From January to April 2019, the study was conducted in 4 regions with large populations of college students in China, including Chongqing in southwest China, Guangdong in south China, Shandong in east China, and Tianjin in north China. Individuals were eligible for participation if they were male; aged 16 years or older; university student (technical diploma and undergraduate students); MSM (having sexual behaviors including mutual masturbation, oral sex, or anal sex with male partners); HIV negative or unknown; and willing to provide informed consent. Each participant received free HIV testing services and 50 Chinese Yuan (~US $7.9) as compensation after enrollment. All data were stored anonymously in encrypted storage space, and personally identifiable information was stored separately from other behavioral and test result data.

The study has been approved by the Ethics Review Board of Sun Yat-sen University (SYSU-SPH2018044) and registered with the Chinese Clinical Trial Registry (Registration number: ChiCTR1900020645).

For sampling and recruitment, the methodology has previously been published elsewhere [18]. Briefly, recruitment of participants was promoted both online and offline. For online recruitment, study advertising campaigns were placed on the social network. For offline recruitment, the study advertisement was distributed to the collaboration of community-based organizations (CBOs), health clinics, and student associations. After reading study advertisements online or offline, participants could make an appointment to join the study at a preferred CBO. After scanning the QR code on a smartphone, participants finished the online questionnaire at the CBO site. A self-administrated online structured questionnaire was used to collect each participant’s sociodemographic information (including age, ethnicity, educational status, involvement in class leadership, academic performance, student loans, sexual orientation, and gender of sex partner in lifetime), sexual behaviors (including the age at first anal intercourse, commercial sex in lifetime, forced sex in lifetime, group sex in the past 3 months, number of sex partners in the past 3 months, condomless sex in the past 3 months, and HIV serostatus of sex partners in the past 3 months), gay app use, recreational drug use, and HIV testing. Gay app use was assessed by the question, “Have you ever used any gay app?”, and subsequent questions about specific gay app use behaviors, including preference, duration, frequency, and purpose. We also assessed seeking sex partners via gay apps by the question “During the past 3 months, have you ever used gay apps to seek sex partners?”, and subsequent questions including the number of sex partners, group sex, condomless sex, and HIV serostatus of sex partners. Recreational drug use was assessed by asking “Have you ever used any of the below-listed recreational drugs during sex? Popper (alkyl nitrites), ecstasy, crystal methamphetamine, and ketamine.” The HIV testing was conducted with rapid HIV testing (Alere Determine HIV-1/2; Alere Medical Co., Ltd.) at the location of CBOs; those who tested positive were referred to the local Centers for Disease Control and Prevention (CDC) for confirmation testing and subsequent treatment services [19]. HIV infections in this study were confirmed results.

We used mean and SD to describe continuous variables, and proportions to describe categorical variables. Chi-square test for trend was used to test the linear trend in the proportion of engaging in sexual risk behaviors with the increase in the frequency of app use. Univariable log-binomial regressions were performed to assess correlates (sociodemographic variables, sexual behaviors) of seeking sex partners via gay apps. Multivariable log-binomial regressions were conducted to use seeking sex partners via the gay apps as the independent variable and sexual behavior in past 3 months (including group sex, number of sex partners, condomless sex, and HIV status of sex partners) as the dependent variable. Each variable of sexual behaviors was examined independently in separate regression models, adjusted for sociodemographic variables. Variables with P<.05 were retained in the final model. Results were presented with an adjusted prevalence ratio (APR) and 95% CI. All data were analyzed using R (version 3.6.1; R Foundation).

A total of 447 MSM attending university with an average age of 20.4 (SD 1.5) years were recruited in the study (Table 1). The majority of participants self-identified as gay (382/447, 85.5%). Most participants were undergraduate (355/447, 79.4%) students and self-reported academic performance above the average (354/447, 79.2%).

Among MSM attending university enrolled in the study, 439/447 (98.2%) reported ever using gay apps (Table 1). The most popular gay apps were Blued (428/439, 97.5%) and Aloha (226/439, 51.5%); likewise, the most frequently used gay apps in the past 3 months were also Blued (318/439, 72.4%) and Aloha (84/439, 19.1%). Most participants reported using only 1 type of app (393/439, 89.5%), over 12 months of usage time (354/439, 80.6%), and using less than 1 hour per day (362/439, 82.5%). The top 3 purposes for using gay apps were looking for a romantic relationship (199/439, 45.3%), following others’ updates (185/439, 42.1%), and making friends (171/439, 39.0%). The majority of participants (197/240, 82.1%) reported engaging in anal sex with sex partners met via apps and over half (141/240, 58.8%) did not know the HIV status of last sex partner.

The frequencies of gay app use were months (125/439, 28.5%), weeks (173/439, 39.4%), and days (141/439, 32.1%) in the past 3 months among MSM attending university (Table 1). Chi-square test for trend showed that higher proportions of sexual behaviors, including seeking sex partners via apps (P<.001), engaging in group sex (P<.001), having multiple sex partners (P<.001), unknown sex partners’ HIV status (P<.001), and using recreational drugs during sex (P<.02), were positively associated with the increase in the frequency of gay app use (Figure 1).

Table 2 shows the characteristics of seeking sex partners via gay apps among MSM attending university. As many as 240 reported using gay apps to seek sex partners. In multivariable analysis, participants who used gay apps to seek sex partners were more likely to have multiple sex partners in the past 3 months (APR 1.53, 95% CI 1.33-1.76, P<.001), engage in group sex in the past 3 months (APR 1.55, 95% CI 1.35-1.78, P<.001), and have sex partners with unknown or positive HIV status (APR 1.72, 95% CI 1.46-2.01, P<.001).

To our knowledge, this is the first study to characterize gay app use among MSM attending university in China. We found that almost all MSM attending university used gay apps, and those who used gay apps to seek sexual partners were likely to be more engaging in high-risk sexual behaviors, including multiple sexual partners, group sex, commercial sex, and recreational drug use during sex.

We found a high prevalence (439/447, 98.2%) of gay app use among MSM attending university, which was much higher than that in nonstudent MSM (ranging from 40% [20] to 60% [21,22]). Previous studies have shown the differences in age distribution for app use among MSM, with younger age groups using them more frequently [21,23]. The mean age of participants in our study was 20 (SD 1.5) years, indicating that MSM attending university were a younger subgroup of MSM. Therefore, they are more likely to use gay apps compared with nonstudent MSM.

The popularity of gay app use among MSM attending universities was accompanied by some high-risk behaviors. Our study found that high-risk behaviors, including engaging in group sex in the past 3 months, having 2 or more sex partners in the past 3 months, failing to know sex partners’ HIV status in the past 3 months, and using recreational drugs during sex, increased with the frequency of gay app use. Previous studies have indicated that men using gay apps could facilitate sexually risky behaviors [7,22,24]. Moreover, MSM attending university who used gay apps to seek sex partners were more likely to have multiple sex partners, engage in group sex and commercial sex, and use recreational drugs. This is consistent with previous studies among nonstudent MSM in China that found gay app users had more sex partners and more group sex compared with those who did not use apps to seek sex partners [20,22,25]. Notably, high-risk behaviors, including multiple sex partners, group sex, and commercial sex, could increase the risk of infection with HIV or other sexually transmitted infections. Therefore, additional interventions are needed to address the sexual risk behaviors associated with seeking sex partners via apps.

Literally all MSM attending university have used gay apps at some point, which points to the potential feasibility of using gay app–based platforms to engage them in HIV prevention and care programs. Barriers, such as stigma, discrimination, and disclosure of sexuality, prevent young MSM from seeking traditional venue–based HIV prevention services. Because social networking apps are very popular among young people [26,27], there may be opportunities to facilitate HIV prevention efforts via these apps. Some app-based HIV interventions achieved desirable results. For example, since an HIV testing campaign using app push messages to promote HIV testing has been launched on Blued in 2015, over 15,000 MSM have tested for HIV, of whom over 2000 self-identified as university students. The number of those who tested for HIV increased substantially from 836 in 2013 before the campaign to 7315 in 2017 after the campaign [12]. Given that gay apps can instantly reach millions of targeted individuals, app-based interventions include information on HIV self-testing, pre-exposure prophylaxis, post-exposure prophylaxis, and HIV treatment as prevention, and thus may benefit a broad group in the MSM community.

We found Blued and Aloha were the most popular gay apps among MSM attending university, and the majority of those reported using only 1 app (393/439, 89.5%). There were 43 brands of gay apps in China in 2018 [28], and among nonstudent MSM 65.9% used only 1 app, and 23.2% used 2 apps [21]. Currently, there are regional differences in gay app use. While English language–based gay apps such as Grindr and Jack’d have subscribers from many countries, the gay app market in China is dominated by local apps such as Blued and Aloha. Those differences in gay app choices may be attributed to linguistic and cultural differences among MSM. Blued and Aloha are also platforms for HIV care services.

There were limitations. First, there may be a selection bias in the study population. Participants were recruited mainly through online advertisements and enrolled offline for HIV testing at CBO site, which may exclude student MSM who hide their sexual orientation and were not willing to attend on-site testing at CBO. This deeply hidden population may be frequent users of gay apps, and the lack of such a population among participants may limit the generalizability of the study. Second, the study results may be affected by social desirability reporting bias because sensitive private information was based on self-reporting by participants. Third, the design of the cross-sectional study prevents making causal inferences between using an app to seek sex partners and risk behaviors.

Our study suggests high gay app use among MSM attending university and using apps to seek sex partners may be associated with increased high-risk sexual behaviors. The causality between seeking sex partners via gay apps and increased high-risk sexual behaviors should be further investigated so as to inform potential policies for HIV prevention.