Secondary Logo

Journal Logo


Prevalence of Bisexual Behaviors Among Men Who Have Sex With Men (MSM) in China and Associations Between Condom Use in MSM and Heterosexual Behaviors

Lau, Joseph T. F. PhD*; Wang, Ming; Wong, Hong Nei*; Tsui, Hi Yi MPhil*; Jia, Manhong MD; Cheng, Feng PhD§; Zhang, Yun MD, MBA; Su, Xiaoyou MSc*; Wang, Ning PhD

Author Information
Sexually Transmitted Diseases: April 2008 - Volume 35 - Issue 4 - p 406-413
doi: 10.1097/OLQ.0b013e318164467f
  • Free

HIGH-RISK BEHAVIORS AMONG MEN who have sex with men (MSM) have been documented in different parts of the world, including China.1–4 By the end of 2005, 7.3% of all estimated human immunodeficiency virus (HIV) cases in mainland China could be attributed to MSM behaviors.5 The prevalence of bisexual behaviors among MSM varies according to a country’s culture and people’s acceptance of MSM behaviors.6 In many countries, MSM behaviors are highly unacceptable and strong social pressure against MSM behaviors is anticipated. Such social environment may cause MSM to conceal their sexual orientation and to unwillingly engage in some heterosexual relationships.7 These bisexual and marital relationships would have a strong impact on the HIV epidemic as across different countries the prevalence of condom use between spouses and/or regular heterosexual partners tends to be very low.3,8,9 Some researchers argue that this potential route of HIV transmission may explain the HIV epidemic in certain places.10 Furthermore, different types of HIV-related risk behaviors that are practiced by the same individual may be clustered together.11 This concept of clustered risk behaviors may also be applicable in understanding the association between condom use with MSM’s male and female sex partners (FSP).

The traditional Chinese culture does not openly endorse MSM behaviors.12 Many MSM therefore do not inform their FSP about their MSM behaviors.13 Bisexual behaviors are prevalent among Chinese MSM, but this important phenomenon has not been investigated in detail.10 Few intervention programs that target MSM have specifically promoted condom use with FSP.

China’s Yunnan Province has the highest number of reported HIV cases.14 The HIV prevalence of injecting drug users in Yunnan exceeds 50%.14 The prevalence of HIV was found to be 1.3% to 5.45% among MSM in China.15,16 Between 2000 and 2006, the China-UK HIV/AIDS Prevention and Care Project implemented a series of HIV/AIDS interventions and behavioral surveillance studies (BSS) targeting vulnerable populations, including MSM, in Yunnan.3 Kunming, the capital of Yunnan Province, was one of intervention sites that targeted the MSM population. Interventions included free HIV antibody testing service, free condom and lubricant distribution, discounted sexually transmitted disease treatment services, provision of a telephone hotline counseling service, and social activities for peers.

This study investigated the age-specific prevalence of bisexual behaviors and specific marital status among Chinese MSM living in Kunming. The prevalence of condom use with MSM and their heterosexual sex partners were reported. Factors associated with specific marital status, FSP, and condom use with FSP were investigated. The hypothesis that whether condom use with MSM partners is associated with condom use with FSP was also tested.


Study Population and Sampling

The study population was comprised of adult males (aged 15–75) in Kunming, China, who reported ever having engaged in MSM sexual behaviors. Data were obtained from 3 BSS carried out as part of the China-UK HIV/AIDS Prevention and Care Project and were conducted from December 1, 2003 to February 22, 2004, July 18 to August 29, 2005, and April 8 to May 8, 2006. No sampling frame for MSM exists and multiple sampling strategies were adopted to recruit MSM from different sources. Respondents were recruited from venues including gay bars, gay saunas, and parks that MSM use to meet one another. Convenience sampling method was used in such venues to recruit respondents. Nonvenue-based methods were also used, including snowballing referrals made by peer educators and respondents, recruitment from the internet, and recruiting attendees of recreational activities organized by several NGOs. Peer interviewers briefed prospective respondents about the purpose and logistics of the study and answered relevant questions. Respondents of the 2005 and 2006 surveys that self-reported having been interviewed in similar BSS studies were excluded from the data analysis.

Verbal informed consent was obtained from respondents before the anonymous, face-to-face interview commenced. The interviewers signed a form pledging that they had clearly explained the details of the study to the respondents. The response rates of the individual surveys were around 80% (defined as the number of respondents who completed the questionnaire divided by the number of prospective respondents that were invited to join the study). An incentive of 30 Yuan (equivalent to about 4 US$) was given to respondents as compensation for the time spent being interviewed. Ethics approval was obtained from the Ministry of Health in China. Field work was supervised by staff members of the China-UK Project and was coordinated by one of the coauthors (M.W.) and was further monitored by a team of international and national experts. Field and office editing were implemented and logical inconsistency was cross-checked. A pilot study was also conducted. Funding and technical support were provided by the United Kingdom’s Department for International Development and Family Health International.


Background information such as sociodemographic characteristics, mode of recruitment, and year of study was recorded. The level of HIV-related knowledge was assessed by 5 questions that were related to modes of HIV transmission and 3 questions that were related to misconceptions about prevention. These questions have been used in other China-UK Project BSS and intervention studies17 and were selected by a panel of consultants of China-UK project. Respondents were asked whether they had ever received Voluntary Counseling and Testing (VCT) services and whether they had utilized various types of HIV-related services in the past 12 months. The relevant items are listed in the footnotes of Table 1. Respondents were not asked about the result of their HIV antibody testing.

Background Characteristics of Respondents (N = 896)

Several questions were related to bisexual behaviors, such as whether respondents had ever had FSP, whether they had had an FSP in the past 6 months, and whether they had had multiple FSP in the past 6 months. Respondents were also asked both if they had used a condom in the last episode and consistently over the past 6 months in the following 3 scenarios: (a) when having sex with FSP, (b) when having oral sex with MSM partners, and (c) when having anal sex with commercial/noncommercial MSM sex partners. Respondents were further asked whether they perceived homosexuality to be a normal life-style and whether they had experienced much stress arising from their family.

Statistical Analyses

The distributions of the background characteristics were tabulated. Age-specific prevalence of specific current marital status (e.g., currently married, currently divorced/separated) and bisexual behaviors were reported. Age-group differences were compared using univariate odds ratios (ORs) and respective 95% confidence intervals. Factors predicting whether one was currently married and whether one had had FSP in the past 6 months (respective reference categories including never married or never had had FSP) were identified using univariate ORs (and 95% confidence intervals). Multivariate logistic regression models were then fitted, using univariately significant variables as candidate variables. Similar analyses were conducted to identify factors that were associated with condom use (in the last episode and in the past 6 months) when having sex with FSP. Factors considered for univariate and multivariate analyses included background variables, HIV-related knowledge and service utilization variables, and variables related to perceived stress and unprotected MSM sexual behaviors.


Background Characteristics

A total of 896 MSM respondents were recruited from parks, bars, saunas, and nonvenue-based methods. Among all respondents, 33.6% had ever been married, with 18.2% being currently married and 14.9% being currently divorced or separated (Table 1). Their other sociodemographic characteristics are summarized in Table 1. The majority (86.6%) of the respondents answered at least 4 of the 5 questions on HIV-related knowledge correctly, but 30% still gave some misconceived responses (Table 1). Furthermore, 14.7% of all respondents believed that homosexuality was an abnormal life style and 39.8% of them perceived much stress from their family. VCT services were received by 28.1% of the respondents whereas 59.7% of the respondents had received more than 2 types of HIV-related prevention services (Table 1).

Prevalence of Unprotected MSM Sex Behaviors

Of all respondents, 69.1% did not use a condom in the last episode of oral sex with MSM partners in the past 6 months and 76.2% used condoms inconsistently in the past 6 months for such behavior (Table 2). A low percentage of all respondents had had unprotected anal sex with male commercial sex partners (last episode: 2.5%; inconsistent use in the past 6 months: 4.5%). Furthermore, 27.6% of all respondents did not use a condom in the last episode of anal sex with noncommercial sex MSM partners in the past 6 months and 44.5% of all respondents had used condoms inconsistently when having anal sex with such partners in the past 6 months (Table 2). Among those who had had anal sex with commercial sex partners in the past 6 months, 30.8% (40 of 130) were inconsistent condom users; similar figures were 54.7% (399 of 730) for those who had had anal sex with noncommercial sex partners in the past 6 months (Table 2).

Prevalence of Unprotected MSM Sexual Behaviors (Among All Respondents)

Bisexual Behaviors

Among those over the age of 30, 70.8% had been married: 35.7% and 33.8%, respectively of those aged above 30 were either currently married or were divorced/separated (Table 3). Age was strongly associated with whether respondents were currently married, had ever been married, or were currently divorced or separated (OR = 7.39–98.23, P <0.01, Table 3).

Age-Specific Prevalence of Different Marital Status and Bisexual Behaviors

Of all respondents, 58.5% had ever had sex with FSP, with 30.6% having done so in the past 6 months. Similar significant associations between age and the aforementioned bisexual behaviors were detected (OR = 1.64–5.98, P <0.05, Table 3). About 10% (9.8%) of all respondents self-reported having had sex with more than 1 FSP in the past 6 months and such prevalence was significantly lower in the ≥31 age group, as compared with other age groups (OR = 0.57, P <0.05, Table 3).

Condom Use With FSP

Among those who self-reported having had FSP in the past 6 months, only 28.1% used condoms consistently with such FSP. With regard to the last episode of sex with FSP, condoms were only used by 22.9% of those who ever had had sex with FSP and by 40.9% of those who had had sex with FSP in the past 6 months. Those of age ≥31 were less likely than others were to use condoms with FSP (OR = 0.37–0.46, P <0.05, data not tabulated).

Factors Predicting Current Marital Status and Sex With FSP in the Past 6 Months

The multivariate results show that besides older age (OR = 13.88 and 85.24, P <0.05, Table 4), provision of at least one inappropriate response to the 3 questions related to misconceptions about HIV prevention and perceiving homosexuality as an abnormal life style were associated with higher likelihood of being currently married (reference category being those who were never married: OR = 1.65 and 1.76, P <0.05, Table 4). Those with a secondary education or higher were less likely to be currently married, as compared with those having only primary education (OR = 0.41–0.5, P <0.05, Table 4). Four variables (mode of recruitment, year of study, having had anal sex with MSM in the past 6 months, and having unprotected sex in the last episode of anal sex with commercial sex MSM partner in the past 6 months) were univariately but not multivariately significant in predicting whether one was currently married (Table 4).

Factors Predicting Currently Married Status and Female Sex Partnership (vs. Those Who Were Never Married or Those Who Never Have had FSP)

Similarly, older age, perceived stress arising from one’s family and those who had used condoms consistently when engaging in anal sex with male commercial sex partners in the past 6 months (as compared with those without such partners) were associated with higher likelihood in having had sex with FSP in the past 6 months (OR = 1.57–5.88, P <0.05, Table 4). The reverse was true for higher education level (Table 4). Furthermore, respondents surveyed in 2005 and 2006 were more likely than those surveyed in 2004 to have had sex with FSP in the past 6 months (reference category being never had had FSP; OR = 1.95 and 2.43, P <0.05, Table 4). In this case, mode of recruitment and misconceptions with regard to HIV prevention were univariately but not multivariately significant.

Factors Predicting Condom Use With FSP

The results of the multivariate analyses show that among those who had ever had FSP, those who were recruited via gay bars or other channels and those who had ever received VCT were more likely than others to have used a condom in the last episode of sex with their FSP (OR = 1.66–2.73, P <0.05, Table 5). On the other hand, those who had not used a condom in the last episode of oral sex with MSM partners in the past 6 months and those who were inconsistent condom users when having anal sex with male noncommercial sex partners in the past 6 months were less likely than others to have used a condom in the last episode of sex with FSP (OR = 0.48 and 0.43, P <0.05, Table 5). Three variables (age group, those who had the experience of not using a condom in the last episode of anal sex with male noncommercial sex partners in the past 6 months, and consistent condom use during anal sex with male commercial sex partners in the past 6 months) were univariately associated with condom use in the last episode of sex with FSP; these variables however did not remain statistically significant in the multivariate analyses (Table 5).

Factors Predicting Condom Use With FSP

Similar analyses were conducted to predict consistent condom use with FSP (among those with FSP in the past 6 months). Older age (≥31), those who had not used a condom in the last episode of anal sex with male noncommercial sex partners were less likely than others to be consistent condom users when having sex with FSP in the past 6 months (OR = 0.43 and 0.32, respectively, P <0.05, Table 5). Lack of condom use in the last episode of oral sex with MSM partners in the past 6 months was significant in the univariate but not in the multivariate analyses (P >0.05, Table 5).


Around one-third of the MSM respondents had ever been married and the prevalence increased to 70% or greater among those MSM who were over 30 years old. These prevalence were higher than corresponding figures obtained from Western countries.18,19 As a noticeable number of the respondents believed that homosexuality is an abnormal life style and such perception was significantly associated with being currently married, it is argued that a substantial number of MSM might be using marriage as a means to disguise their MSM behaviors or sexual orientation. The very high divorce/separation rates among the MSM suggest that many of those married MSM were unhappy with their marriage, supporting the speculation that many MSM might have married unwillingly under social or familial pressure. China is still a relatively conservative country, and Chinese culture does not render homosexuality as a subject open for discussion.12 Stigmatization against MSM in Chinese societies may be still quite serious.4 Relevant advocacy efforts in China are at most preliminary.

Better-educated MSM were less likely than others to be currently married. These MSM may be better informed and more empowered. Marital relationships have other important implications for HIV prevention as married MSM were less likely to be exposed to preventive messages. Furthermore, the sampled married MSM and those MSM with FSP were more likely than other MSM to indicate misconceptions about HIV prevention.

A high percentage of the MSM respondents also had FSP. The prevalence almost doubled the prevalence of respondents who had ever been or were currently married. Again, social pressure may partially explain the observed high prevalence of bisexual behaviors. For instance, those who perceived much stress arising from their family were more likely than others to currently have FSP. Furthermore, those with FSP in the past 6 months were less likely than were others to have been recruited from gay venues. It is logical that MSM with FSP were less likely than other MSM to visit gay venues, if FSP are assumed to be a means of concealing one’s MSM status.

Older age was strongly associated with both the prevalence of having ever been married and being currently married. This seems natural for heterosexual men but it has particular implication for the MSM population. Following the same reasoning outlined above, social pressure for marriage increases with age as there exists some normative age for getting married in China. Older single men may be seen as gay, which is not socially acceptable. Older MSM and their significant others (e.g., parents) were expected to be more traditional and were more likely to have internalized homophobia. Interestingly, older age was also associated with current divorce/separation status. Age correlates with duration of marriage and spousal conflict may build up over time. The cultural context therefore plays an important role in HIV prevention in the MSM population.

The presence of bridging effects is supported by the very high prevalence of unprotected sex practiced by the respondents both when they were having sex with MSM and FSP. It is found that among bisexual MSM, those who did not use condoms with MSM partners were also less likely to be using condoms with their FSP. This phenomenon of clustered risk behaviors were observed in other multiple risk groups, such as male IDUs who were also engaged in sexual behaviors.11 Those who cared little about HIV/sexually transmitted disease transmission would be more likely to engage in different forms of risk behaviors as the patterns of decision making involved are likely to be very similar. Furthermore, those having commercial MSM partners were more likely than others to have had sex with FSP and about 10% of all MSM respondents had had multiple FSP in the past 6 months. The bridging effect of transmitting HIV from the MSM population to the female population is therefore evident, and is particularly true for older MSM. Although the linkage between MSM behavior and female partnership was very strong, the intensity of the bridge effect is likely to be culturally determined. International comparisons with contextual and cultural considerations are warranted.

Only about 30% of the respondents had received VCT. This study suggests that VCT may be functional to slow down the aforementioned bridging effects as having VCT was negatively associated with unprotected sex with FSP. In the literature, the evidence on effectiveness of VCT has been mixed.20,21 Exposure to other HIV prevention services has not associated condom use with FSP. Current HIV prevention work for MSM may have overlooked the need for protecting MSM’s FSP from HIV transmission. Such efforts should especially target older MSM, as they were more likely to have bisexual behaviors.

The study has a number of limitations. First, convenience sampling method was used as random sampling was not feasible. Many published MSM studies have used similar recruitment methods.3,22 Second, responses may be subjected to reporting bias because of social desirability. The direction of such bias, however, would underestimate the prevalence of unprotected sex. Third, we did not directly ask about reasons for getting married and/or engaging in bisexual behaviors. An assumption was made that such behaviors were related to social pressure against their MSM status. Furthermore, bisexual commercial sex behaviors had not been investigated in this study. Some of these limitations are because of the constraint on the length of the questionnaire. Biologic markers were not included in the BSS, as taking blood samples at survey venues would have made recruitment more complicated. Additional questions and measurement of biologic markers should be considered in future studies.

Only a few studies have briefly mentioned bisexual behaviors in China.23–26 These studies, however, did not focus primarily on bisexual behaviors. MSM condom use in association with FSP was also not investigated by any these studies. This study highlights an important area that warrants HIV prevention and research work. We argue that some of the FSP relationships were “not necessary” and/or “not desired,” having been formed under social pressure. The number of women exposed to HIV-related risk via their MSM partners could have been decreased if social pressure was alleviated. Therefore, social pressure such as stigmatization against MSM partially accounts for the intensity of aforementioned bridging effects. This study demonstrates that social and cultural context is an important determinant of HIV transmission.27


1. Elford J, Hart G. If HIV prevention works, why are rates of high-risk sexual behavior increasing among MSM? AIDS Educ Prev 2003; 15:294–308.
2. Mansergh G, Naorat S, Jommaroeng R, et al. Inconsistent condom use with steady and casual partners and associated factors among sexually—Active men who have sex with men in Bangkok, Thailand. AIDS Behav 2006; 10:743–751.
3. Choi KH, Gibson DR, Han L, et al. High levels of unprotected sex with men and women among men who have sex with men: A potential bridge of HIV transmission in Beijing, China. AIDS Educ Prev 2004; 16:19–30.
4. Lau JTF, Kim JH, Lau M, et al. HIV-related behaviors and attitudes among Chinese men who have sex with men in Hong Kong: A population-based study. Sex Transm Infect 2004; 80:459–465.
5. Ministry of Health, People’s Republic of China, UNAIDS, & WHO. 2005 Update on the HIV/AIDS Epidemic and Response in China. Beijing, China: National Center for AIDS Prevention and Control, 2006.
6. Ross MW, Essien EJ, Williams ML, et al. Concordance between sexual behavior and sexual identity in street outreach samples of four racial/ethnic groups. Sex Transm Dis 2003; 30:110–113.
7. Khan SI, Hudson-Rodd N, Saggers S, et al. Men who have sex with men’s sexual relations with women in Bangladesh. Cult Health Sex 2005; 7:159–169.
8. Hernandez AL, Lindan CP, Mathur M, et al. Sexual behavior among men who have sex with women, men, and Hijras in Mumbai, India—Multiple sexual risks. AIDS Behav 2005; 10(suppl 7):5–16.
9. Folch C, Marks G, Esteve A, et al. Factors associated with unprotected sexual intercourse with steady male, casual male, and female partners among men who have sex with men in Barcelona, Spain. AIDS Educ Prev 2006; 18:227–242.
10. He Q, Wang Y, Lin P, et al. Potential bridges for HIV infection to men who have sex with men in Guangzhou, China. AIDS Behav 2006;10(suppl 4):S17–S23.
11. Lau JTF, Cheng F, Tsui HY, et al. Clustering of syringe sharing and unprotected sex risk behaviors in male injecting drug users in China. Sex Transm Dis 2007; 34:574–582.
12. Liu JX, Choi K. Experiences of social discrimination among men who have sex with men in Shanghai, China. AIDS Behav 2006; 10:S25–S33.
13. Weatherburn P, Hickson F, Reid DS, et al. Sexual HIV risk behaviour among men who have sex with both men and women. AIDS Care 1998; 10:463–471.
14. Ministry of Health China. 2005 Update on the HIV/AIDS Epidemic and Response in China. Beijing: Ministry of Health China, UNAIDS, WHO, 2006.
15. South China Morning Post. First Official Survey Puts Gay Male Community at 5–10 m. South China Morning Post, December 2, 2004:A4.
16. Choi KH, Liu H, Guo YQ, et al. Emerging HIV-1 epidemic in China in men who have sex with men. Lancet 2003; 361:2125–2126.
17. Lau JTF, Wang R, Chen H, et al. Evaluation of the overall program effectiveness of HIV-related intervention programs in a community in Sichuan, China. Sex Transm Dis 2007; 34:653–662.
18. Earl WL. Married men and same sex activity: A field study on HIV risk among men who do not identify as gay or bisexual. J Sex Marital Ther 1990; 16:251–257.
19. Stanekova D, Habekova M, Wimmerova S, et al. HIV infection and sexual behavior among homosexual and bisexual men in Bratislava. Cent Eur J Public Health 2000; 8:172–175.
20. The Voluntary HIV-1 Counseling and Testing Efficacy Study Group. Efficacy of voluntary HIV-1 counseling and testing in individuals and couples in Kenya, Tanzania, and Trinidad: A randomized trial. Lancet 2000; 356:103–112.
21. Landis SE, Earp JL, Kock GG. Impact of HIV testing and counseling on subsequent sexual behavior. AIDS Educ Prev 1992; 4:61–70.
22. Wade AS, Kane CT, Diallo PA, et al. HIV infection and sexually transmitted infections among men who have sex with men in Senegal. AIDS 2005; 19:2133–2140.
23. He Q, Wang Y, Lin P, et al. Potential bridges for HIV infection to men who have sex with men in Guangzhou, China. AIDS Behav 2006;10(suppl 1):17–23.
24. Choi KH, Ning Z, Gregorich SE, et al. The influence of social and sexual networks in the spread of HIV and syphilis among men who have sex with men in Shanghai, China. J Acquir Immune Defic Syndr 2007; 45:77–84.
25. Liu H, Yang H, Li X, et al. Men who have sex with men and human immunodeficiency virus/sexually transmitted disease control in China. Sex Transm Dis 2006; 33:68–76.
26. Wong WCW, Zhang J, Wu SC, et al. The HIV related risks among men having sex with men in rural Yunnan, China: A qualitative study. Sex Transm Infect 2006; 82:127–130.
27. Bajo N. Social factors and the process of risk construction in HIV sexual transmission. AIDS Care 1997; 9:227–237.
© Copyright 2008 American Sexually Transmitted Diseases Association