Objectives The expansion of gay sex-seeking application (gay app) use among men who have sex with men (MSM) may create new virtual risk environments that facilitate STI transmission. The goals of this study were to compare sexual behaviours between gay app users and non-users, and to describe sexual behaviours among gay app users in China.
Methods In October 2014, we recruited MSM from three Chinese gay websites. Data on sociodemographics, sexual behaviours and gay app use were collected. Logistic regressions were used to compare gay app users with non-app users and to identify factors associated with condomless sex among gay app users.
Results Of the 1424 participants, most were <30 years old (77.5%), single (83.8%) and self-identified as gay (72.9%). Overall, 824 (57.9%) had used gay apps for partner-seeking in the last 6 months. Among gay app users, 36.4% met their last partner within 24 hours of first message exchange through gay apps, and 59.0% negotiated condom use before in-person meeting. Compared with non-users, gay app users reported engaging in more condomless sex in the last 6 months (adjusted OR (aOR) =1.52, 95% CI 1.19 to 1.94) and more group sex (aOR =1.49, 95% CI 1.02 to 2.18). Negotiating condom use before in-person meeting was positively associated with condom use with partners met through gay apps (aOR =1.83, 95% CI 1.29 to 2.60).
Conclusions Gay apps are linked to risky sexual behaviours and may foster a virtual risk environment for STI transmission among Chinese MSM. App-based interventions could target young gay man and facilitate condom negotiation.
- SEXUAL BEHAVIOUR
- GAY MEN
Statistics from Altmetric.com
Key populations, which include men who have sex with men (MSM), transgender individuals, persons who inject drugs, sex workers and people in prisons and other closed (institutional) settings, account for more than half of all new HIV infections in many low-income and middle-income countries (LMICs).1 Many MSM in LMIC contexts face harsh local sanctions and cultural stigma and, for a number of reasons, turn to MSM-specific smartphone-based sex-seeking applications (gay apps) to find new sex partners.2 Gay app use around the world has rapidly expanded over the past 5 years.3
Gay apps use and its associated patterns may influence sexual risk behaviours. Gay apps' geospatial technology facilitates rapid sex partner identification regardless of time or location. Gay app use has been associated with an increased number of sex partners4 and condomless sex in the USA.4–6 However, other data suggest that sex-seeking gay apps and other gay-specific apps may increase social support,7 facilitate HIV service delivery and decrease risk behaviours.8–10
Although a number of studies have examined gay app use and sexual behaviours in high-income settings, there are limited data from low-income and middle-income settings.2 China presents a unique opportunity to examine the influence of gay app use because the adoption of gay app use has been relatively widespread2 and swift.11 Virtually non-existent 5 years ago, approximately 40% of gay men reported using gay apps to find sex partners in 2013 and the largest locally developed gay app reported approximately 7 million new users last year alone.11
Despite the importance of gay apps, we have very limited knowledge on specific patterns of gay app use and their relationship with high-risk sexual behaviours. Therefore, we conducted a cross-sectional survey among Chinese MSM to: (1) compare sexual risk behaviours between gay app users and non-users; (2) describe specific patterns of gay app use among users and (3) identify factors associated with condomless anal sex among gay app users.
Materials and methods
We conducted an online survey in September–October 2014 among Chinese MSM. We hosted the survey on three large popular gay web portals in Northern China (http://www.danlan.org), Southern China (http://www.yntz.net) and Eastern China (http://www.jstz.org). These web portals provide MSM-specific news, social networking opportunities and advertisements. These three web portals were selected because of their geographically diverse users and high web traffic. Together they reach an estimated 90 000 unique daily users according to internal analytics.
Survey development and recruitment
We interviewed MSM and other key stakeholders to inform development of our online survey. In addition, our draft survey was reviewed by local MSM who had previously participated in online surveys, local community-based organisation leaders and staff, social media experts and physicians and public health experts.12 ,13 Prior to final survey launch, we piloted the survey with 144 individuals on one of the web portals (data not included in final analysis). We followed a checklist for reporting results of Internet e-surveys (CHERRIES) throughout the process to improve quality and reporting of our web survey.14 Participants were recruited via banner links on each web portal homepage as well as an announcement sent to registered users. After clicking the banner, participants were directed to the informed consent and eligibility screening questions. Eligibility criteria included being assigned male sex at birth, 16 years of age or older, engaged in anal or oral sex with a man in the previous year, willing to provide their cell phone number and completed the informed consent process. At the end of our survey was an additional section investigating the efficacy of HIV testing promotion messages. Participants eligible for the additional section (never tested for HIV) received a small (US$10) phone card reimbursement, but were not aware of eligibility criteria before the study. The phone number was used to both screen for duplicates as well as provide reimbursement to eligible participants.
Sociodemographic information collected in the survey included age, occupation, living situation, marital status, education and income. Participants were asked about their sexual orientation as well as the gender of their current sexual partners, number of sexual partners in the last 6 months and history of receptive or insertive condomless anal intercourse with male partners and condomless sex with female partners. They were also asked about participation in group sex and commercial sex.
Participants were asked whether they used gay apps to find sex partners in the last 6 months. If participants reported any use of gay apps to find sex partners in the last 6 months, they were categorised as gay app users, while others were categorised as non-users. Those who used gay apps were asked about behaviours with sex partners they met through the applications, including number of partners, condom use, time from initial conversation to in-person meeting and location of meeting. They were also asked whether they negotiated condom use and discussed HIV status with their sex partner before meeting via gay apps. Here, we defined condom negotiation as talking about condom use before sex. If participants did not negotiate condom use via gay apps, they were asked whether condom negotiation occurred in-person. In our study, gay app partners were defined as male partners that were found via gay apps.
Descriptive analysis was performed to describe sociodemographics, risk behaviours and patterns of gay app use. Univariate and multivariate logistic regression were used to compare gay app users and non-users, while demographic characteristics, including age, residence, education, marital status and income were adjusted for in the multivariate logistic regression models. Logistic regressions (univariate and multivariate) were also used to compare gay app users who did not use a condom during last anal intercourse with partners found through gay app and those who did, adjusting for age, residence, education, marital status and income in the multivariate logistic regression models.
Ethics review committees in China (Guangdong Provincial Center for Skin Diseases and STI Control), and the USA (University of North Carolina at Chapel Hill (14-1685) and the University of California, San Francisco (14-14877)) approved the study prior to launch.
In the 2 months of survey availability, the links to the survey were clicked a total of 5339 non-unique times. Of these clicks, 2192 withdrew before screening for eligibility (1536 withdrew prior to reading the consent form and 656 did not sign the consent form). In addition, 1328 of those who clicked did not meet the eligibility criteria and 395 were excluded due to duplicates (same phone number). Overall, a total of 1424 eligible MSM from 290 cities within 32 provinces finished the online survey.
Demographics and behaviours
Of the 1424 participants, 14.5% (206) were <20 years old, 63.1% (898) were between 20 and 29 years old and the majority (88.9%) resided in an urban area. In addition, 68.0% (969) had a college degree, while another 6.0% (86) had a graduate degree or higher. Over four-fifths (83.8%) of participants were never married, two-fifths (41.3%) were students and about three-quarters (74.0%) had an annual income over US$3000 (table 1).
Among survey participants, 48.5% (691) reported that they currently have a main partner and 29.1% (414) had ever engaged in vaginal or anal sex with women. Among those who ever engaged in sex with women, 44.0% (182/414) had reported condomless sex with women in the last 6 months; 29.6% (421) of participants reported engaging in condomless anal sex with their last male partner in the last 6 months. When engaging in anal intercourse, 36.8% (524) of participants preferred insertive, 43.6% (621) preferred receptive and the rest 19.6% (279) had no preference.
Patterns of gay app users
In the last 6 months, 824 (57.9%) participants used a gay app to seek sex partners. Among gay app users, 337 (40.9%) also found partners using other online media. Of the gay app users, 334 (40.5%) started to use gay apps within the last year, and 398 (48.3%) have been using gay apps for 1–3 years (table 2).
About one-third (32.3%) of gay apps users met four or more sex partners via gay apps in the last 6 months. Approximately one-quarter (23.7%) reported six or more receptive anal sex acts with partners met through a gay app in the last 6 months, and about one-quarter (25.5%) had six or more insertive anal sex acts with partners met through gay app in the last 6 months.
For the last gay app partner, 36.4% reported meeting in person within 24 hours of initiating contact on a gay app, and 38.5% met within 2–7 days. About half of them had sex in a hotel.
Before meeting in person, 59.0% of participants reported negotiating condom use with their last gay app partners. Among those who did not discuss condom use via gay app, 32.3% talked about condom use in person. About one-quarter (25.4%) gay app users did not use condom during last anal sex with the last partner met through gay app. Only one-third of gay app users discussed HIV status with their last gay app partner before meeting in person.
Factors associated with gay app use
Participants who had never engaged in sex with women were more likely to be gay app users (crude OR =1.85, 95% CI 1.47 to 2.32). Compared with non-users, gay app users were more likely to have condomless anal intercourse in the last 6 months, with crude OR of 1.31 (95% CI 1.04 to 1.66) (table 1).
After adjusting for age, residence, education, marital status and income, we found gay orientation, never having engaged in sex with women, and reported condomless anal intercourse in the last 6 months were still positively associated with gay app use, with adjusted ORs (aOR) of 1.30 (95% CI 1.01 to 1.66), 1.49 (95% CI 1.12 to 1.96) and 1.52 (95% CI 1.19 to 1.94), respectively. In addition, gay app users were more likely reported engaging in group sex, with aOR of 1.49 (95% CI 1.02 to 2.18) (table 1).
After adjusting for age, residence, education, marital status and income, we found gay app users who reported more than 20 receptive anal sex acts with partners met through a gay app in the last 6 months were more likely to report condomless anal intercourse with their last gay app partner. Negotiating condom use before meeting in person was associated with reporting condom use during last anal sex with gay app partner (aOR =1.83, 95% CI 1.29 to 2.60). However, no correlation was identified between in-person condom negotiation and subsequent condom use (aOR =0.84, 95% CI 0.58 to 1.21). In addition, having sex at home or at venues were also associated with a higher likelihood of reported condomless anal intercourse during last anal sex with the last gay app partner, compared with having sex at hotels, with aORs of 1.65 (95% CI 1.15 to 2.38) and 1.92 (95% CI 0.99 to 2.60), respectively (table 3).
We found a differential pattern of behaviour based upon time to meeting partner on the gay application. Individuals who met their partners in <1 week after initiating conversations on the application were more likely to report multiple partners in the last 6 months (78.3%) when compared with individuals who took >1 week (78.3 vs 56.5%, p<0.05, table 4). We also found 28.4% gay app users who met their partners in-person within 1 week after online meeting asked the sero-status of their last gay app partners, versus 47.8% (p<0.05) for those meeting partners longer than 1 week after online meeting (table 4).
Sexual behaviours based on the means of meeting partners were also compared (see online supplementary appendix A). Rates of reported condom use with last sexual partner met via gay application were similar to those of last partner met via websites. There were higher reported rates of condomless sex with last partners met via in-person venues (see online supplementary appendix A).
The increasing use of gay apps among Chinese MSM reduces many barriers for MSM to meet new sex partners. Most of the literature about gay app use among MSM has come from high-income contexts.2 ,15 Our study extends the existing literature by recruiting participants from many big cities throughout China, not requiring in-person disclosure of sexual orientation to participate, and gathering detailed gay app data. Our findings indicate that gay app users report more high-risk behaviours than non-users, which may correlate with increased disease transmission among gay app users.
We found that gay app users were more likely to report condomless sex. This was consistent with some literature,16 but contrasted our earlier research in China.2 A potential reason for this contrast is that the population of gay app users is evolving and higher numbers of men with high-risk behaviours have become users.17 For example, in 2013, only about 40% MSM reported that they were gay app users, while this proportion increased to 58% in 2014. China's most popular gay app has grown to 15 million users in only 2 years.11 Another potential reason is the behaviours of the users may also change over time. Perhaps users exercised a greater degree of caution when first using the application and participate in higher risk behaviours as app use becomes normalised within the community. We did not collect information on sexual behaviours before and after the use of gay apps. Future studies should investigate temporal behaviour changes among app users.
Our study suggests that men who reported negotiating condom use through gay apps were more likely to use a condom. This finding is consistent with literature that reports enhanced communication about condom use is associated with condom use.18 ,19 We did not find any relationship between in-person condom negotiation and subsequent condom use, suggesting the importance of condom negotiation on gay apps. More people negotiated about condom use before met in person may reflect how men might find condom negotiation easier via an app rather than face-to-face. Coordinating efforts with gay app companies to encourage regular discussion about condom use or other harm reduction behaviours before in-person meetings might be useful. Several gay app companies have taken a more public health-focused approach (Hornet), but have not been as effective in drawing and retaining gay app users over time as conventional gay apps in China. It is unclear whether the decreased effectiveness is due to their public health messages, specific app features or it not being a domestic (mainland China-based) application.
Our results indicated that the time from online meeting to in-person meeting was often in 1 week. The short duration between online meeting and in-person meeting may limit communication between gay app users in regards to discussing HIV sero-status and condom use. Shorter duration between meeting online and in person was associated with less discussion about HIV status. The short duration between online meeting and in-person meeting may also facilitate increased number of partners for gay app users. We found evidence for this trend in our study, finding that a higher proportion of gay app users who met their last gay app partners in-person within 1 week of initiating conversation (vs >1 week) have two or more gay app partners in last 6 months. Combined with the higher condomless anal intercourse rate, these features may foster a virtual environment for gay app users, and in turn facilitate HIV transmission.
Our study has several limitations. First, we only recruited online gay men who tend to be younger and more highly educated20 compared with non-online gay men in China. However, national data21 suggest that young gay men have a disproportionate burden of syphilis and HIV. Second, all the behaviours measured in our study were based on self-report. This may result in a social desirability reporting bias, especially given stigma surrounding HIV and MSM behaviours. However, our survey did not collect personal information except participants' cell phone numbers for reimbursement, although collecting phone number may also put some men off participating, especially those who did not identify as gay. Third, a large number of MSM who clicked the survey link withdrew before being screened for eligibility, which may have resulted in selection bias. Fourth, the failure to distinguish between partner types—and particularly to distinguish between whether they were in a monogamous ongoing relationship or not is also an important limitation of our study. A recent study of Hong Kong app users found relationship status to be a predictor of condom use.22 Finally, this is a cross-sectional survey and cannot identify causality. Regardless, our study identifies app users as a population of increased sexual risk for future health campaigns.
Gay apps have become a widely accepted part of gay life in China. Our survey data suggest that gay apps foster a virtual environment that is associated with condomless sex and other high-risk behaviours in China. We found specific features and patterns of gay app use that were correlated with higher rates of risky sexual behaviours. Longitudinal prospective studies that target how gay app use shifts behavioural norms of MSM are needed. In addition, policy makers, researchers and gay app companies should work together to explore the role of gay apps in facilitating HIV and other STI transmission among MSM and to also explore its role in facilitating HIV/STIs prevention. Further implementation research on using the gay apps of MSM may be useful for preventing STI and for promoting other STI services such as partner services.
Gay sex-seeking applications (gay apps) are widely used by Chinese men who have sex with men who participated in an online survey.
Compared with non-users, gay app users reported engaging in more condomless sex in the last 6 months and reported engaging in more group sex.
Gay app users who did not report negotiating condom use before meeting in person reported more condomless anal sex.
We thank all the study participants and staff members at SESH Global, Danlan, Jiangsu Tongzhi, Yunnan Tongzhi and the Guangdong Provincial Centers for Skin Diseases and STI Control who contributed. We thank Dr Heping Zheng, Dr Peter Vickerman, Dr Paul Volberding, Dr Kate Mitchell, Dr Kate Muessig, Larry Han, Dee Poon and Gang Meng for their support. A complete list of acknowledgements is available in the protocol.
Abstract in Chinese
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
- Abstract in Chinese - Online abstract
WT and JB contributed equally and are co-first authors.
Handling editor Jackie A Cassell
Contributors WT and JB participated in all stages and wrote the manuscript, YZ and F-yL helped collect the data; LST, SH, BY, CW and JDT helped design the study and reviewed the manuscript.
Funding JDT and CW are receiving grants (NIAID 1R01AI114310, R00MH093201 and FIC 1D43TW009532) from National Institutes of Health, WT and F-yL is receiving grants (5R25TW009340) from UJMT Fogarty Fellowship and JB is receiving grants (R25TW0093) from UJMT Fogarty Fellowship.
Competing interests None declared.
Patient consent Obtained.
Ethics approval Guangdong Provincial Center for Skin Diseases and STI Control, University of North Carolina at Chapel Hill and the University of California, San Francisco. IRB approval number 11413.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.