate to a range of traits and/or sex characteristics that lie outside the binary, medical or social norms for male and female bodies (Zeeman and Aranda 2020). In this paper, LGBTQ+ is defined as lesbian, gay, bisexual, trans, queer and other people who do not identify as part of a heterosexual (not LGBQ), cisgender (not trans) and endosex (not intersex) majority.

Alcohol consumption varies between individuals, nations and countries influenced by factors such as local laws, traditions, religious persuasions, incomes, education and work status. Alcohol use also varies by sexual orientation and gender identity (SOGI). For example, in the USA, the rate of alcohol use in lesbian, gay or bisexual (LGB) populations aged 18–25 in 2018 was 64.3% compared with 55.1% in the general population, and that in those aged 26+ was 64.7% compared with 55.3% (SAMHSA 2018). Also, US trans college students were likely to consume alcohol over more days and have more total drinks than cisgendered college students (Tupler et al., 2017). In Australia, people identifying as LGB were more likely to drink at risky levels (AGIHW 2022) and were 1.5 times more likely to exceed lifetime risk guidelines than heterosexual people (25% compared with 16.9%). Trans women were more likely to exceed two drinks on a typical day (32%), as were gays (39%), when compared with the general Australian population (17.6%) (AGIHW 2022).

Results from a nationally representative survey from France also showed higher consumption of alcohol in non-heterosexual people compared with heterosexual people (Llomond 2014), and a nationally representative survey from Canada showed higher rates of heavy drinking in gay/lesbian and bisexual people compared with heterosexual Canadians (Pakula et al., 2016).

Many health professionals do not consider sexual orientation or gender identity to be important when delivering care (Somerville 2015, Berner et al., 2020). For some time, there has been more dissatisfaction with UK health services reported by LGBTQ+ people compared with their heterosexual and cisgender peers (Elliott et al., 2015, McDermott et al., 2021). The UK Government LGBT survey (2018) found that, in the preceding 12 months, 40% of trans respondents had had at least one negative experience of health care because of their gender identity and that 13% of cisgender respondents had had at least one negative experience of health care because of their sexual orientation (GEO 2018). Such dissatisfaction is also found within alcohol support services; for example, one recent report found that LGBT issues were not well understood, and although this varied from service to service and worker to worker, there was an impression that this was not necessarily due to prejudice but rather a lack of awareness or experience (Moncrieff 2014). The impact of this was felt and noticed, creating a barrier for LGBT clients (Moncrieff 2014). Sometimes the barriers were about homophobia or transphobia from other clients not being challenged by the staff. As the proportion of the UK population that is not heterosexual or cisgender is gradually rising (OECD 2019, ONS 2021a), in addition to more people coming out, these issues may become more common.

A systematic scoping review is a preliminary assessment of the potential size, nature and scope of available research evidence on a specific topic, and attempts to be systematic, transparent and replicable (Grant and Booth 2009). It can also summarise findings from a heterogeneous body of research or identify gaps to aid the planning and commissioning of future research (Tricco et al., 2018). A systematic scoping review is more useful than a systematic review, when the aim is to provide an overview or map of the evidence, rather than producing a critically appraised and synthesized result or answer to a specific question (Munn et al., 2018). As the main aim of this project was to determine the prevalence of alcohol use amongst sexual and gender minorities in the UK from heterogenous sources of research evidence, a systematic scoping review was considered to be the most appropriate review type.

Methods

The objectives for the systematic scoping review were to determine the current or recent prevalence of hazardous/harmful drinking or alcohol misuse amongst SOGI communities in the UK, and how alcohol use plays a specific role for some LGBTQ+ people. At the time of writing, no official government-backed publication was available on alcohol-related problems in the LGBTQ+ community in the UK, so this systematic scoping review aims to fill this gap. It is now needed because there are increasing numbers of primary studies available, but not easily found, which suggest higher levels of alcohol use in SOGI communities. It may be that mainstream alcohol support services are not addressing the needs of their SOGI clients.

A protocol for this systematic scoping review project is available on the Drinkware Trust’s website (Drinkaware 2022). Inclusion criteria were cross-sectional surveys or cohort studies, reported in published or grey literature from 2010 onwards, where harmful alcohol use at any time in SOGI people in the UK was compared with heterosexual/cisgender people. Studies could report any type of prevalence measure regarding alcohol use or problematic drinking or qualitative quotes about issues regarding problematic alcohol use.

Searches were undertaken in six electronic databases (MEDLINE, Embase, Web of Science, PsycINFO, CINAHL and the Cochrane Library) in October 2021. Google Scholar was searched, and Google investigated with the first 100 hits. Websites of known LGBT+ charities in the UK [LGBT Foundation, London Friend, Sigma Research, Stonewall (London, Scotland and Cardiff), GIRES, Brighton Switchboard, London Metro, Kaleidoscope Trust, Galop, Scottish Trans, MindOut] and mental health and alcohol charities (Drinkaware, Mind, Alcoholics Anonymous, Institute of Alcohol Studies) and the UK Office for National Statistics (ONS) website were searched for includable evidence. Relevant systematic reviews and references of included papers were checked to identify further relevant primary research articles. Hand-searching, informed by existing knowledge of the team, found information that may be harder to reach, for example, in grey literature reports or reports on government websites and the first author’s private library of LGBT+ research.

Search terms and appropriate synonyms (MeSH terms) included gender minorit^*, sexual minorit^*, LGBT+, LGBT, LGB, lesbian, gay, bisexual, trans, queer, intersex, asexual, non-binary AND alcohol^* AND prevalence, incidence, rate^*.

Database management software (EndNote) was used to store citations, identify included studies and detect duplicates. A master table was created containing key information from each included study. Quality assessment was by study design, whether the sample was a random, consecutive or convenience sample, and whether there was any statistical analysis of the results. We initially attempted to use the Newcastle-Ottawa Scale adapted for cross-sectional surveys (Herzog et al., 2013), but a formal quality assessment checklist was not used because, when tested, it did not give useful information, and because systematic scoping reviews do not necessarily need a formal critical appraisal of included studies (Tricco et al., 2018). In general, better quality data are assumed to be derived from a population rather than convenience samples, having an appropriate concurrent comparator and having a full statistical analysis of results, including 95% confidence intervals. Data extraction and quality assessment were undertaken by one author (CM) and checked by another (LZ).

Narrative synthesis

Synthesis was principally qualitative by using tables with related text to present results. Meta-analysis was not appropriate as studies reported different subsections of the sexual orientation population (LGBT+, LGBT, LGB, or gay and bisexual men, or lesbians and bisexual women, or lesbian and gay women and men, or bisexual women and men), and because a wide variety of categorical and continuous alcohol outcomes were reported.

Results

Searches found 6607 citations (6394 database records, 37 from websites, 7 from citation searching and 169 from the private library maintained by the first author). Reviewed for eligibility were 505 full-text publications and reports, and 20 studies were included, available in 21 texts [12 publications; (Bloomfield et al., 2011, Chakraborty et al., 2011, Hagger-Johnson et al., 2013, Pesola et al., 2014, Woodhead et al., 2016, Booker et al., 2017, Shahab et al., 2017, Rimes et al., 2019, Amos et al., 2020, Bauer et al., 2020, Becares et al., 2021, Pitman et al., 2021), and 9 grey literature reports (Balding 2014, 2018, Guasp 2011 (Stonewall LGB People in Later Life), Guasp 2013 (Stonewall GB Men’s Health Survey), ONS Health Survey for England 2021b, Nodin et al., 2015 (RaRe Report), What About Youth? (PHE 2014), Whybrow et al., 2012 (Scottish Health Survey), Youth Chances 2014]. See Fig. 1 for a summary PRISMA Flow diagram.

Figure 1

PRISMA flow diagram.

A wide variety of alcohol measures were used in the included studies, including categorical measures such as consuming more than a certain number of units, being a regular drinker and scoring above a certain AUDIT score, and continuous measures such as the mean number of units drunk and mean AUDIT score.

Four of the studies reported results from population surveys from the UK (Bauer et al., 2020), England (ONS Health Survey for England 2021b), What About Youth? (PHE 2014) and Scotland (Whybrow et al., 2012, Scottish Health Survey). Results from these are shown in Table 1. The remaining studies provided results from large cohort studies,

Table 1

Population surveys. Sexual orientation

Report	Outcome	Total N	LGB	Lesbian/gay	Bisexual	Asexual	Grey Asexual	Straight	Allo-sexual
Bauer et al., (2020)	OR of being a non-drinker of alcohol	N = 14 826				0.13 (0.09–0.22)	0.41 (0.35–0.48)		1 (reference case)
Health Survey for England (2021b)	More than 14, up to 35/50 units per week (increased risk)	N = 56 794	25% (n = 1132)	-	-	-		20% (n = 55 662)
	More than 35/50 units per week (higher risk)		7%	-	-	-		4%
	Mean number of units		17.7	-	-	-		12.7
Whybrow et al., (2012) (Scottish Health Survey)	Drink at hazardous or harmful levels	Age standardized	-	34%	29%	-		23%
	Exceed daily limits		-	50% (n = 195)	49% (n = 227)	-		39% (n = 23 811)
What About Youth (via PHE Fingertips (2014)	Drunk in the last 4 weeks	N = 120 115	-	24.6%	26.5%	-		14.3%
	Regular drinker (in young people only)		-	10.7%	12.4%	-		6.0%

Sexual orientation

Population survey data on the prevalence in LGB people compared with heterosexual people was available in 3 grey literature reports [(Whybrow et al., 2012, ONS Health Survey for England 2021b) (Scottish Health Survey), What About Youth? (PHE 2014)] and 10 journal articles (Bloomfield et al., 2011, Chakraborty et al., 2011, Hagger-Johnson et al., 2013, Pesola et al., 2014, Woodhead et al., 2016, Booker et al., 2017, Shahab et al., 2017, Amos et al., 2020, Becares et al., 2021, Pitman et al., 2021).

The Health Survey for England results from 2021 (Table 1) showed higher rates of alcohol use in LGB (not split by gender or sexuality) compared with heterosexual people. It cannot be determined whether these rates were statistically significantly higher in men compared with women, or in lesbian/gay compared with bisexual people. The Scottish Health Survey from 2012 (Whybrow et al., 2012) split results by lesbian/gay and bisexual, but not by gender. Again, this survey showed higher rates in LGB people than in heterosexual people. Unlike the ONS Health Survey for England, the Scottish Health Survey gave 95% confidence intervals around the prevalence estimates (gay or lesbian (n = 194) 34% (95%CI 25–43%), bisexual (n = 229) 29% (95%CI 21–37%), heterosexual (n = 23 851) 23 (95%CI 23–24%). The 95% confidence intervals suggested that their hazardous or harmful drinking rates were statistically significantly higher in gay/lesbian people than heterosexual people, but not in bisexual people. What About Youth? (PHE 2014) was a survey contracted by the Health and Social Care Information Centre on behalf of the UK Department of Health, and conducted by Ipsos Mori. Participants were 15 years old and sampled from the Department for Education’s National Pupil Database, of which 120 115 responded with usable data, giving an unadjusted response rate of 40%. Sexual orientation was asked using the ONS-validated questions. The results by sexual orientation have been hosted on the Public Health England Fingertips site since 2015 (accessed via the Child and Maternal Health part) but are not in the What About Youth main report (Niblett 2015) or spreadsheets hosted by NHS Digital (NHS Digital 2014). The results for alcohol showed that the percentage of regular drinkers and the percentage who have been drunk in the last 4 weeks was higher in gay/lesbian (24.6%) and in bisexual groups (26.5%) compared with the heterosexual group (14.3%).

Included peer-reviewed publications (Table 2) all showed higher alcohol misuse for the LGB groups compared with the heterosexual groups. Amos et al., (2020) investigated adolescents (14 years) in the Millennium Cohort study, and, statistically, significantly more in the sexual minority group ever tried alcohol compared with the heterosexual group. Becares et al., (2021) investigated lesbians and bisexual women in the UK Household Longitudinal Study, and found a statistically significantly higher rate of scoring 3 or more, as measured by the AUDIT-C questionnaire, compared with heterosexual women, which was considered to be harmful. Bloomfield et al., (2011) was a much smaller UK sample from an international study and found that in their samples of men (n = 14 compared with n = 210) and of women (n = 18 compared with n = 270), there were no significant differences in being high volume drinkers by sexual orientation, but this may be because of small sample sizes. Booker et al., (2017) reported two measures of alcohol use in 16- to 21-year olds from The UK Household Longitudinal Study (Understanding Society). There were no significant differences in the frequency of past month alcohol consumption by sexual orientation, but gay/lesbian and bisexual respondents reported significantly more binge drinking four or more times in the previous month compared with the heterosexual respondents. Chakraborty et al., (2011) analysed results from the Adult Psychiatric Morbidity Survey by sexual orientation and partnership preference. She found that significantly more in the non-heterosexual group (n = 650) reported symptoms of alcohol dependence in the past 6 months than in the heterosexual group (n = 6811), but no significant difference between the ‘only opposite gender’ (n = 6794) compared with ‘any same gender’ (n = 667) partnered groups. The unadjusted odds ratio for alcohol dependence was 2.04 (95% CI 1.46–2.86) and adjusted odds ratio was 2.05 (95%CI 1.45–2.80) for non-heterosexual orientation compared with heterosexual orientation, and unadjusted OR for same-sex partnered was 1.31 (95%CI 0.91–1.90) compared with opposite-sex partnered people. Hagger-Johnson et al., (2013) analysed results from the Longitudinal Study of Young People in England (18–19 years), and found significantly higher rates of alcohol drinking more than w days per week and risky single occasion drinking in the lesbian or gay group (n = 88) compared with the heterosexual group (n = 7464) (P < 0.001 for both outcomes) but not for the bisexual (n = 146) compared with the heterosexual group (P = 0.057 and 0.09, respectively). Pesola et al., (2014) gave means and standard deviations for alcohol problems at ages 16 and 18 but no statistical comparisons between sexual minority and heterosexual groups. Pitman et al., (2021) analysed results from two waves of the Adult Psychiatric Morbidity Survey and found that there were significantly higher proportions of alcohol risk, as measured by having an AUDIT score of 8 and above, in the lesbian/gay (n = 163) and bisexual (n = 116) groups compared with the heterosexual group (n = 10 016) (P = 0.006). Shahab et al., (2017) reported three alcohol-related outcomes – mean urge to drink, mean motivation to cut down drinking and mean spent per week on alcohol. For both women and men, all outcomes were significantly worse in the lesbian/gay and bisexual groups than the heterosexual groups, except for no difference in the motivation to cut down in women. Woodhead et al., (2016) reported results from a Southeast London Community Health Survey and the Adult Psychiatric Morbidity Survey, and reported that a significantly higher proportion of the non-heterosexual group (n = 63) drank alcohol ‘to cope with unfair treatment’ than the heterosexual group (n = 978) (P = 0.027). (Unfair treatment was defined in this study as being treated unfairly in a list of 12 specific situations and responses were self-reported.)

Table 2

Publications containing prevalence data—sexual orientation and mixed sexual orientation and gender identity

Publication	Outcome	LGBT	LGB	Lesbian/gay	Bisexual	Lesbian/bi women	Gay/bi men	Straight	Additional information
Amos et al., (2020)	Ever drank alcohol		67.45% (62.52–72.02) n = 628					51.51% (50.17–52.84) n = 9227	OR = 1·85 (1·47 to 2·33) P < 0.0001 Adolescents aged 14 years old
	Regular drinking (of those saying yes to drinking alcohol)		1.07% (0.36–3.11) n = 385,					1.27% (0.94–1.72) n = 4048	OR 0·50 (95%CI 0·14 to 1·81) P = 0·288
Becares (2021)	Harmful alcohol use (AUDIT-C)					71.6% (SE = 4.8)		59.0% (SE = 0.7)	Odds ratio = 1.74 (95%CI =1.09, 2.79, P < 0.05)
Bloomfield et al., (2011)	High volume drinking			Men 36% (n = 14) Women 11% (n = 18)				Men 17% (n = 210) Women 11% (n = 270)	Partnered or recently partnered. Men OR 3.32 (SE 2.07, P = 0.055) Women OR 1.04 (SE 0.81, P = 0.960)
Booker et al., (2017)	Alcohol consumption once per week or more			57% (n = 482)	54% (n = 407)			57% (n = 38 073)	UKHLS total sample
	Alcohol consumption once per week or more			45%	43%			43%	16 to 21–year olds in sample (numbers not given)
	Binge drinking in last 4 weeks, 4+ times			15%	14%			11%	16- to 21-year olds in sample
Chakraborty et al., (2011)	Alcohol dependence in past 6 months	10.4% (n = 650). 7.3% (n = 667)						5.4% (n = 6811). 5.7% (n = 6794)	Heterosexual vs. non-heterosexual identity. Same-sex partner vs. opposite-sex partner
Hagger-Johnson (2013)	Alcohol drinking >2 days/week (age 18/19)			37.5% (n = 33)	26.0% (n = 38)			19.7% (n = 1467)
	Risky single occasion drinking (age 18/19)			45.5% (n = 40)	32.9% (n = 48)			26.6% (n = 1985)
Pesola et al., (2014)	Alcohol use in sexual minority compared with heterosexual young people age 18								Total N = 3710 Percentages not given [Btotal 0.12 (95%CI 0.04–0.20, P = 0.003)]
	Alcohol problem use at age 16 (mean, SD)		7.9 (5.5) (n = 237)					6.6 (4.8) (n = 1628)
	Alcohol problem use at age 16 (mean, SD)		8.8 (5.9)					8.2 (5.2)
Pitman et al., (2021)	AUDIT score 8 or above			37.4%	31.0%			23.8%
Shahab et al., (2017)	Mean (SD) urge to drink				1.6 (1.2) (n = 60)	1.8 (1.2) (n = 50)		1.5 (0.9) (n = 3952)	Women
	Mean (SD) Motivation to cut down drinking				1.9 (1.4)	2.1 (1.9)		1.8 (1.6)
	Mean (SD) spent per week (£)				16.5 (16.6)	19.0 (16.5)		14.6 (14.0)
	Mean (SD) urge to drink				2.0 (1.3) (n = 51)		1.7 (1.2 (n = 87)	1.7 (1.0) (n = 4426)	Men
	Mean (SD) Motivation to cut down drinking				1.9 (1.7)		2.1 (1.7)	1.7 (1.4)
	Mean (SD) spent per week (£)				21.3 (21.5)		25.4 (23.0)	21.6 (20.2)
Woodhead et al., (2016)	Harmful alcohol use	18.8% (n = 12)						3.9% (n = 41)

Five grey literature reports [(Balding 2014) (schoolchildren in Year 10 only), the Stonewall Gay Men’s Health Survey (men only), LGB people in later life (older people aged over 55), RaRe report (lesbians and bisexual women) and Youth Chances (young people aged 16–25)] reported prevalence of alcohol use in LGB people compared with heterosexuals, and one (Balding 2018) compared LGBT schoolchildren in Year 10 to heterosexual children (Table 3). Most of the results showed more alcohol use in the LGB groups compared with heterosexual peers. Only the RaRe report (Nodin et al., 2015) conducted significance tests and found no significant difference in risky drinking as measured by the AUDIT questionnaire between LGB (n = 534) and heterosexual women (n = 470) and between lesbian (n = 324), bisexual (n = 170) and heterosexual women. There were, however, significantly more LGB women who scored as ‘possibly alcohol dependent’ and significantly higher patterns for the LGB group on ‘drinking to intoxication’.

Table 3

Non-population surveys with comparative results.

Report	Outcome	LGBT	LGB	Lesbian/ gay	Bisexual	Lesbian/bi women	Gay/bi men	Straight	Notes
Balding (2014)	Alcoholic drink in the last 7 days		Boys 30% (n = 60) Girls 51% (n = 92)					Boys 37% (n = 2002) Girls 35% (n = 1916)	School Year 10s in Cambridgeshire
	Drank over 14 units in the last 7 days		Boys 7% Girls 1%					3% 2%
Balding (2018)	Alcoholic drink in the last 7 days	Boys 34% (n = 74) Girls 38% (n = 139)						Boys 34% (n = 1467) Girls 36% (n = 1538)	School Year 10s in Cambridgeshire
	Drank over 14 units in the last 7 days	Boys 8% Girls 6%						Boys 3% Girls 2%
Nodin et al., (2015) (RaRe Report)	Hazardous alcohol use (AUDIT)					37.1% (n = 534)		31.9% (n = 470)
	Hazardous alcohol use					35.8% (n = 324) lesbian 40.0% (n = 170) (bi women)		31.9% (n = 470)
	Dependent alcohol use					4.5% (n = 534)		4.0% (n = 470)
	Dependent alcohol use					4.3% (n = 324) lesbian 4.7% (n = 170) (bi women)		4.0% (n = 470)
	Drink alcohol to intoxication once per week or more					14.6% (n = 587)		13.1% (n = 533)
	Drink alcohol to intoxication once per week or more					15.1% (n = 357) lesbian 12.6% (n = 189) (bi women)		13.1% (n = 533)
Guasp (2013)   [Stonewall GB men’s health survey (also Bourne et al., 2016)]	Had a drink in the last week						78% (n = 6861)	68% (n =???)
	Drank alcohol on 3 or more days in the last week						42%	35%
	Drunk or hung over while working/school or other responsibilities in past 6 months						19%
	Missed or late for work/school or other activities because of the above						13%
	Drunk alcohol even though doctor suggested they stop drinking						4%
Guasp (2011) (Stonewall LGB people in later life)	Drink alcohol every day or 5–6 days per week		Men 35% Women 19%					25% 15%
Youth Chances (2014)	Alcoholic drink 4 or more times per week		8% (n = 4332)					6% (n = 341)	16–25 year olds only
	Drinking more than 4 units in a typical day		57% (n = 3933)					57% (n = 300)	Of those that drink
	Drink twice the recommended limit in the last year, monthly or more		51% (n = 3930)					45% (n = 300)	Of those that drink

Sexual orientation and mixed sexual orientation and gender identity.

Asexuality

Bauer et al., (2020) published results of alcohol consumption in asexual and grey-asexual people compared with allosexual people in four UK surveys—1990 (NATSAL I), 2000 (NATSAL II), 2010 (NATSAL III), and a follow-up survey based on these censuses, and the Towards Better Sexual Health: A Study of Sexual Attitudes and Lifestyles of Young People in Northern Ireland (TBSH) survey in 2000. They also reported odds ratios of drinking by sexual attraction from NATSAL III, adjusted for age, sex, health and importance of religion only, with opposite-sex attraction being the reference case. Results for each group were as follows: Mostly opposite-sex attraction OR = 1.60, About equal attraction OR = 0.87, Mostly same-sex attraction OR = 1.69, Only same-sex attraction OR = 1.55, Grey-asexual OR = 0.42 and Asexual OR = 0.14. This showed that asexual people drank less alcohol than allosexual people, and same-sex-attracted people drank more alcohol than opposite-sex-attracted people.

Gender identity

We found no population surveys in the trans and non-binary communities (Table 4). The only evidence was from a grey literature report from the Youth Chances survey (2014) and one publication by Rimes et al., (2019) based on this survey. Assuming that the small heterosexual comparator group was also cisgender, the relative proportions of having an alcoholic drink four or more times per week were slightly higher in the trans sample than in the heterosexual group. However, the percentages having a higher number of units on a typical day and the frequency of having six (if female) and eight or more units (if male) were lower in the trans sample.

Table 4

Non-population surveys, published and grey literature.

Report	Outcome	Trans	Non-binary	Cisgender/comparator	Intersectionality
Youth Chances (2014)	Alcoholic drink 4 or more times per week	8% (n = 665)		6% (n = 341)	16–25 year olds only
	Drinking more than 4 units in a typical day	44% (n = 551)		57% (n = 300)	Of those that drink
	Drink twice the recommended limit in the last year, monthly or more	38% (n = 549)		45% (n = 300)	Of those that drink
Rimes et al., (2019) (Youth Chances)	Mean AUDIT score (SD)	Trans female 5.1 (2.6) Trans male 4.3 (2.4)	MAAB 5.2 (2.8) FAAB 4.9 (2.5)

Gender identity.

Intersex

Although intersex people were in some of the samples, no data on alcohol use for people with intersex variance were found.

Quotes from respondents relevant to alcohol use

Several grey literature reports gave insights into how respondents see alcohol use for themselves or as part of the LGBT+ community. Some quotes highlight the societal role that alcohol use plays culturally in helping people deal with heteronormativity or gendernormativity, or in response to emotional pain, or related to being ‘othered’ that may lead to further harm:

‘It’s difficult to find other gay men unless you go clubbing or pubbing and so much of gay socialising involves alcohol and often drugs’ (Nodin et al., 2015).

Some quotes reflected the key role alcohol plays in socialization whilst acting as emotional support during current difficult times:

‘I just learned to use it as a crutch to support me when times got emotionally tough and yeah I just woke up to realise what I was doing wasn’t healthy’ (Nodin et al., 2015).

Although other quotes show the impact of past trauma or emotional harm related to bullying and its current results:

‘I am a trainee doctor and would be considered successful, but I hide the fact that it’s a daily struggle. I often deal with depression and alcohol dependence due to absence of self-esteem, both resulting from childhood bullying. I've little support and live a fairly lonely life. Many LGBT people have similar experiences. It’s great that things are moving forward, but for many, significant damage of the past remains a factor in the present’ (Guasp 2013).

Moreover, other quotes show how alcohol may be used to deal with normative heterosexuality associated with unattainable societal roles:

‘I felt at the time my parents, or my mother, was alive I’d broken her spirit and her hope for me as a young woman because there was a lot of peer pressure surrounding me getting married and having grandkids, so the white wedding and everything like that. So when I did come out there was the disappointment. I’d let my parents down and having to deal with their excuses of why I might be a lesbian. So again it’s upsetting and guilt and I didn’t live up to their expectations so drink again basically blanked all that out. So I can deal with it as long as I’ve got a drink’ (Nodin et al., 2015).

As reflected in these quotes, alcohol may act as a mechanism to help people cope with marginalization when they do not conform to gender and sexual norms. For some people, alcohol may make life more liveable, but could also result in harm when used in excess.

Discussion

Included in this systematic scoping review were 12 publications and 9 grey literature reports (20 studies in total). As alcohol use was reported in a wide variety of ways in different groups of the LGBTQ+ population, no meta-analyses could be conducted to give an overall estimate of the higher use of alcohol in minority sexual and gender identity populations. However, all studies had higher rates of alcohol use in minority sexual orientation or mixed LGBTQ+ populations compared with heterosexual or cisgender populations. One study investigating alcohol use in asexual people (Bauer et al., 2020) found that they drank less than allosexual people. One study investigating trans and non-binary populations found possibly lower rates of alcohol use compared with the comparator (Youth Chances 2014), but was apparently high when compared with average rates of drinking in the UK population (Zambon 2021). There was no information available regarding alcohol intake for UK intersex people. The qualitative quotes from grey literature illustrate how some LGBT+ people use alcohol to give some temporary relief from social or mental health difficulties as well as some reporting that alcohol is part of their social milieu.

These results are consistent with findings from international systematic reviews. For example, Hughes et al., (2020) investigated alcohol and drug use in sexual minority women and found three to seven times higher odds of past-year alcohol dependence when compared with heterosexual women, controlling for demographic factors including age. King et al., (2008) investigated lifetime and 12-month prevalence of alcohol dependence in men and women together and separately. They found statistically significantly higher risk ratios (RR) of lifetime prevalence of alcohol dependence in both sexes [RR 2.59 (95%CI 1.62–4.15)] and women [RR 6.51, (95%CI 2.74, 15.44)] but not in men [RR 1.60 (95%CI 0.91–2.80)]. For 12-month prevalence, the RRs were 2.22 (95%CI 1.78–2.77) in men and women, 1.51 (95%CI 1.13–2.02) in men only and 4.00 (95%CI 2.85–5.61) in women only.

Limitations

A range of alcohol measurements is currently used in research. In this systematic scoping review, a range of SOGI descriptors was also used, such as LGBT+, LGB, lesbians and gay men, bisexual women and men, lesbian and bisexual women, gay and bisexual men. Because of this, we were unable to conduct meta-analyses and generate any overall prevalence estimates. The vote-counting approach used here is acknowledged to be very limited in that it does not consider the size or quality of studies.

The use of the first author’s private library means that full search results are not replicable by other researchers because some of the studies are no longer available on the internet. Only two studies were available from this source (Balding 2014, 2018).

We did not formally quality assess any included studies using a critical appraisal tool and have used the study design only. We chose this because a checklist such as the Newcastle-Ottawa Scale adapted for cross-sectional surveys (Herzog et al., 2013) did not give any additional information, as almost all studies did not give a full description of the sampling strategy, justification of the sample size, description of response rate or justification of alcohol measurement used.

Implications for policy and practice

If a service provider does not record the SOGI status of their service users, they may not be aware that their service has a relatively high or low proportion of LGBTQ+ people. An understanding of the key role of personal and social identity in drinking practices in the LGBTQ+ community is required by those who provide support, as well as the individual and social factors that may be at work when LGBTQ+ people drink alcohol to excess. Service providers need to be more aware of LGBTQ+ issues, which will probably entail specialist training. As all studies showed higher rates of alcohol use in LGBTQ+ populations, there are policy implications for programmes aimed at reducing population rates of alcohol-related harms. Specific interventions for LGBTQ+ people systematically reviewed by Dimova et al., (2022) showed some promising results using cognitive-behavioural therapies and motivational interviewing. Changing infrastructure such as reducing targeted alcohol advertising to the LGBTQ+ community may be of benefit, but the LGBTQ+ community has few resources, so the reduction of sponsorship revenue may reduce community events, increase loneliness and, ultimately, increase alcohol consumption. Public awareness and acceptance of LGBTQ+ people to reduce stigma and isolation could, in turn, reduce alcohol consumption.

Implications for research

This systematic scoping review has identified gaps in UK-relevant research. Regarding general research into alcohol use, ways of its measurement need to be standardized across research and generally agreed upon by alcohol researchers and other stakeholders. The agreed methods should be clearly described and available to anyone wishing to conduct alcohol research.

The UK ONS validated a measure of sexual orientation in 2008 (Vizard 2014), and the 2021 UK Census measured SOGI status (ONS 2021c), so these tools are available to researchers. We recommend that all UK-based cohort studies of alcohol use incorporate these measures and use them consistently. It is also essential that results by sexual orientation and by gender identity are reported. In addition, these results should be disaggregated by sex or gender—we found few data that disaggregated alcohol use by sex or gender.

Conclusions

This systematic scoping review has demonstrated that there is consistent evidence from a number of studies showing that the prevalence of alcohol use is higher in LGBT+ people compared with their heterosexual peers in the UK. We have highlighted some missing information. Fewer asexual people drink alcohol than allosexual people. There is little information on alcohol use in trans compared with cisgender people, and there is no evidence in intersex people. These UK findings are novel but in line with systematic review evidence from other countries. Government-funded cohort studies should, as a matter of routine, include SOGI status in demographic data, so that better quality information becomes available. Alcohol service providers need to assess SOGI status routinely and ensure that their staff is adequately trained in equality and diversity issues. Standardizing of the reporting of sexual orientation, gender identity and alcohol misuse measures would improve comparability across studies.

Acknowledgements

Our thanks to the peer reviewers for their comments on an earlier version of this manuscript. Thanks also to Drinkaware, Emma Catterall and the members of the project’s scientific advisory board for their steer throughout.

Author contributions

Catherine Meads (Conceptualization-Equal, Data curation-Equal, Formal analysis-Lead, Funding acquisition-Supporting, Investigation-Lead, Methodology-Lead, Validation-Equal, Writing – original draft-Lead), Laetitia Zeeman (Conceptualization-Lead, Data curation-Supporting, Funding acquisition-Lead, Investigation-Supporting, Methodology-Equal, Project administration-Lead, Writing – original draft-Supporting, Writing – review and editing-Equal), Nigel Sherriff (Conceptualization-Equal, Funding acquisition-Supporting, Investigation-Supporting, Supervision-Lead, Validation-Equal, Writing – review and editing-Supporting), (Kay Aranda) Validation-Equal, Writing – original draft-Supporting, Writing – review and editing-Equal).

Conflict of interest: The authors have no conflicts of interest and have taken the risk of bias into account, where possible, when appraising the included studies.

Funding

This project was funded by the Drinkaware Trust, UK. The information and views set out in this journal article are those of the authors and do not necessarily reflect the opinion of the Drinkaware Trust.

Data availability

The data underlying this article are available in the article and in its online supplementary material.