Article Text
Abstract
Objective There has been an increase in new HIV diagnoses among young men who have sex with men (YMSM) over the past decade in both UK and US contexts, with online sex-seeking implicated in driving this development. This study sought to examine YMSM's use of a variety of social and sexual networking websites and ‘apps’, and assess sexual risk behaviours.
Design YMSM were recruited from across four countries in Britain and Ireland, via an online survey using convenience sampling. Data were collected from 2668 men, of whom 702 were aged 18–25 years.
Results Facebook use was almost ubiquitous and for largely social reasons; sexual media use was common with 52% using gay sexual networking (GSN) websites frequently and 44% using similar apps frequently. We found increased odds of high-risk condomless anal intercourse associated with the length of time users had been using GSN websites and lower levels of education. We found no significant differences across the four countries in sexual risk behaviours.
Conclusions YMSM are a heterogeneous population with varied sexual health needs. For young men with digital literacy, individual-level online interventions, targeted and tailored, could be directed towards frequent users with lower levels of education. Variation in demographic characteristics of GSN websites and app users may affect who interventions are likely to reach, depending on where they are targeted. However, interventions, which may catch young men earlier, also provide a major opportunity for reducing sexual health inequalities.
- HIV
- SEXUAL BEHAVIOUR
- ADOLESCENT
- GAY MEN
Statistics from Altmetric.com
Introduction
HIV diagnoses among young men who have sex with men (YMSM) have increased. Two examples show the trend: an 8.7% increase in the UK in 2004–2013; a 34% increase in the USA between 2006 and 2009. (HIV in the UK, 2014)1 A recent study by Beyrer et al2 drew on data from across 15 countries to assess the global epidemiology of HIV, and concluded while there was a lack of data on adolescent men who have sex with men (MSM), what existed suggests young men are a high-incidence group. While the literature on the wider MSM population has grown to trial stage, and a meta-analysis has been conducted on the burgeoning evidence base on online sex-seeking and sexual risk behaviours,3 there remain, in comparison, far fewer studies that focus exclusively on YMSM or disaggregate data to convey clear findings on YMSM. Accordingly, to date, the evidence base in terms of intervention development and trials remains insufficient.4 Adolescence represents a particularly critical and sensitive phase within the life course,5 with families, peers and supportive environments fundamental to adolescent health.6 Given the circumstances and experiences of YMSM may differ to those of their heterosexual peers or to those of older MSM, it is vital that interventions for YMSM are not simply adaptations of those developed for heterosexual or adult MSM populations.7
We reviewed studies that focused on YMSM and found an emerging picture. US-based studies have, for example, revealed the sex-seeking behaviour of YMSM: Winetrobe et al8 recruited YMSM (n=146) in Los Angeles to explore associations between unprotected anal intercourse (UAI) and partner-seeking behaviours on Grindr (a geosocial networking application geared towards gay, bisexual and bi-curious men), finding higher levels of UAI associated with greater familiarity with the app; Horvath et al9 considered whether YMSM met partners online, offline or both, with the latter more likely to report a greater number of sexual partners and UAI. However, this is far from a consistent picture, as other work has shown no association between partners met online and increased sex risks.10 Issues around the operationalisation of sex-seeking behaviour may be a factor in explaining such conflicting findings.11 As regards testing behaviours, YMSM appear to test less frequently for HIV than older MSM.12 ,13 Yet, YMSM find apps acceptable for interventions, and are willing to engage with prevention efforts via these technologies.14 ,15
The predominance of US data warrants evidence from other contexts and subpopulations; intervention development requires an understanding of the transferability of US findings to Europe, the patterning and frequency of digital technology use and equally understandings of the subpopulations within YMSM who may warrant particular targeting. Finally, this is a rapidly changing field, so even fairly recent studies do not maintain a monopoly on knowledge for long; we need to stay abreast of changing tastes and behaviours for optimising intervention effectiveness.
We conducted a cross-sectional, online survey of YMSM's use of a variety of social and sexual networking media across four countries in Britain and Ireland. We address two key questions here, first, describing frequency of, and reasons for using websites and apps. Second, we assessed men's sexual risk behaviours and its correlation with use of social and sexual networking websites and apps. While our survey recruited n=2668 men aged 18+ years, overall, this paper focuses on the group (n=702) of YMSM (aged 18–25 years). It represents the first study of its kind within Europe to convey data specific to YMSM.
Methods
Participants
We invited MSM residents across Scotland, Wales, Republic of Ireland (ROI) and Northern Ireland (NI) to complete an online survey on their use of various social and sexual networking websites and apps, their sexual behaviours and their views towards the use of websites and apps for HIV prevention. Facebook refused us permission to recruit men from England and so we chose not to recruit English men across all recruitment sites. Pop-up message ‘blasts’ and/or banner adverts were created in order to invite men using gay-specific social media websites (Gaydar, Recon and Squirt), smartphone apps (Grindr and Gaydar) and Facebook to participate via Survey Monkey. For recruitment via Facebook, we sought those who had chosen to ‘like’ ‘LGBT’, ‘gay’ or other such tags. We sought men aged 18+ years, and as Gaydar, Grindr, Squirt and Recon have a minimum user age of 18, we set this age for the other sites including Facebook. Users located in Scotland, Wales, NI or the ROI were targeted, based on their computer IP addresses or, for smartphone social media, their GPS coordinates. Irrespective of which social media people were recruited by, they were all asked the same questions regarding the range of social media they used. Recruitment took place between November 2012 and February 2013.
Survey
The survey link took men to an online information sheet, which provided men with details of the study, the questions, the confidentiality and anonymity assurance. We sought data regarding: participants’ sociodemographics (age, area of residence, education, sexual orientation, employment, ethnicity, partnership status, living situation); participants’ frequency and reasons for use of Facebook, gay sexual network (GSN) websites, GSN smartphone apps and barebacking websites/apps (such as http://www.barebackrt.com); sexual partners in the previous 12 months, whether these were casual partners, knowledge of HIV status of partners, use of condoms, whether seroadaptive strategies had been employed and details of most recent HIV and sexually transmitted infection (STI) tests.
For social and sexual media use, we defined frequent use as between ‘every few days’ up to ‘all the time’. We computed a high-risk sex variable to include men who reported any one of the following three behaviours: condomless anal intercourse (CAI) with two or more partners; CAI with casual partners and/or CAI with HIV status unknown/serodiscordant partners in the previous 12 months. This measure has been used previously in the Medical Research Council/Chief Scientist Office (MRC/CSO) Gay Bar Studies.16 We refer to CAI rather than UAI to reflect the changing context of viral suppression, but note that at the time of writing, the drugs used for pre-exposure prophylaxis (PrEP) are not licensed for use in the four countries included in this study. Ethical approval was granted by Glasgow Caledonian University School of Health and Life Sciences Ethics Subcommittee (HLS id: B11/59).
Analysis
Statistical analyses were conducted in IBM SPSS V.21. We sought to undertake two main analyses: first, we examined patterns of social and sexual media use by demographic and behavioural characteristics (country, age, highest level of education, sexual identity, partnership status, level of outness and reporting high-risk CAI), using the χ2 test; second, we used our measure of high-risk CAI as a dependent variable to examine differences by various demographic and behavioural characteristics (country, age, highest level of education, sexual identity, partnership status, level of outness—given their potential as conceptual confounders) as well as duration and frequency of websites and app use. For this latter analysis, variables significant at the bivariate level (p<0.05) were included in a multivariate logistic regression model used to estimate ORs and 95% CIs for high-risk CAI.
Findings
We recruited 702 YMSM, aged 18–25 years, with a mean age of 21.4 years (SD=2.17). The majority of YMSM were based in Scotland (47.4%, n=334), with 22.4% (n=158) ROI, 15.8% (n=111) NI and 14.3% (n=101) Wales. Most identified as white (95.4%, n=669) and reported being HIV negative (89%. N=496) with 1.4% (n=8) positive and the rest status unknown. Many young men were students (50.8%, n=356); a large minority were employed (42.2%, n=296). The majority identified as gay (78.8%, n=551), with 17.7% bisexual (n=124) and 3.4% straight (n=24). Most were educated to degree level (58.5%, n=402), or further education beyond high school (35.2%, n=242), were single (71.1%, n=496) and out to almost everyone (68.2%, n=376). Most were recruited via the Facebook survey (33.8%, n=237), followed by the Grindr survey (26.1%, n=183) and Gaydar instant message (24.5%, n=172).
Patterns of social and sexual media use
Table 1 shows the percentage of YMSM who use Facebook, GSN websites, GSN apps and barebacking sites (web or app). While the vast majority (96.1%, n=667/694) of YMSM use Facebook, 75.1% (493/656) use GSN websites, 60.4% (n=373/618) use GSN apps and 10.1% (n=61/601) use barebacking sites. The frequency of use of these sites varied, with more YMSM using the GSN websites frequently (55.6%, n=365/656), compared with apps (50.5%, n=312/618), but fewer used these frequently compared with Facebook (92.1%, n=639/694). The majority of YMSM do not use barebacking sites, although 5.8% (n=35/601) use them frequently. Figure 1 shows the pattern of use of social and sexual media.
The reasons men provided for their Facebook, GSN website and app use are shown in table 2. Common Facebook use reasons were: social networking, chatting to family/friends and to check family/friends’ status. Reasons for using GSN websites were: meeting guys for sex, to kill time and to meet guys for dates. Reasons for using GSN apps were: meeting guys for sex, to kill time and to swap naked pictures. Barebacking site use was for: meeting guys for sex, to look for bareback sex and to swap naked pictures. Thus, sites such as Facebook are used for social and GSN websites and apps used for sexual reasons; however, a substantial proportion of men are using them simply to ‘kill time’.
Several demographic and behavioural factors were associated with being a frequent user of Facebook, GSN websites and apps, shown in table 3. These univariate analyses found that frequent Facebook use was significantly associated with being out to all and using the gay scene; frequent use of GSN websites was associated with being unemployed or on sickness benefit (a government benefit for disability or ill health), higher levels of educational attainment, being single, not being ‘out’ and frequent use of GSN apps was associated with country of residence (Wales), employment (employed/self-employed), highest qualification (secondary/high school), being single, being gay identifying, having ever had an HIV test and use of the gay scene.
Sex-risk associations with use of GSN websites and apps
As to our second main question for this work, we assessed men's sexual risk behaviours and correlations with use of social and sexual networking websites and apps. Regarding testing behaviours, overall, just over half (53.3%, 302/567) of young MSM ever had an HIV test. Of those, 82.1% (248/302) reported having one in the past year. Almost half (45.8%, 257/561) never had an STI test, while 44.2% (248/561) reported having an STI test in the past year, of whom 12.8% (72/562) reported an STI. Regarding CAI, just under half (48.5%, 275/567) reported CAI in the past 12 months. Of these YMSM, 32.4% (88/272) reported CAI always with a casual partner, 29.8% (81/272) sometimes and 37.9% (103/272) never with a casual partner. Using our measure of ‘high risk sex’, we found 39.4% (104/264) of the group who reported CAI in the past 12 months fell into the high-risk category. As such, around one in seven YMSM in this sample had high-risk CAI in the past 12 months with a similar proportion for casual partnering.
We sought to examine the associations with high-risk CAI, but found no association between our high-risk CAI variable and country, employment, sexual orientation or length of time using the GSN apps. However, being a frequent user of GSN websites was significantly associated with reporting high-risk CAI (χ21=8.79; p=0.003), as was using GSN websites for a longer duration (with increasing length of time associated with increased risk of CAI) (χ25=19.14; p=0.002), being educated to high school level rather than degree or postgraduate (χ22=12.32; p=0.002) and having a partner rather than being single (χ21=6.76; p=0.009).
Based on these bivariate results, a multivariable logistic regression model examined associations with high-risk CAI, including the significant variables: education, duration of GSN website use, frequency of GSN website use and partner type. We found lower levels of education (ie, high school only) associated with increased odds of high-risk CAI (OR 3.1, 95% CI 1.5 to 6.9; p=0.002). There were also increased odds of high-risk CAI by the length of time using GSN websites, with users of 1–5 years being 2.4 times (OR 2.4, 95% CI 1.3 to 4.1; p=0.005), and those using the sites for more than 5 years 3.1 times (OR 3.1, 95% CI 1.5 to 6.1; p=0.001), more likely to report high-risk CAI than new users (<1 year). Frequent users of GSN websites were also 1.7 times more likely to report high-risk CAI than infrequent/non-users (OR 1.7, 95% CI 1.0 to 2.9; p=0.039), but relationship status was not related to high-risk CAI.
Discussion
This study examined YMSM's use of social and sexual media and sexual risk behaviours and sought to determine sex-risk associations with use of GSN websites and apps. A majority of YMSM were frequent users of Facebook, and 52% used GSN websites and 44% used GSN apps every few days or more. We found several demographic and behavioural factors were associated with frequent use, including how ‘out’ men were, their employment status, their level of education and whether they were in a relationship. We did not find country-level differences, indicating similar patterns of use across the four countries. Overall, these young men favoured GSN apps, but those who were non-gay identifying were more likely to use GSN websites. We found increased odds of CAI associated with the length of time users had been using GSN websites and lower levels of education. Overall, around one in seven YMSM had high-risk CAI in the past 12 months, and one in seven had CAI with a casual partner. We found no significant differences across the four countries in sexual risk behaviours.
For those seeking to develop interventions targeting YMSM in general, our data may be helpful as it illuminates who may be found amid the plethora of sites. It is worth considering that YMSM are regular users of both GSN websites and apps, with daily use of websites and apps being common; as such, there may be opportunities to layer health promotion over time, offer higher doses of intervention or boost intervention components. However, if young men who engage in high-risk CAI are sought, then targeting GSN websites may be advantageous, as our findings suggest it is here you are likely to find more of these particular young men. Nevertheless, we should also bear in mind the effectiveness of even brief interventions delivered online, particularly if they are theoretically based.17 ,18 Of note, our findings do not indicate a causal relationship between duration, or frequency, of use of GSN websites and high-risk CAI; however, they do suggest the viability of exploring the targeting of frequent and longer users of GSN websites who are at elevated risk of HIV transmission and enable complex intervention designs facilitated by the delivery of multiple and synergistic intervention doses.
The association with lower levels of education and increased GSN website use as well as with high-risk CAI highlights inequalities within sexual health and the importance of understanding MSM as a heterogeneous population with varied sexual health needs; they are suggestive of a range of intervention opportunities. Our finding that around half of YMSM use gay social networking websites and apps simply to browse and kill time is similar to other work,14 and suggests an opportunity for more intensive or interactive HIV risk reduction interventions. For example, individual level, targeted and tailored, online gaming interventions may be worthy of exploration. Our findings suggest these could be targeted towards the population of frequent users with lower levels of education attainment who by their participation within this study do not lack digital literacy. Previous studies have shown the importance of paying attention to retention within internet-based studies,19 but also that it is possible to engage a range of YMSM,7 so it remains vital to ensure the best approaches are trialled and are appropriate for different populations.
Mustanski et al10 reviewed the epidemiological literature on factors associated with sexual risk among YMSM, highlighting the importance of exploring multiple factors across microsystems and macrosystems, due to the importance of personal-level factors, relationship and family factors as well as sociosexual context in sexual risks. It may well be more effective (given that sex is not an individual behaviour) to target wider communities of YMSM. Given our data, along with others, reveal sex-risk behaviours among adolescent MSM, there is a case for earlier intervention to occur, whether online, offline or both. Intervening earlier points towards settings such as schools, to affect proximal and distal factors, which the Mustanski et al10 paper highlights. This could be a further opportunity for research to explore the feasibility of online interventions in partnership with school curriculums, certainly around the areas of knowledge improvement and reduced risk behaviours but also the culture that may inhibit positive behaviours. Although caution is warranted given the limitations of untargeted sex and relationships education,20 and educators being able, whether through supportive curriculum, policies and training, to deliver sexual orientation and gender identity topics—as the Ontario, Canada experience revealed can be challenging.21
Limitations to this study include the cross-sectional nature, which by design prohibits us from inferring causality, and the convenience sampling prohibits generalisability to a wider population or other populations of YMSM. Strengths include the multicountry comparison, even though few differences were found, perhaps indicating a similar culture of use by YMSM across these contexts. We also recruited across multiple sites and apps and enquired about a range of use rather than focusing on one, as some studies have done. In the absence of UK-based data explicitly focusing on YMSM, this study provides an important contemporary snapshot of the patterning of YMSM's use of gay social and sexual networking websites and apps. The patterns of young men's use of social and sexual media, revealing who may be using what, when and why, can direct future research and health promotion in these contexts.
Key messages
Young men who have sex with men (YMSM) were frequent users of gay sexual networking (GSN) websites and apps, with many using them daily or every few days.
Variation in demographic characteristics of GSN websites and app users may affect who interventions are likely to reach, depending on where they are targeted.
Risk for HIV transmission increased with the length of time YMSM reported having used GSN websites, and the frequency of their use.
Acknowledgments
We thank the organisations that helped with recruitment and the men who agreed to take part in the survey. We also thank the three anonymous reviewers of the manuscript.
References
Footnotes
Handling editor Jackie A Cassell
Contributors KL helped to design the study, carried out the analysis, wrote the first draft of the manuscript and led revisions. PF and MD helped to design the study and helped to draft the manuscript. JF designed the study, carried out some of the data analysis and helped to draft the manuscript. All authors read and approved the final manuscript.
Funding NHS Lanarkshire BBV Networks. KL, PF and JF are funded by Glasgow Caledonian University; MD is funded by Monash University.
Competing interests None declared.
Ethics approval Glasgow Caledonian University School of Health and Life Sciences Ethics Subcommittee HLS ID: B11/59.
Provenance and peer review Not commissioned; externally peer reviewed.