China is moving from a drug-driven to a bimodal drug/sex-driven HIV/AIDS epidemic.1^–4 The proportion of reported sexually transmitted HIV infections (STI) increased from 10.9% in 2002 to 43.6% in 2005. Among the estimated 70 000 new cases in 2005, 49.8% involved sexual transmission, outnumbering intravenous drug use (48.6%). Crucial to this latest sexual transmission epidemic is the rise of HIV infection among men who have sex with men (MSM), from 0.2% of reported HIV cases in 2001 to 7.3% in 2005.5 6 MSM may soon become the third highest risk group, after intravenous drug users and female sex workers.

MSM studies in China are limited, given difficulties in accessing this stigmatised hidden population and their late inclusion in sentinel surveillance in 2003.2 3 7 8 Surveys report high levels of unprotected sexual practices among MSM,7^–13 with reported HIV prevalence from 1.2% to 3.4%.4 7 9 10 12 There is evidence that MSM HIV transmission may bridge to the general population, as most MSM are married or bisexual and have sex with women.2 3 7 11 13

Male sex workers (MSW) who serve men in China are most frequently called “money boys” (or “yazi” (“duck”) and “haizi” (“child” or “son”)). As an MSM subpopulation, they are of special concern.14^–16 Their number is unknown, although they may constitute 4.9% to 24% of all MSM.8^–11 Studies portray MSW as young, heterosexual, undereducated rural male migrants who provide sexual services solely for money. They are perceived as the highest risk group among the MSM population, most likely to engage in high-risk behaviour.7 8 11

International studies of MSW suggest that subjective meanings, issues of power, control over interaction with clients and other contextual factors are crucial for understanding MSW sexual practices. Position in the industry, length of working, age, migrancy, financial and educational status, drugs and alcohol intake, work and sexual identities have been linked to unsafe sex among MSW.17^–21

This research identifies risk factors affecting MSW sexual practices in Beijing and Shanghai in China. These cities (Beijing, population 15.36 million; Shanghai, population 17.78 million in 2005; National Bureau of Statistics of China, 2006) are believed to be home to the largest MSM/MSW populations and are the most popular migrant destinations in China.13 16 One study estimated that 4000 to 10 000 male rural migrants offered same-sex sexual services in Beijing.13

METHODS

Thirty Chinese MSW were interviewed within four field trips during 2004–5. Participants were Chinese men living in either Beijing (n  =  14) or Shanghai (n  =  16); at least 18 years old and reported having had sex with at least one other man in exchange for money in the previous 6 months. Informants were recruited through referral from a non-governmental organisation (NGO) with a strong MSM network by which their outreach team members were all identified as gay and had established rapport with participants (n  =  10). More participants were obtained through the snowball technique with managers of male brothels with whom the NGO had contact (n  =  15) and with participating workers (n  =  5).

Respondents participated in face-to-face, semi-structured, in-depth interviews that addressed topics concerning their work and private lives in relation to their reported safer sex practices—particularly condom use for anal/vaginal intercourse. Respondents, who had not previously been interviewed by other studies, were informed that the study was funded by UNAIDS and an academic institute. They were guaranteed confidentiality and anonymity and were told of their right to end the interview at any time if they felt any psychological discomfort. Their participation was voluntary with no direct benefits except honorariums of RMB300 (approximately UK£20). Workers usually charged RMB500 (UK£30) for commercial sexual transactions. Consent was obtained from respondents before interviews. The author conducted the interviews in hotels, the NGO’s office, brothels or workers’ homes.

Taped interviews in Mandarin Chinese, of half to 2 h, were transcribed into written Chinese. Data analysis included identifying themes, building codebooks, marking texts, constructing models and testing models against empirical data.22 First-level coding involved several readings of the transcripts. Coding themes, informed by existing research, were revised when new themes emerged from coding. Second-level coding involved comparisons across three coders (the author and two research assistants), clarifying potential discrepancies until a final classification was agreed upon. Quotes were translated into English and back-translated independently.

RESULTS

Sociodemographic characteristics

Respondents were ethnic Han Chinese, aged from 19 to 32 years (mean age 23.3 years). Three identified as heterosexual, seven were ambivalent about their sexuality but were sexually interested in men. The rest identified as homosexual or “gay” (or in local parlance “tongzhi” and “a person in the circle”). One was married, although the majority (n  =  25) said they would marry due to family and social pressures. Four claimed steady non-commercial intimate relationships with men, three with women. Twenty-one had secondary education; three primary education; six post-secondary education. All were born in mainland China in rural or semi-rural villages. They had long work histories in different manual and semi-skilled jobs within and across cities. They came to cities for multiple reasons: economic survival, sexual freedom, social mobility, or to start a new cosmopolitan life.

Four work types were identified, which are commonly reported in other studies:21 (1) “full-time independents” (or “street hustlers”) (n  =  9), who work on their own in parks, bars, and increasingly through the internet; (2) “full-time brothel workers” (n  =  8), who work and some might even live in a sex venue (eg, male brothel or massage parlour) under an agent in which they cannot freely go out without permission; (3) “freelancers” (or “part-time workers”) (n  =  11), including those who have “quit” the job but freelance occasionally when short of money or requested by old clients and (4) “houseboys” (n  =  2), ie, those kept by a “sugar daddy”.

They were highly mobile and temporarily engaged in sex work. They worked at more than one work type (n  =  17), in and out of sex work (n  =  13) and moved between cities (n  =  13). Their length of sex work ran from 6 months to 8 years, but most (66.6%, n  =  30) had worked less than 2 years (see table 1).

Reported condom use, AIDS knowledge, perceived risk and HIV testing

All informants had sexual experiences before entering the sex industry. None reported condom use with men or women in early sexual experiences. They had limited knowledge about AIDS in their home towns and regarded AIDS as distant from their lives. Mass media and government publicised commercial sex with women and drug use as two major transmission routes; respondents apparently believed this and felt “safe” if they avoided commercial sex or drugs.

“When I was young, I had many girlfriends… we didn’t use condoms… you thought only xiaojie (female sex workers) and drug addicts had it.” (B4)

“When I was young, I had sexual experiences with different men… we didn’t use condoms.” (B9)

Not until migrating to cities and entering the sex industry did they start to use condoms at work.

“When I first started (male sex work), I didn’t use condoms…. But then I learned something about AIDS… it’s dangerous, you have sex with different people, but you can’t help it… so I am using condoms now.” (S17)

They then treated condom use at work as the norm.

“I use condoms at work, it’s the norm, like a flowchart, I put it onto him.” (S21)

They had good general knowledge about HIV/AIDS, from their managers, other MSW, mass media and health clinics, and understood the major routes of HIV/AIDS transmission, degrees of risk of different sexual practices and effective means of protection, and perceived higher levels of risk than before.

However, more than half (n  =  17) had not tested for HIV because: (1) they believed they had taken sufficient precautions, making health check-ups unnecessary; (2) they had to pay more for medical care as migrants because they only had rural but not urban hukou, a unique household registration system based on locality and family background exercised since 1958; (3) they did not want to disclose their sexual orientation to others, including medical professionals. No HIV/AIDS cases were reported; six respondents reported STI such as crabs, herpes, syphilis, gonorrhoea and urethritis.

Risk factors affecting condom use

Despite changes in perceived risk and behaviours after entering the industry, four factors kept them at risk.

AIDS knowledge and cultural myths

Some respondents reported misconceptions, stereotyping foreigners as “promiscuous”, believing locals “safer” and considering young MSM and those who looked “clean” or “innocent” as “safe”:

“If you are Chinese, young, and don’t have sex with foreigners, you should be fine.” (S27)

“I can tell after I talk to him. Those who seem to know little about the circle or seem not to have had sex before, I think are safe.” (S16)

Their clients were primarily Chinese men from other Chinese provinces; secondly, other overseas Chinese or Asian men; and lastly Caucasians.

Mistaken preventive measures were reported:

“I used condoms, I even wore two. It’s safer!” (B8)

“I took very hot shower before and after sex, it reduced the chance of infections.” (B13)

Their misconceptions stem from cultural myths dividing “safe”/“risky” sexual practices according to race, nationality, age, appearance and “quality”.23

Financial reasons

The overwhelming reason for entering the sex industry, as well as a major reason for engaging in unsafe sex, was money.

“I don’t want it (not to use condoms), but if he gives more money, I think it’s OK.”(S24)

As rural migrants, economic hardship meant basic survival in the city, supporting parental families, siblings’ tuition fees and family debts incurred from misfortune including natural disasters and illnesses. Like other studies, some, in dire financial straits, accepted pro-offered financial incentives for unprotected sex.11 24

Sexual orientation

Respondents all reported strong discrimination against homosexuality in their home towns.25^–27 Although money was the motive for entering the industry, homosexual MSW and those ambivalent about their sexuality treated sex work as a way to test sexuality and come to terms with homosexual desires, denied under social and family pressures.

“For me, it’s not just about money, we homosexuals are quite repressed… at our age, we have our biological (sexual) need.” (B14)

Homosexual respondents were more likely to engage in unsafe sex at work.

“I always use condoms at work… but if I like him… sometimes I don’t use.” (S28)

As reported elsewhere, those identifying as gay seemed at greater risk because intimacy, love and sexual pleasure were emphasised.18^–20 28 This seems true in China where homosexuality remains taboo.

Trust

The norm of condom use was loosened when clients became regulars, thus more like “friends”, who were perceived as “safe”.

“I’ve known him for a long time, more than half a year. He is my regular. He didn’t like condoms and tested negative, I trusted him, we didn’t use it… he is like my friend.” (B7)

This is particularly significant for homosexual MSW, who did not work solely for money, treating some clients as boyfriends and happy futures. “Freelancers” and “houseboys” were particularly vulnerable as they blurred the work/recreation boundary, regarding regulars as “friends”, “boyfriends” or “sugar daddies”.

Condom use in non-commercial intimate relationships—wife, girlfriend and boyfriend—was rare. The only married respondent (identified as homosexual) said he never used condoms with his wife. Three with steady girlfriends reported similarly, perceiving non-commercial heterosexual intimate sex as “safe”.

“I never use condoms with my girlfriend... I think if there is love, it’s not necessary.” (B11)

Four had steady boyfriends. Although they realised anal intercourse without condoms was dangerous, they took the risk:

“We have been together for four years. Sometimes we used, sometimes we didn’t. He is my first love and so am I for him. We think we are safe.” (S14)

Information about their partners’ sexual behaviour is unknown, but it is doubtful that they had no sex with others or always practised safer sex. Unprotected sex of non-commercial intimate relationships carried significant meaning. Condoms symbolised detached, public and businesslike “work sex”, distinct from intimate, private, and passionate “personal sex”.19 Romantic love “justified” the risk. Like other international studies, MSW in China were more at risk of HIV infection in their private lives than in their work lives.18^–20 29

DISCUSSIONS AND CONCLUSION

This study indicates that MSW who serve men in China are at risk. They are neglected in existing studies and fall into the loophole of HIV/AIDS sentinel surveillance for two reasons. First, most MSW are rural-to-urban migrants, but migrant studies (except two),15 16 and sentinel surveillance of the migrant population assume male migrants are potential clients of female sex workers, failing to acknowledge that they themselves could become MSW. Second, MSM studies tend to focus on non-commercial MSM, portraying MSW as heterosexuals working solely for money,7 8 11 without considering that they may be rural gay men experimenting with sexuality and looking for intimacy through sex work. This study rebuffs some of the myths about MSW.

By engaging with local and international researches, this study reveals the complicated sexual practices of MSW in China. No respondents had sexual protection in early sexual experiences, conforming to studies showing a lack of HIV/AIDS information in rural and semi-rural China, where safer sex prevention is urgently needed.30 Although they began safer sex practices after entering the industry, their initial period in sex work seems the most dangerous for HIV/AIDS infection, as is suggested elsewhere.15

The low reported HIV testing rate reflects structural constraints in accessing medical services due to rural migrant status and psychological barriers due to sexual orientation, which may undermine health-seeking behaviour.31 Some respondents contracted STI, suggesting that sexual health messages specific to MSM sexual behaviour should be conveyed, stressing that STI can be transmitted despite condom use.

Similar to other research, MSW in this study were more knowledgeable about AIDS, had higher awareness of HIV risks and tended to use condoms at work more than heterosexual rural migrants15 and average MSM.9 10 However, they were still vulnerable to HIV infections due to: insufficient AIDS knowledge and preventive measures resulting from cultural myths;23 dire financial straits;11 24 pleasure-seeking at work due to sexual orientation18^–20 28 and ambivalent condom use with regular clients due to blurred work/friendship boundaries and trusting non-commercial intimate partners.18^–20 29 The last two factors, largely reported in international studies, have not been discussed in the Chinese context. Moreover, the MSW vulnerabilities are reflected in work types. Houseboys’ vulnerability is linked to perceptions of “sugar daddies” as “boyfriends”. Full-time independents’ vulnerability stems from serious financial difficulties and limited AIDS information sharing. Freelancers hold a tenuous sex work identity, blurring work and recreation. Full-time brothel workers are relatively less vulnerable but are subject to higher surveillance. As they frequently move between categories, risks among different types may become indistinct. This requires further investigation.

The HIV threat among the MSM population is evident. The potential risk to the general population cannot be underestimated as most, no matter what their sexual orientation, would marry due to external pressures. Their low risk perception and unprotected non-commercial heterosexual sex are alarming.2 3 7 9^–11

HIV/STI prevention should address the social, cultural and economic contexts of these men’s engagement in sex work, paying attention to vulnerabilities resulting from their interlocking identities as rural–urban migrants, sex workers and homosexuals. As migrants, their deprived status in cities restrains their willingness to seek services. They are most vulnerable in the first few months of their commercial sex involvement when condom use is not normally practised. As sex workers, they, especially “freelancers” and “houseboys”, do not develop a strong “worker” identity. Their high mobility and temporary engagement also make HIV prevention difficult. As homosexuals, most were looking for more than money. Love, intimacy and sexual pleasure hinder condom use.

There are limitations to this study. MSW are difficult to locate due to stigma, so non-probability convenient sampling was used for recruitment. We interviewed none who exclusively served women, or who were transsexual/transvestite, who may experience different health vulnerabilities. MSW in small cities or rural areas were excluded. The small sample size, with findings based on self-reported data, potentially implies memory bias and concealment of information. The interviews were conducted in 2004–5; considerable changes in the community might have occurred. The financial incentive, although small, might have led some respondents to expect more resources in the future, possibly affecting their responses.