Typology of drug use in United Kingdom men who have sex with men and associations with socio-sexual characteristics
Introduction
Analyses of drug use patterns in men who have sex with men (MSM) have generally focused on examining individual drugs in isolation (Digiusto & Rawstorne, 2013; Vosburgh, Mansergh, Sullivan, & Purcell, 2012), or have examined polydrug use as a unitary variable (Boone, Cook, & Wilson, 2013; Daskalopoulou et al., 2014) without regard to types of drugs combined. This leaves aside two important points for analysis—namely, understanding how patterns of specific drug use cluster together within MSM, and how different drug use patterns are associated with socio-sexual variables.
Understanding specific drug use patterns is important in targeting HIV prevention and harm reduction interventions and in developing a fuller picture of drug use in context. For example, reports of sexualised drug use among MSM (also known as ‘chemsex’) in the UK focus on the use of three specific drugs—crystal methamphetamine, or ‘crystal meth’; gamma-hydroxybutyric acid, or GHB; and mephedrone (McCall, Adams, Mason, & Willis, 2015), with ketamine at times included as a chemsex drug (Bourne, Reid, Hickson, Torres-Rueda, Steinberg, 2015; Bourne, Reid, Hickson, Torres-Rueda, & Weatherburn, 2015). Other latent class-based approaches to understanding typologies of substance use have drawn primarily from the US context (McCarty-Caplan, Jantz, & Swartz, 2014; Tobin, Yang, King, Latkin, & Curriero, 2016; Yu, Wall, Chiasson, & Hirshfield, 2014). Thus, whether these drugs form a coherent pattern of use, or whether they are part of broader patterns of drug use, has not yet been investigated in the UK context using quantitative data.
In this analysis, we use latent class analysis of data from a cross-sectional community-based survey of MSM living in the UK to develop a typology of recent drug use in MSM, and to explore how distribution of MSM across the classes in this typology differs by socio-sexual characteristics (including demographics, HIV testing history and sexual risk-related variables). Latent class analysis is a type of factor analysis. It constructs underlying and non-observed variables from observed data, specifically multicategorical latent variables, which can be used to develop underlying typologies of people. This analysis is important to understand patterns of use of specific drugs in the context of use of other drugs and to move beyond bivariate associations between use of specific drugs and other person-level characteristics.
Section snippets
Methods
This study used data from the 2014 Gay Men’s Sex Survey. This open-access, internet-based survey recruited a community sample of MSM via general-interest gay dating websites and apps, Facebook, and the websites of community organisations in late summer 2014. While a community sample may lead to underrepresentation or overrepresentation of certain individual characteristics or risk behaviours, it is difficult to establish a probability sample of a group as diffuse in terms of behaviour and
Results
The analysis sample consisted of 16,814 participants resident in the UK with data available for analysis. Of this group, 81.9% were White British and 11.3% were White non-British, 3.3% were Asian, and 1.9% were Black. An additional 1.6% identified with other ethnic groups. Participants were on average 35.1 years of age (SD 13.2). Almost half (48.5%) had a university degree. In terms of HIV status, 24.0% had never received an HIV test result, 8.8% had tested positive, and 67.2% had a last test
Discussion
This analysis developed a set of latent classes to summarise different typologies of drug use in MSM living in the United Kingdom. The five latent classes described a range of drug use patterns, including minimal users, low-threshold users, old-skool users, chemsex-plus users and diverse users. The model achieved good separation and classification of respondents, and key socio-sexual characteristics were found to predict class membership.
Several noteworthy observations emerged from this latent
Conclusion
In our analysis, we were able to discern five distinct drug use subgroups of MSM in the UK and we identified how these relate to key socio-sexual characteristics and other demographic characteristics. This latent class model, like all analyses of its kind, is tentative and should be revisited as cultures and contexts of drug use change over time. However, it points to the need for service providers and clinicians to understand how MSM ‘engage’ with the full range of drugs both individually and
Declaration of interest
The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.
Authors report no conflicts of interest.
Acknowledgement
This work was supported by the Department of Health for England through the HIV Prevention England programme.The work was undertaken with the support of The Centre for the Development and Evaluation of Complex Interventions for Public Health Improvement (DECIPHer), a UKCRC Public Health Research Centre of Excellence. Joint funding (MR/KO232331/1) from the British Heart Foundation, Cancer Research UK, Economic and Social Research Council, Medical Research Council, the Welsh Government and the
References (19)
- et al.
Methamphetamine use among gay men across the UK
International Journal of Drug Policy
(2010) - et al.
Chemsex and harm reduction need among gay men in South London
International Journal of Drug Policy
(2015) - et al.
Latent class analysis of substance use among men who have sex with men in Malaysia: Findings from the asian internet MSM sex survey
Drug and Alcohol Dependence
(2015) - et al.
Poly drug use, chemsex drug use, and associations with sexual risk behaviour in HIV-negative men who have sex with men attending sexual health clinics
International Journal of Drug Policy
(2017) - et al.
Substance use and sexual risk behavior in HIV-positive men who have sex with men: An episode-level analysis
AIDS and Behavior
(2013) - et al.
Illicit drug use in sexual settings (chemsex) and HIV/STI transmission risk behaviour among gay men in South London: Findings from a qualitative study
Sexually Transmitted Infections
(2015) - et al.
Recreational drug use, polydrug use, and sexual behaviour in HIV-diagnosed men who have sex with men in the UK: Results from the cross-sectional ASTRA study
Lancet HIV
(2014) - et al.
Is it really crystal clear that using methamphetamine (or other recreational drugs) causes people to engage in unsafe sex?
Sexual Health
(2013) - et al.
Sensation seeking and alcohol use as markers of sexual transmission risk behavior in HIV-positive men
Annals of Behavioral Medicine
(2002)
Cited by (26)
Drug use typology, demographic covariates, and associations with condomless anal sex: A latent class analysis among a U.S. national cohort of men who have sex with men
2023, International Journal of Drug PolicyCitation Excerpt :While many classes between regions may be similar, some dissimilar patterns of use can be seen. A 2018 study of United Kingdom (UK) MSM identified six latent classes of substance use: 1) minimal use, 2) low-threshold use (poppers, erectile dysfunction (ED) medications, cannabis), 3) old-school use (mostly ecstasy, cocaine, and poppers), 4) chemsex-plus (chemsex drugs as well as low-threshold drugs), and 5) diverse use (all drugs) (Melendez-Torres et al., 2018). Meanwhile, in a 2015 study of internet-recruited US MSM, the authors identified six distinct use classes: 1) no use, 2) recreational/sporadic use, 3) poppers and prescription ED drug use, 4) poppers with non-prescription and prescription ED drug use, 5) recreational, club, and ED drug use, and 6) polydrug use of a variety of substances (Yu et al., 2015).
Patterns of adversity and post-traumatic stress among children adopted from care
2022, Child Abuse and NeglectCitation Excerpt :Inspection of the 6-class model showed that compared to 5-classes, entropy decreased by 0.01, and the BLRT strongly suggested that this was not better than the 5-class model. Based on these results in conjunction with concerns about substantive interpretability of the classes, we chose the model with five classes as it balanced statistical criteria and theoretical judgement (Melendez-Torres, Bourne, Reid, Bonell, & Weatherburn, 2018). The 5-class solution suggested five distinct classes of adversity: a lower exposure class, a perinatal risk class and different patterns of adversity in the care of their birth parents (Table 2).
Non-prescription drug use among HIV positive gay and bisexual men in Australia: A latent class analysis and comparison of health and well-being
2022, International Journal of Drug PolicyCitation Excerpt :In most cases, these latent class analyses of drug use among gay men have focussed on examining associations with demographic and socio-sexual characteristics (i.e. who is most likely to fall into each typology and how they behave sexually). For example, an analysis by Melendez-Torres et al. (2018) of 16,814 men taking part in an online community survey in the UK established the following groupings of drug use: ‘minimal users’ (where use of any drug was low), ‘low-threshold users’ (who were largely defined by their use of historically popular drugs), ‘old-skool users’ (who commonly used drugs such as ecstasy or cocaine, but also occasionally new psychoactive substances), ‘chemsex-plus users’ (who used drugs associated with chemsex [e.g. crystal methamphetamine or GHB/GBL] as well as other drugs) (Bourne et al., 2015) and ‘diverse users’ (who used a very broad range of drugs). Far fewer studies have sought to establish the sequelae of drug use classes for health and well-being.
Using Latent Class Analysis to Explore Complex Associations Between Socioeconomic Status and Adolescent Health and Well-Being
2021, Journal of Adolescent HealthA Latent Class Analysis of Parental Alcohol and Drug Use: Findings from the Avon Longitudinal Study of Parents and Children
2020, Addictive BehaviorsCitation Excerpt :Once the fit criteria were achieved, the latent variable of mothers’ alcohol use was used in the latent class analysis in a one-step method. Latent class analysis, a derivative of factor analysis, was used to explore unobserved constructs in observed data (Melendez-Torres et al., 2018) and therefore the presence of underlying classes in the parental substance use variables were identified. This form of analysis facilitates an understanding of substance use in context, going beyond one-dimensional definitions (i.e. alcohol or drugs) and mutually exclusive conceptualisations (i.e. maternal or paternal).