The prevalence of HIV among men who have sex with men (MSM) varies considerably in different parts of Asia. In China, 12.2% of the 50,000 estimated new infections in 2007 could be attributed to MSM.1 There was only one national sentinel site on MSM in 2005, and just 14 sites in the entire nation in 2008.2 Accordingly, 2008 national survey surveying over 18,000 MSM in 61 Chinese cities reported an HIV prevalence of 4.9%.3 In recent years, HIV prevalence among MSM has increased rapidly in different parts of China. In Chengdu, it increased from 0.6% to 10.6% in 4 years from 2003.3 In Chongqing, rates went from 10.4% in 2006 to 12.5% in 2007.4
Most of the aforementioned studies were cross-sectional and did not report incidence data.5–9 However, reporting HIV incidence data are very important measures that can describe the rate that HIV is spreading in a target population, predict future trends of HIV prevalence, and quantify number of new infections incurred in an amount of time. Reduction of HIV incidence, rather than self-reported condom use behavior, is the strongest measure to evaluate whether HIV prevention services are effective in controlling the epidemic.
A few cohort studies have been conducted in Western countries. In Sydney, a study measuring HIV incidence for 1426 MSM at the end of a 4-year study period in 2006, reported an HIV incidence of 0.87 per 100 person-years.10 In a similar study in the United States and Canada, researchers followed-up on 3257 and 1587 HIV negative MSM, reporting the HIV incidence of 1.55 per 100 person-years and 2 per 100 person-years, respectively.11,12 In Amsterdam, HIV incidence for its cohort of MSM was 1.1 and 1.3 per 100 person-years in the years 1995 and 2002, respectively.13 In Spain, it increased from 1.91 to 3.28 per 100 person-years from 2002 to 2003.14 Despite these studies, there are still very few such HIV incidence studies targeting MSM in South East Asia and none have been reported in China.
Nanjing, the capital city of the Jiangsu province, has a population size of 7.59 million. In 2006, a survey in Nanjing reported that 4.73% MSM had HIV.15 While some voluntary peer groups working on HIV prevention target MSM jointly with the local CDC in Nanjing, there are no special clinical services offered.
This study reports on HIV incidence obtained from a cohort of 397 HIV negative MSM in Nanjing, China. Study participants were recruited using respondent-driven sampling (RDS) methods over a 6-month period. Factors predicting loss to follow-up were identified. Sex-related risk factors were examined, including unprotected anal intercourse (UAI) with sex partners (regular, casual, commercial, or any other), number of male sex partners in the last 6 months, casual sex partners in the last 6 months, and positive syphilis status at baseline. The mediating effects of some sex-related risk factors in explaining the relationships between particular risk factors and seroconversion were also investigated.
MATERIALS AND METHODS
Study Design and Study Population
The Baseline Study.
The baseline survey was conducted from May through July 2008. Study participants were recruited using the RDS method, which approaches random probability sampling under a set of assumptions.16 The method has been used for recruiting hard-to-reach populations, including injection drug users, female sex workers, and MSM.17,18 It permeates deeply into the participants' social networks to recruit a more representative sample through waves of referrals.16,19
RDS starts with an initial set of participants (called “seeds”), who receive an incentive for recruiting a preassigned number (usually 3–4) of peers to participate in the study. These recruited peers would further recruit the same number of peers in return. The process continues until the target sample size is achieved.16,19 In general, RDS can achieve equilibrium in the distributions of the respondents' characteristics after 5 to 6 waves of referrals, independent of the seeds' characteristics.16,19
Participants of the baseline study were at least 18 years old and had previously engaged in anal or oral sex with a man in the past 12 months. A research panel consisting of the researchers and stakeholders assisted the data collection process. A total of 9 seeds were recruited: 2 at gay saunas, 2 at gay bars, 2 via the internet, 1 via the park, and 2 via peer referral. Such venues have been used to recruit MSM study participants in previous studies.5–9 Considerations were given to diversity in age (5 were >30 years old). All but one of the 9 recruited seeds were able to recruit other respondents.
These 9 seeds each received 3 recruitment coupons for recruiting up to 3 MSM peers. Each coupon had an expiry date, a unique study number, information about the testing site, and a contact number for enquires. A research officer confirmed the eligibility of the participants and obtained their written informed consent. Pretest counseling was provided to the participants by a well-trained medical doctor in a private room before a blood sample was drawn and the baseline questionnaire was administered. Those who agreed to further recruit up to 3 of their peers were trained about the recruitment procedures and were given their recruitment coupons.
Participants were informed about the results of the rapid test and posttest counseling was provided by the medical doctor. Those yielding a positive rapid test result were asked to return in person in 2 weeks to receive confirmatory testing result. Those with confirmed HIV-positive status received additional posttest counseling and referrals to receive relevant free services.
Interview and volunteer counseling and testing (VCT) procedures were conducted at a conveniently located HIV clinic at the Jiangsu Provincial Central for Disease Control and Prevention. Privacy was ensured. The HIV antibody testing used rapid blood HIV antibody test (ELISA; Diagnostic Kit for Antibody to Human Immunodeficiency Virus, Acon Biotech Co., Ltd., China). Positive screening results were retested by another rapid test (Serodia-HIV; Fujirebio Inc., Japan), and Western Blot assay was used as the confirmatory test (Genelabs Diagnostics Pte Ltd., Singapore). Syphilis screening was performed by using the Rapid Plasma Regain Test (RPR; Beijing WanTai, Biologic Pharmacy Enterprise Co. Ltd). Positively screened results were confirmed by using the Treponema Pallidum Particle Agglutination Test (TPPA; Livzon Pharmaceutical Group Inc.). All cases with confirmed syphilis testing results received counseling and were referred to receive clinical treatment. Syphilis testing was also performed at month 6. A slightly different cohort is involved and the results are presented in a separate paper. The study was approved by the Ethics Committee of the Chinese Central for Disease Control and Prevention.
All participants who received HIV antibody testing, regardless of whether they succeeded in recruiting other participants, were given a box of free condoms and lubricant and an amount of 30 RMB (4 USD) was added to their mobile phone account. Based the 9 seeds, a total of 407 respondents were recruited from 13 waves of referrals. Equilibrium was reached at the 9th wave with regard to distributions on age, marital status, resident city, education level, monthly income, and venues most frequently visited to recruit male sex partners.
The 6-Month Follow-Up Study.
Of the 416 participants joining the baseline study, 19 were found HIV-positive (4.57%). The remaining 397 HIV negative participants were invited to uptake another VCT after 6 months, which was received by 286 MSM (286/397 = 72%). The median follow-up time was 5.73 months (interquartile range: 5.21–6.32 months). Out of the 111 MSM who did not return for the second VCT, 75 (67.6%) could not be approached (at least 8 contacts by phone plus other contacts attempted via the internet); 18.0% (n = 20) did not show up for at least 2 scheduled appointments with numerous reminders given to them; 7.2% (n = 8) refused to be retested, 5.4% (n = 6) were out of town, and 1.8% (n = 2) had bone fracture. The procedures of the follow-up study were similar to those of the baseline study. Those retested at month 6 were given some condoms/lubricant and 50 RMB (7 USD) was added to their mobile phone account.
The Baseline and Follow-Up Questionnaires
In-person interviews of participants were also conducted by trained peer field workers at both baseline and 6-month follow-up visits, before receiving VCT service. The variables of concern are listed in Table 1.
Binomial logistic regression models were fitted to identify risk factors predicting loss to follow-up. Adjusted relative risks (ARR) were reported, controlling for age, marital status, resident city, education level, and income. HIV incidence was estimated by using the number of seroconversion within the follow-up period as the numerator and the cohort's total number of person-year (PY) exposure to the risk of HIV transmission as the denominator. For those who with seroconversion, half of the follow-up duration (between the 2 visits) was used as their contribution to the total risk exposure.20 The 95% confidence interval of the incidence was estimated by using the Poisson method. Risk factors for HIV seroconversion were identified by estimating the relative risks (RR) and respective 95% confidence intervals (CI), using Poisson regression models.
Double data entry was performed and the 2 data sets were compared (Epi Data for Windows; The Epi Data Association Odense, Denmark). The Statistical Analysis System was used for data analysis (SAS 9.1 for Windows; SAS Institute Inc., NC). Statistical significance was defined by P value <0.05.
Factors Predicting Loss-to-Follow-Up
Retesters were more likely than defaulters to be local residents of Nanjing and were more likely to have attained junior college or higher education level. Adjusting for socio-demographic factors, those who self-reported being homosexual and having been a male sex worker in the past 6 months were more likely than others to a defaulter (Table 2).
HIV Incidence and the 95% Confidence Interval
During the follow-up period, seroconversion was detected among 7 participants, yielding an incidence rate of 5.12 per 100 person-years (7/136.81 PY). The 95% CI of the HIV incidence was from 1.33 to 8.91 per 100 PY.
Factors Predicting Seroconversion
Of all the background factors listed in Table 3, recruiting male sex partners mostly from saunas (RR = 2.35, 95% CI: 1.09–5.08, P = 0.0302) and identifying as an MSM for more than 10 years (>10 vs. ≤10, RR = 3.08, 95% CI: 1.53–6.20, P = 0.0016) are significant risk factors predicting seroconversion. Four other significant sex-related risk factors were identified. They were positive syphilis status at the baseline (RR = 2.82, 95% CI: 1.31–6.09, P = 0.0082), having had UAI with one of more male regular sex partners in the last 6 months (RR = 2.22, 95% CI: 1.10–4.49, P = 0.0261), having a causal sex partner in the last 6 months (RR = 2.17, 95% CI: 1.00–4.70, P = 0.0491), and having multiple male sex partners in the last 6 months (RR = 2.54, 95% CI: 1.18–5.49, P = 0.0174).
Recruiting male sex partners mostly from saunas was significantly associated with syphilis status at the baseline and became statistically nonsignificant in predicting seroconversion, after adjusting for the syphilis variable (RR = 1.97, P = 0.0921). Duration being a MSM >10 years was significantly associated with variables related to recruitment of sex partners from saunas (RR = 1.54, 95% CI: 1.24–1.91, P = 0.0001), positive syphilis diagnosis made at the baseline study (RR = 1.20, 95% CI: 1.03–1.40, P = 0.0214), having a causal sex partner (RR = 1.52, 95% CI: 1.01–2.28, P = 0.0425). The RR of the association between duration being MSM >10 years and HIV incidence decreased from 3.08 (P = 0.0016) to 2.29 (P = 0.038), before and after adjusted for variables related to recruiting male sex partners mostly from the sauna, syphilis status at the baseline and having a casual male sex partner.
This study is the first one of its kind in China, using a unique design combining developed RDS methods with a cohort design. The feasibility of using this model for conducting future incidence studies is demonstrated by its reasonable sample size and acceptable retention rate.
Our findings show that the HIV incidence in Nanjing, China is much higher than rates reported in some Western countries.11,12 The figure is consistent with the sharply rising HIV prevalence among MSM observed in different parts of China over the past few years by cross-sectional design.21 The high incidence suggests more HIV-positive cases will be reported and confirms that the threats are serious and real. Effective intervention services are most warranted.
In China, the social environment is currently not conducive to effective HIV interventions among MSM. Historically, Chinese culture has shied away from the issues surrounding homosexuality22,23 and stigmatization against MSM can be strong and adverse.8,22,23 Intervention programs targeting MSM are still in preliminary stages as HIV prevention work has only just begun over the last few years.24 Because the Chinese government has a tight grip over the development of civil organizations, including nongovernmental organizations serving MSM, such organizations are not prioritized and are hence underdeveloped.22 It is uncertain whether available VCT services would have a large coverage and would be effective.
The literature has previously shown that syphilis increases the risk of HIV infection,25 which is consistent with our results. The prevalence of syphilis diagnosis made at our baseline study was 12.7%; this STD ranges from 7% to 24.3% in other parts of China.26,27 The correlation of these 2 diagnoses show that the screening and treating STDs are critical in controlling HIV among MSM. As a policy point, the prevention of STDs and HIV should be integrated with each other. In many parts of China, many STD clinics/hospitals operate separately from the HIV prevention branch of the Centre for Disease Control and Prevention and private medical care is not well developed.22,28 MSM might find STD services not MSM-friendly and discriminating.28
Saunas should be given much more attention in HIV prevention programs targeting MSM. Other Chinese studies have suggested that sauna-based MSM are associated with higher HIV/STD prevalence and risk behaviors.27 Such is understandable as the sexual networks of the sauna customers might be overlapping to a large extent. Though condoms were distributed in some saunas in China, police actions might hinder a regular and stable supply. The public health implications of having gay-saunas as a high risk hub and a platform for HIV prevention in China needs to be debated openly. It is also seen from this study that the impact of sauna on HIV incidence can be partially explained by its association with syphilis status at the baseline.
It is very interesting to see that UAI with regular sex partners, but not UAI with nonregular sex partners, significantly predicted HIV seroconversion. One plausible explanation is that frequency of UAI may be higher among those with UAI with regular sex partner, hence increasing the risk of HIV transmission. UAI is prevalent among MSM's regular male sex partners (RPs), and trust are often mentioned as reason for not using condoms with RPs.29 Those having RP in China often have multiple sex partners and have unprotected sex with other types of MSM partners,5,6 while there seems to be a norm of condom nonuse with RPs. According to the Theory of Planned Behaviors,30 subjective norm affects the decision on condom use. Much has to be done to inform the MSM about the risk involving RPs and to shift the norm toward protected sex when having sex with RPs.
As expected, those who were nonlocals, less well educated, male sex workers, and were not homosexual in sexual orientation were more likely than others to be defaulters. Homosexual MSM may have a stronger sense of identify with the community and better relationship with the peer workers, and are hence more willing to show up for a retest. Male sex workers are much stigmatized and mobile31 and were hence difficult to be followed up. Researchers on future cohort MSM studies should pay attention to retention of the aforementioned types of study participants. These factors were not predictors of seroconversion.
The results suggest that programs should target those who with longer MSM experience, while age differences were not significant of this study. Some sex-related risk factors were associated with longer duration of being a MSM: recruiting male sex partners through sauna, positive syphilis diagnosis at the baseline and having a casual male sex partners in the last 6 months. These variables can be partial explanations for the relationship between duration of being MSM and HIV incidence. Much attention has previously been given to prevention targeting younger MSM,32 while other studies reported higher risk among elder MSM.33 Some cross-sectional studies suggested that newly initiated MSM were at high risk of contracting HIV.34 Our cohort study, however, showed the opposite might be true in China.
The prevalence of bisexual behaviors was high and was comparable to other studies targeting MSM in China.35 Though in our study, bisexuality was not a predictor of HIV incidence, the bridging effect to the female sex partners of MSM is a serious concern as there is no good intervention methods made available to protect female sex partners of MSM from contracting HIV.
The study has a number of limitations. The retention rate of this study was moderate (72%), which should be acceptable and being comparable to some published incidence studies.36 Loss to follow-up was not statistically associated with risk factors which predicted HIV seroconversion. Therefore, though some biases might exist, the size of such biases should not be too serious. Second, like all other relevant studies, the results of the condom use variables are self-reported. The study was however, anonymous; fake names and study identification numbers were used to match the records. Third, the sample size was only modest and the number of seroconversions was small; Poisson regression models were used. The sample size may be too small to provide an adequate power to detect some potentially important relationships. Lastly, there is no evidence to show that the HIV incidence data obtained from this study can be generalized to other parts of China. China is a big country and there are few studies on rural MSM.
This study was carefully designed, implemented, and monitored. It used the RDS design to yield representative estimations and equilibrium was reached after several waves. It sends a clear message about the urgency of the HIV epidemic in China, and points at some potentially useful considerations for designing future programs.
1. State Council AIDS Working Committee Office, China & UN Theme Group on HIV/AIDS in China. A Joint Assessment of HIV/AIDS Prevention, Treatment and Care in China. Beijing, China: United Nations Development Program, 2007.
2. Sun X, Wang N, Li D, et al. The development of HIV/AIDS surveillance in China. AIDS 2007; 21(suppl 8):S33–S38.
3. Wu ZY. China's assessment and responses to HIV epidemic in MSM. Presented at: WHO meeting on MSM; February 8, 2009; Geneva, Switzerland.
4. Feng L, Ding X, Lu R, et al. High HIV prevalence detected in 2006 and 2007 among men who have sex with men in China's largest municipality: An alarming epidemic in Chongqing, China. J Acquir Immun Defic Syndr 2009; 52:79–85.
5. Ruan S, Yang H, Zhu Y, et al. HIV prevalence and correlates of unprotected anal intercourse among men who have sex with men, Jinan, China. AIDS Behav 2008; 12:469–475.
6. 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 Immun Defic Syndr 2007; 45:77–84.
7. Choi KH, Hudes ES, Steward WT. Social discrimination, concurrent sexual partnerships, and HIV risk among men who have sex with men in Shanghai, China. AIDS Behav 2008; 12:s71–s77.
8. Ruan Y, Li D, Li X, et al. Relationship between syphilis and HIV infections among men who have sex with men in Beijing, China. Sex Transm Dis 2007; 34:592–597.
9. Jiang J, Cao N, Zhang J, et al. High prevalence of sexually transmitted diseases among men who have sex with men in Jiangsu province, China. Sex Transm Dis 2006; 33:118–123.
10. Jin F, Prestage GP, McDonald A, et al. Trend in HIV incidence in a cohort of homosexual men in Sydney: Data from the health in men study. Sex Health 2008; 5:109–112.
11. Buchbinder SP, Vittinghoff E, Heagerty PJ, et al. Sexual risk, nitrite inhalant use, and lack of circumcision associated with HIV seroconversion in men who have sex with men in the United States. J Acquir Immun Defic Syndr 2005; 39:82–89.
12. Lavoie E, Alary M, Remis RS, et al. Determinants of HIV seroconversion among men who have sex with men living in a low HIV incidence population in the era of highly active antiretroviral therapies. Sex Transm Dis 2008; 35:25–29.
13. van der Bij AK, Stolte IG, Coutinho RA, et al. Increase of sexually transmitted infections, but not HIV, among young homosexual men in Amsterdam: Are STIs still reliable markers for HIV transmission? Sex Transm Infect 2005; 81:34–37.
14. Hurtado I, Alastrue I, Ferreros I, et al. Trends in HIV testing, serial HIV prevalence and HIV incidence among people attending a Center for AIDS prevention from 1988 to 2003. Sex Transm Infect 2007; 83:23–28.
15. Guo H, Wei JF, Yang H, et al. Rapidly increasing prevalence of HIV and syphilis and HIV-1 subtype characterization among men who have sex with men in Jiangsu, China. Sex Transm Dis 2009; 36:120–125.
16. Heckathorn DD. Respondent- driven sampling II: Deriving valid population estimates from chain- referral samples of hidden population. Soc Probl 2002; 49:11–34.
17. Heckathorn DD, Semaan S, Broadhead RS, et al. Extensions of respondent-driven sampling: A new approach to the study of injection drug users aged 18–25. AIDS Behav 2002; 6: 55–67.
18. Yeka W, Maibani-Michie G, Prybylski D, et al. Application of respondent driven sampling to collect baseline data on FSWs and MSM for HIV risk reduction intervention in two urban centers in Papua New Guinea. J Urban Health 2006; 83:i60–i72.
19. Heckathorn DD. Respondent- driven sampling: A new approach to the study of hidden population. Soc Probl 1997; 44:174–199.
20. Heyward WL, Osmanov S, Saba J, et al. Preparation for phase III HIV vaccine efficacy trials: Methods for the determination of HIV incidence. AIDS 1994; 8:1285–1291.
21. Li SW, Zhang XY, Li XX, et al. Detection of recent HIV-1 infections among men who have sex with men in Beijing during 2005 – 2006. Chin Med J (Engl) 2008; 121:1105–1108.
22. Liu H, Yang H, Li X, et al. Men who have sex with men and human immunodeficiency virus/sexually transmitted disease control in China. Cell Res 2005; 15:858–864.
23. Neilands TB, Steward WT, Choi KH. Assessment of stigma towards homosexuality in China: A study of men who have sex with men. Arch Sex Behav 2008; 37:834–844.
24. Wu Z, Sullivan SG, Wang Y, et al. Evolution of China's response to HIV/AIDS. Lancet 2007; 369:679–690.
25. Reynolds SJ, Risbud AR, Shepherd ME, et al. High rates of syphilis among STI patients are contributing to the spread of HIV-1 in India. Sex Transm Infec 2006; 82:121–126.
26. Ma X, Zhang Q, He X, et al. Trends in prevalence of HIV, syphilis, hepatitis C, hepatitis B, and sexual risk behavior among men who have sex with men. Results of 3 consecutive respondent-driven sampling surveys in Beijing, 2004 through 2006. J Acquir Immun Defic Syndr 2007; 45:581–587.
27. Hong FC, Zhou H, Cai YM, et al. Prevalence of syphilis and HIV infections among men who have sex with men from different settings in Shenzhen, China: Implications for HIV/STD surveillance. Sex Transm Infect 2009; 85:42–44.
28. Choi KH, Diehl E, Guo Y, et al. High HIV risk but inadequate prevention services for men in China who have sex with men: An ethnographic study. AIDS Behav 2002; 6:255–266.
29. Elford J, Bolding G, Maguire M, et al. Sexual risk behavior among gay men in a relationship. AIDS 1999; 13:1407–1411.
30. van Kesteren NM, Hospers HJ, van Empelen P, et al. Sexual decision-making in HIV-positive men who have sex with men: How moral concerns and sexual motives guide intended condom use with steady and casual sex partners. Arch Sex Behav 2007;36:437–449.
31. Wong FY, Huang ZJ, He N, et al. HIV risks among gay- and non-gay-identified migrant money boys in Shanghai, China. AIDS Care 2008; 20:170–180.
32. Cong L, Ono-Kihara M, Xu G, et al. The characterisation of sexual behaviour in Chinese male university students who have sex with other men: A cross-sectional study. BMC Public Health 2008; 8:250.
33. Dukers NH, Fennema HS, van der Snoek EM, et al. HIV incidence and HIV testing behavior in men who have sex with men: Using three incidence sources, The Netherlands, 1984–2005. AIDS 2007; 21:491–499.
34. Centers for Disease Control and Prevention. HIV incidence among young men who have sex with men—Seven US cities, 1994–2000. MMWR Morb Mortal Wkly Rep 2001; 50:440–444.
35. Lau JT, Wang M, Wong HN, et al. Prevalence of bisexual behaviors among men who have sex with men (MSM) in China and associations between condom use in MSM and heterosexual behaviors. Sex Transm Dis 2008; 35:406–413.
36. van der Snoek EM, de Wit JB, Götz HM, et al. Incidence of sexually transmitted diseases and HIV infection in men who have sex with men related to knowledge, perceived susceptibility, and perceived severity of sexually transmitted diseases and HIV infection: Dutch MSM-cohort study. Sex Transm Dis 2006; 33:193–198.