Intimate partner violence (IPV) is a significant public health concern that has been linked with higher levels of sexual risk behaviors, including lower levels of condom use, increased substance use, and increased participation in transactional sex.1–4 Studies have shown that individuals more likely to perpetrate IPV are also more likely to have multiple sex partners and less likely to use condoms, thus putting victims of IPV at increased risk for HIV and sexually transmitted infections (STIs).5 Exact definitions of IPV vary from study to study. However, studies on IPV commonly include measures of physical violence (e.g., pushing, hitting, shoving, kicking, throwing objects, and choking), sexual violence (e.g., rape, forced sexual contact, fondling, and unwanted sexual touching), and emotional violence (e.g., destroying property, verbal abuse, and threats to harm the individual or loved ones).6 Compared with the extensive research on heterosexual IPV, there is a paucity of studies on IPV among men who have sex with men (MSM), but existing studies from Western countries suggest that MSM have IPV rates similar to those of heterosexual women.7,8 Furthermore, a US Centers for Disease Control and Prevention report indicated that gay and bisexual men are much more likely to be victims of IPV than heterosexual men.9
In China, MSM are considered a key population for HIV acquisition,10 and existing studies indicate that they engage in high levels of risky sexual behavior, including multiple sex partners, condomless anal sex, sex while intoxicated, and intercourse with sex workers.11,12 Despite this, research on MSM and IPV specific to Chinese populations is extremely limited. One study conducted among MSM in Shanghai reported an IPV prevalence of 51%.13 However, this study population contained a large sample of “money boys” (e.g., men who sell sex to other men), who are likely at greater risk for IPV than other MSM due to their profession. To our knowledge, no studies have been conducted examining IPV among the general Chinese MSM population. Given the connection of IPV with increased sexual risk behaviors, greater priority should be placed on researching IPV among MSM in China.
Furthermore, studies from Western countries have indicated that men who have sex with both men and women (MSMW) experience higher rates of IPV victimization than do MSM.14 Some studies have also indicated that MSMW are more likely than MSM to engage in multiple high-risk behaviors, including transactional sex and injecting drug use.15–18 Although the reasons for this disparity are unclear, some researchers hypothesize that MSMW's increased rates of IPV victimization and risk behaviors may be a result of the “double stigma” they experience from both heterosexual and gay individuals.19,20 A study by Dodge et al.21 found that MSMW frequently had difficulty navigating between heterosexual and gay communities and often felt that they did not belong to any community. Although this study was conducted among a US population, MSMW in China may feel a greater sense of isolation due to their relative invisibility in society compared with MSM.22
Men who have sex with men and women are especially important in the context of China, where many gay men are socially compelled to marry women and start families, but still engage in sex with men.23 Existing studies indicate that a high proportion of MSM in China also have sex with women.24 One meta-analysis estimated that more than 30% of Chinese MSM have had sex with a woman in their lifetime.25 High levels of bisexual behavior among MSM have important implications for sexual health research in China, as it creates a situation where an individual's sexual self-identity is often incongruent with their sexual behaviors. Although some gay men who enter heterosexual marriages may not be engaging in sexual intercourse with their wives, many do because of high expectations to conceive children. Furthermore, men who self-identify as bisexual and are in heterosexual relationships may be engaging in extramarital sexual acts with men, or they may only be engaging in sex with women for fear of the stigma associated with same-sex sexual behaviors. Thus, sexual self-identification tells us little about actual sexual behaviors.26–29 This poses a challenge for researchers, who are faced with the decision of classifying participants based on their sexual self-identification or based on their reported sexual behaviors. In terms of HIV and STI risk, classifying participants based on sexual behaviors is generally more useful, as sexual behaviors are considered more salient factors for risk than sexual identification.30 In medical settings, it is recommended that clinicians avoid assumptions about client risk for HIV and STIs based on sexual identity. Instead, it is recommended that a thorough behavioral history is gathered to determine HIV/STI testing, treatment, and harm reduction guidelines for each individual.30
Given the high levels of bisexual behavior among Chinese MSM and increased vulnerability for IPV and risk behaviors among MSMW, it is therefore important not only to examine IPV among MSM in general but also to look at MSMW as a separate group from MSM. The aims of this study were to (1) examine the prevalence of IPV among MSM and MSMW and (2) identify correlations between IPV and risk behaviors and HIV/STI positivity.
Sampling and Recruitment
This study was part of a larger study focused on risk factors and HIV testing among MSM and included a diverse sample of MSM recruited from 31 provinces and other autonomous and administrative regions in China. Recruitment occurred through 3 different MSM specific Web sites: 1 based in Northern China, 1 in Southern China, and 1 in Eastern China. These Web sites are used for education, networking, partner seeking, and LGBT-specific news. Participants were recruited through a banner link on the Web pages and an announcement was sent to registered users. Those who clicked on the link were directed to eligibility screening and informed consent procedures. Participants must have been assigned male at birth, engaged in anal sex with another man at least once in their life, and be at least 16 years old (the age of consent in China) to participate in the study. Biologically born men who currently identify as a woman or transgender were eligible to participate. Data from transgender individuals have been described in a separate paper (in draft) and were therefore excluded from this analysis.
Consenting participants completed an online survey in Mandarin Chinese. Individuals did not receive an inducement for survey completion. All protocols for the study were approved by institutional review boards at the Guangdong Provincial Center for Skin Diseases & STI Control, the University of California–San Francisco, and the University of North Carolina at Chapel Hill.
Survey questions were based on existing IPV measures that have previously been used among MSM in China.13 To evaluate male-to-male IPV, participants were asked if their current male sexual partner had ever threatened to stop helping with money or housing, hit or thrown objects at them, threatened to hurt them or someone they loved, threatened to reveal their sexual orientation to others, or destroyed their property.13 Data were also gathered describing risk behaviors in the past year, such as whether participants had ever used recreational drugs, used drugs before having sex, used a condom during their last sexual encounter, had group sex, or had transactional sex. Participants were asked if they had ever tested positive for HIV or other STIs. Because all participants had to have a history of anal sex with a man, to determine MSMW status, respondents were asked “Have you ever had vaginal or anal sex with a woman?” Participants who responded “yes” to this question were coded as MSMW. Given the frequent incongruences of sexual self-identity labels and sexual behaviors and the high rates of bisexual behavior among gay-identified men in China, sexual behaviors were determined to be a better measure of risk than self-identified sexual orientation.25,31 Demographic variables such as age, province, education, income, and ethnicity were gathered.
A total of 1424 men participated in the larger study, but only men with current male sexual partners were asked questions about IPV. A total of 610 men reported having current male sexual partners and were included in this analysis. Missing data were low. Three participants had no response on the ethnicity question and were therefore automatically removed from all multivariate logistic regression analyses. Eighty-one participants had no response to the question on condom use and were excluded from regression analyses specific to that variable. Descriptive statistics were used to characterize the sample. Prevalence of IPV was calculated overall and by type of sexual behaviors (MSMW vs. strictly MSM). χ2 Tests were used to calculate significant differences in categorical variables between the 2 groups. Univariate and multivariate logistic regression analyses were conducted and odds ratios computed to explore the relationship between MSMW and IPV and the relationship between IPV and risk behaviors and reported HIV/STI positivity. A stratified analysis between IPV and risk behaviors by MSMW and MSM was also conducted. Common demographic variables (age, urban vs. rural location, region, education, income, and ethnicity) were adjusted for in the multivariate models. All analyses were conducted using SPSS version 22 (Durham, NC).
Of 610 participants, the median age was 24 years (interquartile range, 21–29 years), with a range from 16 to 55 years. Participants were from 31 provinces and autonomous and administrative regions in China, with the highest number of participants coming from Jiangsu (17.5%), Yunnan (10%), and Guangdong (7%) provinces and Beijing (6.7%). Table 1 provides information on the main demographic characteristics of the sample.
Most participants (67.7%; n = 413) reported only having sex with men, but 32.3% (n = 197) reported a history of sex with both men and women. The sample as a whole was highly educated, predominately Han Chinese, and lived in urban areas. Men who have sex with men and women were significantly older, had higher levels of monthly income, were more likely to self-identify as bisexual or straight, and were less likely to disclose their sexual orientation than MSM (Table 1).
The prevalence of IPV is shown in Table 2. Overall, nearly 30% of the sample reported experiencing any type of IPV from their current male sex partner. The most commonly reported form of IPV was the threat of disclosing sexual orientation (18.9%), followed by physical violence (16.1%). Fewer respondents reported receiving financial threats from intimate partners (6.7%). A significantly higher proportion of MSMW reported being a victim of IPV overall and for each category of violence, except for being hit or having objects thrown at them, than did MSM. Potential differences in IPV prevalence between gay-identified and bisexual-identified men were also examined (data not shown). However, no significant differences in IPV were found between these 2 groups.
After adjusting for demographic variables categorized in Table 1 (age, urban vs. rural location, region, education, income, and ethnicity), multivariate analysis (Table 3) revealed that MSMW were at significantly greater risk for IPV overall (adjusted odds ratio [AOR], 1.56; 95% confidence interval [CI], 1.04–2.32) and for having an intimate partner threaten to stop financial support (AOR, 2.42; 95% CI, 1.20–4.91) and threaten to reveal their sexual orientation to others (AOR, 1.70; 95% CI, 1.07–2.69). Threats to harm the individual or loved ones and destruction of property approached, but did not reach, significance after controlling for demographic variables. No significant differences between groups were found for having an intimate partner hit them or throw objects at them.
Demographic characteristics were compared between participants who had experienced IPV and those who had not. Participants who had experienced IPV were more likely to be older than the median age (P = 0.026), but no other significant demographic differences were found. We also examined the relationship between experiencing IPV and risk behaviors and HIV/STI positivity. As illustrated in Table 4, univariate analysis indicated differences were not significant for drug-related risk behaviors or reporting a history of STI diagnosis, but they were significantly different for risky sexual behaviors and a positive HIV diagnosis. After adjusting for age, urban versus rural location, region, bisexual behavior, education, income, and ethnicity, multivariate analysis showed that drug-related risk behaviors were still nonsignificant. Participation in risky sexual behaviors was still significantly higher for men who had experienced IPV than for men who had not; however, differences in condom use were no longer significant. Men who had experienced any type of IPV were more likely to have participated in group sex (AOR, 1.84; 95% CI, 1.07–3.18) and to have had sex in exchange for gifts or money in the last 12 months (AOR, 5.43; 95% CI, 2.63–11.23). Men who had experienced IPV were still more likely to report a positive HIV diagnosis (AOR, 2.79; 95% CI, 1.31–5.95), but differences in reporting a positive STI diagnosis remained nonsignificant.
We also conducted multivariate regression analysis by subgroup (Table 5). Experiencing IPV and having had sex in exchange for gifts or money in the last 12 months was found to be significant for both MSM (AOR, 6.02; 95% CI, 2.11–17.20) and MSMW (AOR, 7.54; 95% CI, 2.13–26.61). Relationships between IPV and participating in group sex in the last 12 months (AOR, 3.11; 95% CI, 1.45–6.68), not using a condom during last sexual encounter (AOR, 2.78; 95% CI, 1.45–5.34), and a positive HIV diagnosis (AOR, 5.90; 95% CI, 1.47–23.69) were found to be significant among MSMW.
Our study examined rates of IPV among MSM and MSMW across a broad geographic area in China. Studies on IPV have been primarily concentrated among heterosexual women, and few studies have focused on IPV among MSM in China. This study extends the existing literature by illustrating differences in IPV prevalence rates between MSM and MSMW and by demonstrating links between IPV and sexual risk behaviors and HIV positivity among MSM and MSMW in China. Particular strengths of this study include its large sample size, recruitment from a broad geographic area, and low levels of missing data. Unlike most other studies focused on MSM in China,13,22,32 our study recruited participants from almost every province and administrative region in China.
Our results indicate that IPV is common among MSM in China. Although overall rates were not as high as those found in a previous study in Shanghai,13 our study contained a more general population of MSM from a broad geographic region and did not recruit “money boys.” Although our sample is lower risk than money boys, almost a third of our sample reported being a victim of IPV. We found that MSMW were significantly more likely to experience IPV overall, withdrawal of financial support, and threats to disclose their sexual orientation. Although studies comparing IPV rates among MSM and MSMW are few, our results are consistent with US data showing that MSMW have higher rates of IPV than MSM.9 The reasons for these differences are unclear, but it may be a result of minority stress and the stigma and discrimination MSMW feel from both heterosexual and gay communities.19,20 A lack of social support has been correlated with IPV outcomes,33 and it may be that MSMW's isolation contributes to their increased prevalence of IPV. Further research needs to be conducted to determine the underlying causes for differences in IPV rates between MSM and MSMW.
Consistent with data from other studies,1,34 IPV had a strong association with risky sexual behaviors. Intimate partner violence was also associated with a positive HIV diagnosis. Other studies have indicated increased levels of IPV among HIV-positive individuals.2,35 Not only does IPV increase the risk of HIV infection, HIV-positive individuals are also more likely to be victims of IPV.2 Men who are HIV positive often have limited financial resources, which may increase dependency on their sexual partner.36 Furthermore, HIV-positive men often lack social support and face a substantial amount of stigma, making it difficult for them to find and maintain intimate relationships.37 Thus, HIV-positive individuals may be less likely to leave an abusive intimate relationship because they lack adequate financial resources and fear they may not be able to find another intimate relationship. Interestingly, unlike other studies, our study did not find an association between IPV and an STI diagnosis. This may be because our study relied on self-reported STI diagnoses, and rates of STI testing were much lower than rates of HIV testing. Had we been able to test all participants for STIs, we may have found a significant association with IPV.
Our study has several limitations. First, our findings are limited by the fact that this was a cross-sectional study, and thus, causal relationships between IPV, risk behaviors, and HIV could not be established. Second, because most of our sample participants tended to be younger and urban, their experiences may not be generalizable to older or rural MSM. Furthermore, data on the number of current sexual partners, the duration of partnerships, and types of partnerships were not gathered, so the role of these factors is unclear. Furthermore, questions about IPV were asked only in relation to participants' current male partner, not over the life course; therefore, estimates of IPV in this study are likely conservative.
Our study design used a lifetime measure of bisexual behavior rather than a 12-month period to reduce confounding by number of sexual partners.38 However, a lifetime measure of bisexual behavior could also result in misclassification bias, particularly among older participants who may have had sex with a woman in their youth, but have only had sex with men since then. There may be factors inherently different about men who did not immediately recognize their orientation, but came to that conclusion later in life that lead to increased vulnerability for IPV. However, there is no single best approach to studying bisexuality because each measure (self-identity, behaviors in the last 12 months, or behaviors over a lifetime) carries its own limitations. Men who have sex with men who had sex with a woman more than a year before the study would change classification if a 12-month cutoff were used. Given the social pressure for marriage in China, these men may be in a nonsexual relationship with a woman, but that status could change if reproduction becomes a goal. Thus, the decision was made to use the most inclusive definition (lifetime behavior). As a result, our MSMW population is somewhat larger than if we had limited the definition to behaviors within the last 12 months. However, if these men classified as MSMW really belong in the MSM group, then the bias would tend toward the null, and thus, any significant results identified in this study are robust estimates. As mentioned earlier, the decision to use behaviors rather than self-identity was driven by the social-contextual factors related to men's lives in China. Although we did examine IPV between gay-identified and bisexual-identified individuals, no significant differences were found. It is not clear why significant differences existed between behavior-based instead of identity-based categorizations, but it may be possible that in this cultural context, the labels of “gay” or “bisexual” may be less reflective of participants' experiences because self-identity was often incongruent with participants' reported sexual behaviors.
Our findings indicate that there is a hidden epidemic of IPV among MSM and MSMW. These findings have important implications for violence prevention and intervention programs. Currently, IPV programs in China primarily focus on women and there are no specific interventions targeting MSM and MSMW.32 However, based on the high levels of IPV found among these populations, the development of such programs is needed to help mitigate the impact of IPV on risk behaviors and HIV. Given the higher rates of IPV among MSMW, targeted interventions specific to MSMW may be necessary and may need to include content specific to MSMW, such as developing resilience against stigma from gay and heterosexual communities and establishing social support outside these communities. Special attention should also be given to HIV-positive MSM and MSMW, as they seem to have an increased vulnerability for IPV. Further research is needed to understand the pathways linking IPV and HIV risk among MSM to optimize the design of effective interventions. Additional research is also needed to elucidate the underlying causes in IPV prevalence differences between MSM and MSMW. The development of supportive clinical environments will be crucial to the establishment of programs aimed at reducing the high burden of IPV among MSM and MSMW in China.
1. Raj A, Santana M, La Marche A, et al. Perpetration of intimate partner violence associated with sexual risk behaviors among young adult men. Am J Public Health 2006; 96: 1873–1878.
2. Ramachandran S, Yonas M, Silvestre A, et al. Intimate partner violence among HIV-positive persons in an urban clinic. AIDS Care 2010; 22: 1536–1543.
3. Bauer H, Gibson P, Hernandex M, et al. Intimate partner violence and high-risk sexual behaviors among female patients with sexually transmitted diseases. Sex Transm Dis 2002; 29: 411–416.
4. Decker M, Miller E, McCauley H, et al. Recent partner violence and sexual and drug-related STI/HIV risk among adolescent and young adult women attending family planning clinics. Sex Transm Infect 2014; 90: 145–149.
5. El-Bassel N, Fontdevila J, Gilbert L, et al. HIV risks of men in methadone maintenance programs who abuse their intimate partners: A forgotten issue. J Subst Abuse 2001; 12: 29–43.
6. Duvvury N, Callan A, Carney P, et al. Intimate Partner Violence: Economic Costs and Implications for Growth and Development. World Bank, 2013””.
7. Waldner-Haugrud L, Gratch LMB. Victimization and perpetration rates of violence in gay and lesbian relationships: Gender issues explored. Violence Vict 1997; 12: 173–184.
8. Tjaden P, Thoennes N, Allison C. Comparing violence over the life span in samples of same-sex and opposite-sex cohabitants. Violence Vict 1999; 14: 413–425.
9. Centers for Disease Control and Prevention. The National Intimate Partner and Sexual Violence Survey: 2010 Findings on Victimization by Sexual Orientation. Atlanta, GA: Centers for Disease Control and Prevention, 2010.
10. Wong F, Huang Z, Wang W, et al. STIs and HIV among men having sex with men in China: A ticking time bomb? AIDS Educ Prev 2009; 21: 430–446.
11. Tao X, Gai R, Zhang X, et al. Prevalence of HIV infection and HIV-related sex risk behaviors in men who have sex with men in Shandong Province, China. Biosci Trends 2008; 2: 97–100.
12. Choi K, Hudes E, Steward W. Social discrimination, concurrent sexual partnerships, and HIV risk among men who have sex with men in Shanghai, China. AIDS Behav 2008; 12: S71–S77.
13. Dunkle K, Wong F, Nehl E, et al. Male-on-male intimate partner violence and sexual risk behaviors among money boys and other men who have sex with men in Shanghai, China. Sex Transm Dis 2013; 40: 362–365.
14. National Center for Injury Prevention and Control. National Intimate Partner and Sexual Violence Survey: 2010 Summary Report. Atlanta, GA: Centers for Disease Control and Prevention, 2010.
15. Munoz-Laboy M. MSM: Sexual desire among bisexually-active Latino men in New York City. Sexualities 2004; 7: 55–80.
16. Doll L, Beeker C. Male bisexual behavior and HIV risk in the United States: Synthesis of research and implications for interventions. AIDS Educ Prev 1996; 8: 205–225.
17. Doll L, Myers T, Kennedy M, et al. Bisexuality and HIV risk: Experiences in Canada and the United States. Annu Rev Sex Res 1997; 8: 102–147.
18. Sandfort T, Dodge B. Homosexual and bisexual labels and behaviors among men: The need for clear conceptualizations, accurate operationalizations, and appropriate methodological designs. In: Reddy V, Sandfort T, Rispel R, eds. Perspectives on Same-Sex Sexuality, Gender and HIV/AIDS in South Africa: From Social Silence to Social Science. Pretoria, South Africa: Human Sciences Research Council, 2009: 51–57.
19. Welzer-Lang D. “”Speaking out loud about bisexuality: Biphobia in the gay and lesbian community. J Bisexual 2008; 8: 81–95.
20. McLean K. Inside, outside, nowhere: Bisexual men and women in the gay and lesbian community. J Bisexual 2008; 8: 63–80.
21. Dodge B, Schnarrs P, Reece M, et al. Community involvement among behaviourally bisexual men in the midwestern USA: Experiences and perceptions across communities. Culture Health Sex 2012; 14: 1095–1110.
22. Cai R, Zhao J, Cai W, et al. HIV risk and prevention behaviors in men who have sex with men and women: A respondent-driven sampling study in Shenzhen, China. AIDS Behav 2014; 18: 1560–1568.
23. Wang Y, Li L, Zhang G, et al. Analysis on the intention of marriage and the influence factors among unmarried men who have sex with men. Zhonghua Liu Xing Bing Xue Za Zhi 2012; 33: 1031–1035.
24. Tao J, Ruan Y, Yin L, et al. Sex with women among men who have sex with men in China: Prevalence and sexual practices. AIDS Patient Care STDs 2013; 27: 524–528.
25. Yun K, Xu J, Reilly K, et al. Prevalence of bisexual behaviour among bridge population of men who have sex with men in China: A meta-analysis of observational studies. Sex Transm Infect 2011; 87: 563–570.
26. Ross M, Essien E, Williams M, 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.
27. Nield J, Magnusson B, Brooks C, et al. Sexual discordance and sexual partnering among heterosexual women. Arch Sex Behav 2015; 44: 885–894.
28. Pathela P, Hajat A, Schillinger J, et al. Discordance between sexual behavior and self-reported sexual identity: A population-based survey of New York City men. Ann Intern Med 2006; 145: 416–425.
29. Shingae A. Discordance between sexual behavior and self-reported sexual identity among MSM participating in the NLGR (Nagoya Lesbian and Gay Revolution) HIV testing event: Implications for developing future community-based prevention interventions. Eur J Med Res 2009; 14: 341–342.
30. Rankow E. Sexual identity vs sexual behavior. Am J Public Health 1996; 86: 1822.
31. Dodge B, Schnarrs P, Reece M, et al. Sexual behaviors and experiences among behaviorally bisexual men in the Midwestern United States 2013; 42: 247–256.
32. Chow E, Wilson D, Zhang L. What is the potential for bisexual men in China to act as a bridge of HIV transmission to the female population? Behavioural evidence from a systematic review and meta-analysis. BMC Infect Dis 2011; 11: 242–258.
33. Kamimura A, Parekh A, Olson L. Health indicators, social support, and intimate partner violence among women utilizing services at a community organization. Womens Health Issues 2013; 23: e179–e185.
34. Houston E, McKirnan D. Intimate partner abuse among gay and bisexual men: Risk correlates and health outcomes. J Urban Health 2007; 84: 681–690.
35. Greenwood G, Relf M, Huang B, et al. Battering victimization among a probability-based sample of men who have sex with men. Am J Public Health 2002; 92: 1964–1969.
36. Letellier P. Gay and bisexual male domestic violence victimization: Challenges to feminist theory and responses to violence. Violence Vict 1994; 9: 95–106.
37. Craft SM, Serovich JM. Family-of-origin factors and partner violence in the intimate relationships of gay men who are HIV positive. J Interpers Violence 2005; 20: 777–91.
38. Bauer G, Brennan D. The problem with ‘behavioral bisexuality’: Assessing sexual orientation in survey research. J Bisexual 2013; 13: 148–65.