Original research article

International Journal of STD & AIDS

2014, Vol. 25(6) 410–419
! The Author(s) 2013
Prevalence of syphilis infection and Reprints and permissions:
sagepub.co.uk/journalsPermissions.nav
associations with sexual risk behaviours DOI: 10.1177/0956462413512804
std.sagepub.com
among HIV-positive men who have sex
with men in Shanghai, China

Huan He1, Min Wang1, Nickolas Zaller2, Jun Wang1,

Dandan Song1, Yuhuang Qu3, Xin Sui3, Zhengxin Dong1,
Don Operario4 and Hongbo Zhang1

Abstract
The aims of this study were to understand the prevalence and correlates of syphilis infection among HIV-positive men
who have sex with men (MSM) in Shanghai, China. A total of 200 HIV-positive MSM participants were recruited using
‘‘snowball’’ sampling. Participants were tested for syphilis and completed a one-time questionnaire which included
demographic characteristics, sexual behaviours with male and female sexual partners, substance use, and use of anti-
retroviral medications. Prevalence of syphilis infection was 16.5%. Among HIV/syphilis co-infected participants, 63.6%
reported having anal sex with male partners and 24.2% did not use condoms consistently during the past six months;
66.7% reported having oral sex with male partners and 51.5% reported unprotected oral sex during the past six months.
Factors associated with testing seropositive for syphilis infection included receptive anal sex with a male partner in the
past six months (AOR ¼ 12.61, 90% CI ¼ 2.38–66.89), illicit drug use in the past six months (AOR ¼ 11.47, 90%
CI ¼ 2.47–53.45), and use of antiretroviral medication (AOR ¼ 4.48, 90% CI ¼ 1.43–14.05). These data indicate a
need for ‘‘positive prevention’’ interventions targeting HIV-positive MSM in China.

Keywords
HIV, AIDS, sexually transmitted infections, syphilis, Treponema pallidum, men who have sex with men, MSM, China, risk
behaviour

Date received: 8 May 2013; accepted: 28 August 2013

Introduction HIV and syphilis share a common mode of transmis-

HIV prevalence among men who have sex with men sion. Thus, the interaction between HIV and syphilis is
(MSM) in China has been rapidly increasing in recent of importance in both sentinel surveillance and clinical
years such that sexual transmission between men has treatment of both infections. Among MSM popula-
now become an important mode of HIV transmission tions, HIV infection has been shown to be strongly
in China.1–4 Concurrent with increases in HIV inci- associated with syphilis infection.7,8 A large body of
dence and prevalence, syphilis rates have also been on
the rise among MSM populations in China. A recent 1
Department of Epidemiology and Health statistics, School of Public
meta-analysis on co-infection of HIV and syphilis
Health, Anhui Medical University, Hefei, PR China
among MSM in China found that the syphilis preva- 2
Department of Infectious Diseases, Alpert Medical School, Brown
lence among Chinese MSM nearly doubled from 6.8% University, Providence, RI, USA
3
in 2003–2004 to 13.5% in 2007–2008; the HIV/syphilis Beautiful Life Health Promotion Center, Shanghai, PR China
4
co-infection prevalence also nearly doubled from 1.4% Program in Public Health, Brown University, Providence, USA
to 2.7% between 2005–2006 and 2007–2008.5 Another
Corresponding author:
meta-analysis of surveys conducted in China between Hongbo Zhang, School of Public Health, Anhui Medical University,
2001 and 2008 estimated an overall 9.1% prevalence of 69 Meishan Road, Hefei 230032, Anhui Province, RI, PR China.
syphilis among MSM in China.6 Email: [email protected]

Downloaded from std.sagepub.com at UNIV OF WESTERN ONTARIO on June 11, 2015

He et al. 411

research conducted internationally among MSM sug- Methods

gests that the prevalence of HIV and syphilis
co-infection may be as high as 20% in some MSM
Participants
populations.9–11 In U.S. settings, the prevalence of A cross-sectional survey was administered to HIV-
HIV infection has been shown to be two-to-three positive MSM between June and December of 2010 in
times higher among MSM who are co-infected with Shanghai, China. Inclusion criteria for the study were:
syphilis compared with MSM not infected with syph- (1) HIV-positive diagnosis confirmed by the local
ilis.12 In addition, syphilis outbreaks have occurred in Shanghai CDC prior to the study; (2) infected with
large cities in Italy where syphilis infection was more HIV via sex with men; (3) 18 years and older; and (4)
prevalent in HIV-infected MSM.13,14 A study con- reside in Shanghai during the study period. To facilitate
ducted in Amsterdam found a significant increase in our ability to reach the target populations, we collabo-
prevalent syphilis among MSM after the introduction rated with a non-governmental organization in
of combination antiretroviral therapy (cART), rising Shanghai (Beautiful Life Health Promotion Center)
from 0.5% to 0.8% of the MSM population.15 Many with a history of outreach to HIV-positive MSM.
studies posit that syphilis outbreaks among HIV posi- Participants were recruited using snowballing methods.
tive MSM may be related to changes in sexual beha- Initial participants were referred to the study by staff
viour among this group. For example, effective cART members at Beautiful Life Health Promotion Center.
may lead to reductions in safer sex practices among Beautiful Life Health Promotion Center services
MSM who no longer fear AIDS-related complications include providing HIV/AIDS education, counseling
such as opportunistic infections.15–19 for cART, and mental health referrals. After complet-
Shanghai has 17 million permanent residents and ing the survey, participants were asked to refer other
5 million migrants. By the end of 2007, 3010 HIV/ HIV-positive MSM in their peer networks to the study;
AIDS cases had been reported, of which 70% were participants were not compensated for providing refer-
migrants.20 Shanghai is one of the largest metropol- rals to the study. All referrals were screened via the
itan areas in China where the HIV/AIDS transmis- inclusion criteria, specified above, and invited to par-
sion route is largely attributed to homosexual ticipate in the study if they were eligible. Recruitment
contact. Some studies conducted in Shanghai continued until 200 participants were enrolled in the
found that substantial rates of unprotected sexual study.
intercourse among MSM, and highlights the import-
ant role that social and financial contexts play in
these behaviours. The rates of HIV-1 and syphilis
Procedures
infection were 4.4% and 10.3% of the MSM, Surveys were conducted in private rooms at our colla-
respectively.21 borating organization. Written informed consent was
Importantly, among many MSM, unprotected oral obtained from those who agreed to participate.
sex is considered a low-risk behaviour for HIV trans- Participants provide their written informed consent to
mission, but is a route of transmission for syphilis and participate in this study. The survey was administered
other sexually transmitted infections (STIs).7,22 One face-to-face by trained staff in a confidential space. All
recent study found that having more oral sex partners, participants received 30 Yuan (about $4) for participat-
receptive anal sex and unprotected receptive anal sex ing in the study. This process was documented by the
were all independently associated with greater odds of work staff of our research team. Study procedures and
being infected with syphilis.23 consent procedures were approved by the Anhui
Few known studies in China have described preva- Medical University IRB.
lence and correlates of syphilis infection among HIV-
positive MSM. Moreover, Shanghai had a great
Measures
number of HIV-infected MSM with some typical
social and behaviour characteristics, so we chose We assessed socio-demographic characteristics includ-
Shanghai as our survey area. The aims of this ing age, education, marital status, monthly income, and
study are: (1) to better understand the prevalence of sexual orientation identity. STI- and HIV/AIDS-
syphilis in a sample of HIV-positive MSM and (2) to related knowledge were measured using a validated
examine the relationship between syphilis infection 18-item measure. We made the 18-item measures
and demographic and behavioural factors. The find- based on scale of HIV/AIDS related knowledge of
ings of this paper will establish a basis for future China Ministry of Health. The eight questions of this
interventions to prevent further transmission of scale only focus on general population, but in our
STIs among HIV-positive MSM populations research we want to explore the level of knowledge
in China. about HIV/AIDS in HIV-positive MSM. The questions

Downloaded from std.sagepub.com at UNIV OF WESTERN ONTARIO on June 11, 2015

412 International Journal of STD & AIDS 25(6)

we made in this scale included many domains which Results

were closely related to HIV treatment and transmission
and prevention of opportunistic infections in HIV-posi-
Participant characteristics and syphilis infection
tive MSM; (sample item, ‘‘Does using a condom reduce A total of 200 HIV-infected MSM were enrolled in the
the risk of HIV/ STI transmission?’’; response options study. Table 1 depicts the demographic characteristics
were ‘‘Yes,’’ ‘‘No,’’ and ‘‘I don’t know’’). We assessed of the sample population. The mean age was 36.3
behavioural health information, including sexual beha- (SD ¼ 10.1) years and ages ranged from 21 to 68, with
viours with male partners (e.g. sexual position, the 45.0% of the sample being 26 to 35 years of age.
number of male anal sexual partners, HIV status of Approximately one-third (30.0%) of the respondents
male sexual partners, etc.) and female sexual partners reported being currently married to a woman.
(e.g. the number of female sexual partners, HIV status Participants currently living with a male partner and
of female sexual partners, the type of female sexual living with a female partner accounted for 17.0% and
partners, etc.) in the past six months, condom use 9.0% of the total sample population, respectively. The
(e.g. condom use in the past six months and during majority (54.5%) had a Shanghai residence card. Over
the most recent sexual episode), alcohol consumption, half of participants (55.0%) had at least a college edu-
alkyl nitrite use, and illicit drug use before having sex cation. Among all participants, 69.0% identified them-
during the previous six months. We also collected the selves as homosexual, 23.5% identified themselves as
physical health information of the participants (the bisexual, and 7.5% identified themselves as heterosexual
time of HIV diagnosis, receiving cART, reported or unsure. Of the 200 participants who provided blood
CD4 lymphocyte counts at last test). samples for this study, 33 (16.5%) tested positive for
syphilis. Among participants living with male partners
and female partners, approximately one-fifth (20.6%)
Laboratory testing and one-third (33.3%), respectively, tested positive for
Blood samples were collected from all eligible partici- syphilis. No statistically significant differences in syphilis
pants for syphilis testing. Syphilis seropositivity was seroprevalence were observed among participants with
determined by rapid plasma reagin (RPR; Shanghai respect to demographic characteristics.
Kehua Biotechnology Co., Ltd Shanghai, China) and
a Passive Particle Agglutination Test for detection of Relationship between syphilis infection and
antibodies to Treponema pallidum (TPPA;
behaviour characteristics
Singapore MP Biomedical Asia Pacific Ltd Singapore,
Singapore). Over half (57.5%) of participants in this sample reported
having had anal sex with male partners and 16% of par-
ticipants reported not using condoms consistently with
Statistical analysis male partners during the six months prior to the survey.
EpiData 3.0 (The EpiData Association, Odense, Sixty percent of participants in the sample reported
Denmark) software was used to input the survey data having had oral sex with male partners, and 47.0%
and Statistical Product and Service Solution (SPSS) 10.0 reported any unprotected oral sex in the previous six
was used to analyse the data. We performed descriptive months. Substance use measures included three cate-
analyses to describe the demographic characteristics of gories: any alcohol consumption, any alkyl nitrite use,
participants and to examine the relationship between and any other illicit drug use. Prevalences of alcohol con-
demographic characteristics and syphilis infection. sumption, alkyl nitrite use, and illicit drug use before sex
Chi-square tests were conducted to examine the relation- in the prior six months were 16.5%, 17.0%, and 2.5%,
ship between syphilis infection and behavioural charac- respectively.
teristics during the six months prior to the participant’s Table 2 shows the relationship between syphilis infec-
interview. We conducted multivariable regressions to tion and behaviour characteristics among participants
identify independent correlates of our primary depend- who were syphilis/HIV co-infected; 63.6% reported
ent variable: syphilis infection. In order to identify vari- having anal sex with male partners and 24.2% of partici-
ables for inclusion in the regression models, we used pants reported unprotected anal sex during the previous
bivariate analyses to identify correlates of both depend- six months. Approximately, two-thirds (66.7%) of co-
ent variables and included any co-factor that was asso- infected participants reported having oral sex with male
ciated with each dependent variable at P < 0.10. partners and 51.5% reported unprotected oral sex
Regression models also included controls for sociode- during the previous six months. The prevalences of alco-
mographic variables expected to be associated hol consumption, alkyl nitrite use, and illicit drug use
with both dependent variables (Hosmer & Lemeshow, before sex among participants with co-occurring syphilis
2000). infection were 15.2%, 27.2%, and 6.1%, respectively.

Downloaded from std.sagepub.com at UNIV OF WESTERN ONTARIO on June 11, 2015

He et al. 413

Table 1. Sociodemographic characteristics and syphilis infection of HIV-positive MSM sample (n ¼ 200).

Syphilis infection

Variables n (%) n (%) 2 P value

Age group (years) 1.927 0.588

18-25 19 (9.5) 2 (6.1)
26-35 90 (45.0) 17 (51.5)
36-45 45 (22.5) 5 (15.2)
46 46 (23.0) 9 (27.2)
Marital status 0.398 0.881
Single 120 (60.0) 19 (57.6)
Currently married 60 (10.0) 10 (30.3)
Divorced/widowed 20 (30.0) 4 (12.1)
Living with a male partner 0.497 0.481
Yes 34 (17.0) 7 (21.2)
No 166 (83.0) 26 (78.8)
Living with a female partner 2.836 0.092
Yes 18 (9.0) 6 (18.2)
No 182 (91.0) 27 (81.8)
Having a Shanghai residence card 0.577 0.448
Yes 109 (54.5) 16 (48.5)
No 91 (45.5) 17 (51.5)
Education level 1.204 0.548
Junior high school or less 31 (15.5) 7 (21.2)
High school or vocational school 59 (29.5) 8 (24.2)
College or higher 110 (55.0) 18 (54.6)
Sexual orientation 0.093 1.000
Homosexual 138 (69.0) 23 (69.7)
Bisexual 47 (23.5) 8 (24.2)
Heterosexual/undecided 15 (7.5) 2 (6.1)

Among participants who reported engaging in Logistic regression analysis of factors associated
receptive anal sex with male partners in the previous
with syphilis infection
six months, the prevalence of syphilis infection was
36.8%. Participants who reported having HIV-negative Multivariable logistic regression models examining cor-
male sexual partners were more likely to be syphilis relates of syphilis infection are reported in Table 3.
infected than those whose male sexual partners’ HIV Significant correlates of syphilis infection included
statuses were HIV-positive or unknown (44.4% receptive anal sex with male partners in the past six
vs.17.4% and 12.5%, 2 ¼ 9.88, P ¼ 0.01). In addition, months (AOR ¼ 12.61, 90% CI ¼ 2.38–66.89), illicit
respondents whose CD4 lymphocyte counts were drug use in the past six months (AOR ¼ 11.47, 90%
500 cells/mm3 or higher were more likely to be syphilis CI ¼ 2.47–53.44), and use of antiretroviral medication
seropositive (40.0% vs.14.9%, 2 ¼ 4.57, P ¼ 0.03). (AOR ¼ 4.48, 90% CI ¼ 1.43–14.05).
Individuals who reported that the Internet was
their most frequent place for finding male sexual part-
ners had a syphilis prevalence of 21.1%, whereas
Discussion
those who reported gay gathering sites (e.g. entertain- This study specifically examines HIV-infected MSM in
ment venues, bathhouses) as their most frequent China who are aware of their HIV infection. Our study
place for finding male sexual partners had a syphilis showed a higher prevalence of syphilis (16.5%) among
prevalence of 15.1%. The frequency of sex and the HIV-infected MSM compared with the general popula-
number of sexual partners were not related to syphilis tion of MSM in China (11.8%) and in Shanghai
status. (11.7%).24 Findings reported here are consistent with

Downloaded from std.sagepub.com at UNIV OF WESTERN ONTARIO on June 11, 2015

414 International Journal of STD & AIDS 25(6)

Table 2. Relationship between syphilis infection and sexual behaviours during the past six months among HIV-positive MSM par-
ticipants (n ¼ 200).

Syphilis infection

Variables n (%) n (%) 2 P value

The most frequent venue for seeking male sexual partners 0.79 0.38
Internet 57 (28.5) 12 (36.4)
Gay venuesa 73 (36.5) 11 (33.3)
No sexual behaviours with men 70 (35.0) 10 (30.3)
Sex with male partners 0.86 0.84
Anal sex only 10 (5.0) 1 (3.0)
Oral sex only 15 (7.5) 2 (6.1)
Both anal and oral sex 105 (52.5) 20 (60.6)
None 70 (35.0) 10 (30.3)
Anal sex positioning 5.73 0.02
Insertive only 26 (13.0) 2 (9.5)
Both insertive and receptive 70 (35.0) 12 (57.1)
Receptive only 19 (9.5) 7 (33.4)
Number of anal sex partners 2.29 0.32
1 42 (21.0) 5 (23.8)
2 28 (14.0) 5 (23.8)
3 45 (22.5) 11 (52.4)
Frequency of anal sex per week 3.02 0.17
<1 time 90 (45.0) 14 (66.7)
1  2 time 20 (10.0) 5 (23.8)
3 time 5 (2.5) 2 (9.5)
Had regular male sexual partner 0.73 0.39
Yes 61 (30.5) 8 (24.2)
No 139 (69.5) 25 (75.8)
Had casual male sexual partner 1.48 0.23
Yes 61 (30.5) 13 (39.4)
No 139 (69.5) 20 (60.6)
Had commercial male sexual partner 0.00 1.00
Yes 9 (4.5) 2 (6.1)
No 191 (95.5) 31 (93.9)
Inconsistent condom use during anal sex 2.00 0.16
Yes 32 (16.0) 8 (24.2)
No 168 (84.0) 25 (75.8)
Condom use in the last anal sex 0.00 1.00
Yes 107 (53.5) 19 (90.5)
No 8 (4.0) 2 (9.5)
Frequency of oral sex weekly 3.93 0.12
<1 time 98 (49.0) 15 (68.2)
1-2 times 17 (8.5) 5 (22.7)
3 times 5 (2.5) 2 (9.1)
Had vaginal sex with women 0.28 0.59
Yes 27 (13.5) 3 (9.1)
No 173 (86.5) 30 (90.9)
(continued)

Downloaded from std.sagepub.com at UNIV OF WESTERN ONTARIO on June 11, 2015

He et al. 415

Table 2. Continued.

Syphilis infection

Variables n (%) n (%) 2 P value

Inconsistent condom use during in vaginal sex 0.16 0.69

Yes 5 (2.5) 0 (0.0)
No 195 (97.5) 33 (100)
HIV status of sexual partners 9.88 0.01
HIV-positive 46 (23.0) 8 (24.2)
HIV-negative 18 (9.0) 8 (24.2)
Unknown 136 (68.0) 17 (51.6)
Alcohol use before sex 0.05 0.82
Yes 33 (16.5) 5 (15.2)
No 167 (83.5) 28 (84.8)
Alkyl nitrite abuse before sex 2.96 0.09
Yes 34 (17.0) 9 (27.3)
No 166 (83.0) 24 (72.7)
Illicit drug use before sex 0.68 0.41
Yes 5 (2.5) 2 (6.1)
No 195 (97.5) 31 (93.9)
Duration of HIV infection (years diagnosed) 0.32 0.85
<0.5 37 (18.5) 5 (15.2)
0.5-1 31 (15.5) 5 (15.2)
>1 132 (66.0) 23 (69.9)
CD4 count (cells/mm3) 4.57 0.03
<500 175 (87.5) 26 (81.2)
500 15 (12.5) 6 (18.8)
Receiving antiretroviral therapy 3.24 0.07
Yes 144 (72.0) 28 (84.9)
No 56 (28.0) 5 (15.1)
a
Gay venues defined as gay bar, entertainment venues, bathhouse.

Table 3. Multivariable logistic regressions: correlates of syphilis infection in HIV-positive MSM in China (n ¼ 200).

Syphilis infection

Univariate regression Multivariate regression

Variables OR (95% CI) OR (90% CI)

Age group (years)

18-25 Ref
26-35 1.979 (0.417–9.395)
36-45 1.062 (0.187–6.025)
46 2.068 (0.403–10.619)
Marital status
Single Ref
Currently married 1.063 (0.460–2.456)
Divorced/widowed 1.329 (0.400–4.413)
Living with a male partner
No Ref
Yes 1.396 (0.550–3.541)
(continued)

Downloaded from std.sagepub.com at UNIV OF WESTERN ONTARIO on June 11, 2015

416 International Journal of STD & AIDS 25(6)

Table 3. Continued.

Syphilis infection

Univariate regression Multivariate regression

Variables OR (95% CI) OR (90% CI)

Sexual orientation
Heterosexual/undecided Ref
Homosexual 1.300 (0.275–6.153)
Bisexual 1.333 (0.251–7.096)
Sex with male partners
Anal sex only Ref
Oral sex only 1.385 (0.108–17.670)
Both anal and oral sex 2.118 (0.254–17.689)
None 1.500 (0.171–13.160)
Anal sex positioning
Insertive only Ref Ref
Both insertive and receptive 2.483 (0.516–11.942) 3.781 (0.856–16.712)
Receptive only 7.000 (1.257–38.993) 12.608 (2.377–66.888)
Had regular male sexual partner
No Ref
Yes 0.688 (0.291–1.627)
Inconsistent condom use during anal sex
No Ref
Yes 0.524 (0.212–1.298)
Any illicit drug use in the past six months
No Ref Ref
Yes 4.469 (1.132–17.638) 11.486 (2.469–53.438)
Duration of HIV infection (years diagnosed)
<0.5 Ref
0.5-1 1.231 (0.321–4.716)
>1 1.350 (0.475–3.837)
CD4 count (cells/mm3)
<500 Ref
500 3.821 (1.254–11.637)
Receiving antiretroviral therapy
No Ref Ref
Yes 2.462 (0.899–6.739) 3.848 (1.492–9.925)

previous studies of syphilis prevalence in MSM. A condoms consistently with male partners. Previous stu-
survey of 477 MSM in Shanghai found a syphilis sero- dies in China had reported higher prevalence of unpro-
prevalence of 13.5%.25 Another study of 2087 HIV tected anal sex in MSM. One study conducted in
patients in Shandong found a seroprevalence of syphilis Chongqing included 1166 MSM, 14.8% of the partici-
of 19.6%.26 In our study, the high prevalence (16.5%) of pants had syphilis infection; furthermore, 35.1%
syphilis among MSM may indicate a high prevalence of reported unprotected anal sex in the past six months.27
unprotected sexual behaviours and suggests a potential Among some MSM, unprotected oral sex is con-
risk of rapid HIV spread among MSM, although syph- sidered a safer sexual practice thus making oral sex a
ilis is more infectious than HIV. During the past six risk factor for transmission and acquisition of syphilis.
months, 57.7% reported having anal sex with male part- Of note, 60.0% of participants in our study reported
ners and 16% of participants reported not using having oral sex with male partners with 47.0%

Downloaded from std.sagepub.com at UNIV OF WESTERN ONTARIO on June 11, 2015

He et al. 417

reporting unprotected oral sex in the previous six optimistic about their HIV disease being well-
months. Among the HIV/syphilis co-infected partici- controlled, such that they perceive themselves as
pants in our study, two-thirds reported having oral having low risk for additional STIs and subsequently
sex with male partners and half engaged in unprotected engage in unsafe sexual behaviours.33 Other inter-
oral sex. A review of 65 articles found that the propor- national epidemiological studies highlight that the
tion of syphilis transmission attributable to oral sex is increased survival of high-risk HIV-infected MSM
estimated to be between 20% and 46% in the USA and could lead to population level behaviour changes with-
Europe.28 Consequently, ‘‘positive prevention’’ efforts out necessarily changing the behaviour of specific indi-
for HIV-infected MSM in China should caution against viduals.26,3,34 However, although some MSM are
the STI risks associated with unprotected oral sex. reporting high rates of unprotected sex, many others
One interesting finding from our research was the are adopting HIV risk reduction strategies, such as
relationship between of anal sex positioning and syphilis ‘‘sero-sorting’’ which refers to choosing partners with
infection among HIV-positive MSM. We observed that the same HIV status. Although sero-sorting prevents
participants had a more than 12-fold greater adjusted HIV transmission from an HIV-positive to an HIV-
odds of testing positive for syphilis if they reported negative partner, the risk of syphilis and other STI
receptive anal sex versus insertive anal sex in the previ- transmission still remains.30,32 It is also important to
ous six months. It is possible that these men might per- note that one of the most significant consequences of
ceive low added risk for engaging in unprotected syphilis co-infection among HIV-positive MSM is its
receptive anal sex because they are already HIV infected. impact on the natural history of HIV infection and
Our study design was cross-sectional, which limited our the increased risk of transmission of HIV. Thus, there
ability to define the relationship between engaging in is a need for a continued emphasis on syphilis preven-
receptive anal sex and syphilis infection. Receptive tion, as well as early diagnosis and treatment of syphilis
anal intercourse with male sexual partner has been in HIV-infected MSM in China.
repeatedly reported as a risk factor for HIV infection In addition, consistent with other studies of risk
since early in the AIDS epidemic.29 Our study included behaviours related to syphilis infection among HIV-
more participants who engaged in receptive anal sex positive MSM,26,32 our multivariate analysis found
than men who engaged in insertive anal sex. However, that illicit drug use was independently associated with
efforts also are needed to educate HIV-positive MSM of syphilis infection among HIV-positive MSM. This may
the potential for syphilis infection and other STIs due to indicate that the use of illicit drugs is associated with
unprotected sex, and of the adverse health consequences unprotected sex, as some particular drugs are primarily
of co-infection of HIV and other STIs. used to enhance sexual pleasure and can also signifi-
The majority of HIV-positive MSM diagnosed with cantly impair judgment and/or reduce the ability to
syphilis in our study was already aware of their HIV- negotiate condom use through direct effects on mental
positive status and had been on cART for several function.35,36
years.30 Interestingly, our study found that receiving There are limitations to our study. The cross-sec-
cART was a risk factor for syphilis infection. On the tional nature of the study prevents ascertainment of
one hand, in our research, the time sequence of HIV causal associations between related behaviours and
and syphilis diagnosis was unclear. However, the inter- syphilis infection. Longitudinal studies are needed to
action between HIV and syphilis is complex. In add- determine more causative relationships. Our relatively
ition, syphilis infection may increase the immune small sample size may limit the ability to recognize stat-
activation of host cells and the secretion of cytokines, istically significant behavioural associations associated
and thus enhance HIV replication as well as decreases with HIV/syphilis co-infection. Due to the hard-to-reach
in CD4 cell counts.31 According to the national policy nature of the population of HIV-positive MSM, we used
of China, patients receive cART only if their CD4 convenience sampling based on a snowball recruitment,
counts fall to below 250. But we cannot define whether which limits the generalizability of our findings.
in our study the reason for the fall in CD4 cell counts Additionally, we only interviewed MSM in Shanghai
was syphilis infection. On the other hand, cART has and, therefore, the sample might not be generalizable
significantly reduced AIDS-related mortality and is to HIV-infected MSM in other parts of China. In add-
responsible for improved physical well-being, which ition, self-reported data are subject to response bias.
may allow higher rates of sexual activity. These data Participants were asked about sexual and drug use beha-
are consistent with a study in Tel Aviv, Israel which viours in the past six months, so recall bias may also be
indicated that HIV and syphilis co-infection was observed in our study. Finally, due to the scale of know-
found to be more common in HIV-positive patients ledge level in HIV-positive MSM which included three
receiving cART.32 Together, these findings suggest different domains, the validity and reliability of this scale
that some HIV-positive patients may feel overly might not be not high.

Downloaded from std.sagepub.com at UNIV OF WESTERN ONTARIO on June 11, 2015

418 International Journal of STD & AIDS 25(6)

Data presented in this paper show a great deal of References

overlap among HIV and syphilis among MSM in 1. Chow EP, Wilson DP, Zhang L, et al. Human immuno-
China. High prevalence of syphilis in this population deficiency virus prevalence is increasing among men who
is associated with ongoing sexual risk behaviour have sex with men in China: Findings from a review and
which might be due in part to the success of cART meta-analysis. Sex Transm Dis 2011; 38: 845–857.
which minimizes risk of opportunistic infections and 2. Ministry of Health, People’s Republic of China. 2005
reduces perceptions of illness severity and susceptibil- Update on the HIV/AIDS epidemic and response in
ity. The tendency for people to increase risk behaviours China. Beijing: National Center for AIDS/STD
when they perceive they are protected against harm Prevention and Control, China CDC, 2006.
(e.g. due to biomedical treatment) is referred to as 3. China Ministry of Health. UNAIDS, WHO, 2009 esti-
mates for the HIV/AIDS epidemic in China. Beijing:
risk compensation,37 and this phenomenon might be
http://www.docin.com/p-86177025.html (2010, accessed
contributing to increasing incidence of syphilis among
10 November 2013).
HIV-positive MSM in China. Additional research is 4. China Ministry of Health. UNAIDS, WHO, 2011
needed to examine understanding about sexual risk estimates for the HIV/AIDS epidemic in China. Chin J
behaviours among HIV-positive MSM in China and AIDS STDS 2012; 18: 1–5.
the perceived susceptibility and consequences of further 5. Chow EPF, Wilson DP and Zhang L. HIV and syphilis
STIs. ‘‘Positive prevention’’ interventions are needed in co-infection increasing among men who have sex with
China to continue educating HIV-positive individuals men in China: A systematic review and meta-analysis.
about behavioural and biological risks of STIs and stra- PLoS ONE 2011; 8: e22768.
tegies for reducing those risks, such as consistent 6. Gao L, Zhang L and Jin Q. Meta-analysis: Prevalence of
condom use and partner sero-sorting practices. In add- HIV infection and syphilis among MSM in China Sex
ition, integrating HIV testing with testing and treat- Transm Infect 2009; 85: 354–358.
ment for STIs – particularly in populations at high 7. Peterman TA and Furness BW. The resurgence of syph-
ilis among men who have sex with men. Curr Opin Infect
risk for co-infection such as MSM – may offer a
Dis 2007; 20: 54–59.
useful strategy for education and behavioural risk
8. Blocker ME, Levine W and St Lou ME. HIV prevalence
reduction to avert new STIs in vulnerable populations in patients with syphilis. United States. Sex Transm Dis
in China. In China, we should set a STI sentinel sur- 2000; 27: 53–59.
veillance system to evaluate the prevalence of HIV and 9. Hourihan M, Wheeler H, Houghton R, et al. Lessons
other STI co-infection among MSM, changes in the from the syphilis outbreak in homosexual men in east
HIV and STI epidemiology and increase in funding London. Sex Transm Infect 2004; 80: 509–511.
for prevention. This sentinel surveillance system also 10. Seña AC, Torrone EA, Leone PA, et al. Endemic early
can provide behavioural characteristics and related syphilis among young newly diagnosed HIV-positive men
information for analysis of trends in the HIV epidemic. in a southeastern U.S. State. AIDS Patient Care STDs
2008; 22: 955–963.
11. Ashton M, Sopwith W, Clark P, et al. An outbreak no
Author Contributions longer: factors contributing to the return of syphilis in
HZ was responsible for designing this research, MW, Greater Manchester. Sex Transm Infect 2003; 79: 291–293.
12. Zetola NM and Klausner JD. Syphilis and HIV infection:
YQ, and XS were responsible for conducting this
an update. Clin Infect Dis 2007; 44: 1222–1228.
research, ZD and DS were responsible for statistical ana- 13. Cusini M, Ghislanzoni M and Bernardi C. Syphilis out-
lysis, HH was responsible for writing this paper, and HZ, break in Milan. Sex Transmit Infect 2004; 80: 154.
NZ, and DO were responsible for revising this paper. 14. Giuliani M, Palamara G, Latini A, et al. Evidence of an
outbreak of syphilis among men who have sex with men
Acknowledgements in Rome. Arch Dermatol 2005; 141: 100–101.
We thank the staff of Beautiful Life Health Promotion Center 15. Stolte IG, Dukers NHTM, de Wit JBF, et al. Increase in
in Shanghai city for their help in recruiting participants, and sexually transmitted infections among homosexual men
we also the staff from the Center for Disease Control of in Amsterdam in relation to HAART. Sex Transm Infect
Shanghai for their assistance in syphilis testing, as well as 2001; 77: 184–186.
all participants in this study. 16. Battegay M, Fehr J, Flückiger U, et al. Antiretroviral
therapy of late presenters with advanced HIV disease.
Conflict of interest J Antimicrob Chemother 2008; 62: 41–44.
The authors declare no conflict of interest. 17. Ven van de P, Kippax S, Knox S, et al. HIV treatments
optimism and sexual behaviour among gay men in
Sydney and Melbourne. AIDS 1999; 13: 2289–2294.
Funding
18. Vanable PA, Ostrow DG, McKirnan DJ, et al. Impact of
This study was supported by Natural Science Foundation of combination therapies on HIV risk perceptions and
China (grant no. 30771850).

Downloaded from std.sagepub.com at UNIV OF WESTERN ONTARIO on June 11, 2015

He et al. 419

sexual risk among HIV-positive and HIV-negative gay 28. Peterman TA and Furness BW. The resurgence of syph-
and bisexual men. Health Psychol 2000; 19: 134–145. ilis among men who have sex with men. Curr Opin Infect
19. Elford J, Bolding G, Maguire M, et al. Combination Dis 2007; 20: 54–59.
therapies for HIV and sexual risk behavior among gay 29. Tabet SR, de Moya EA, Holmes KK, et al. Sexual behav-
men. J Acquir Immune Defic Syndr 2000; 23: 266–271. iors and risk factors for HIV infection among men who
20. He N, Zhang J, Yao J, et al. Knowledge, attitudes, and have sex with men in the Dominican Republic. AIDS
practices of voluntary HIV counseling and testing among 1996; 10: 201–206.
rural migrants in Shanghai, China. AIDS Educ Prev 2009; 30. Dougan S, Evans BG and Elford J. Sexually transmitted
21: 570–581. infections in western Europe among HIV-positive men
21. Choi KH, Hudes ES and Steward WT. Social discrimin- who have sex with men. Sex Transm Dis 2007; 34:
ation, concurrent sexual partnerships, and HIV risk 783–790.
among men who have sex with men in Shanghai, 31. Palacios R, Jiménez-Oñate F, Aguilar M, et al. Impact of
China. AIDS Behav 2008; 12: S71–S77. syphilis infection on HIV viral load and CD4 cell counts
22. Centers for Disease Control and Prevention. in HIV-infected patients. J Acquir Immune Defic Syndr
Transmission of primary and secondary syphilis by oral 2007; 44: 356–359.
sex: Chicago, Illinois, 1998–2002. MMWR Morb Mortal 32. Brosh-Nissimov T, Mor Z, Avramovich E, et al. Syphilis
Wkly Rep 2004; 53: 966–968. outbreak among men who have sex with men, Telaviv,
23. Mrie J, Lambert N, Mercer CH, et al. Refocusing health Israel, 2008–2009. Isr Med Assoc J 2012; 14: 152–156.
promotion for syphilis prevention: Results of a case-con- 33. Park WB, Jang HC, Kim SH, et al. Effect of highly active
trol study of men who have sex with men on England’s antiretroviral therapy on incidence of early syphilis in
south coast. Sex Transm Infect 2006; 82: 80–83. HIV-infected patients. Sex Transm Dis 2008; 35: 304–306.
24. Wu Z, Xu J, Liu E, et al. HIV and syphilis prevalence 34. Boily M-C, Bastos FI, Desai K, et al. Changes in the
among men who have sex with men: A cross-sectional transmission dynamics of the HIV epidemic after the
survey of 61 cities in China. Clin Infect Dis 2013; 57: wide-scale use of antiretroviral therapy could explain
298–309. increases in sexually transmitted infections: results from
25. Choi KH, Ning Z, Gregorich SE, et al. The influence of mathematical models. Sex Transm Dis 2004; 31: 100–113.
social and sexual networks in the spread of HIV and 35. Parsons JT, Kutnick AH, Halkitis PN, et al. Sexual risk
syphilis among men who have sex with men in behaviors and substance use among alcohol abusing
Shanghai, China. J Acquir Immune Defic Syndr 2007; HIV-positive men who have sex with men.
45: 77–84. J Psychoactive Drug 2005; 37: 27–36.
26. Zhao YS, Su SI, Lv CX, et al. Seroprevalence of hepatitis 36. Forrest DW, Metsch LR, LaLota M, et al. Crystal meth-
C, hepatitis B virus and syphilis in HIV-1 infected amphetamine use and sexual risk behaviors among HIV-
patients in Shandong, China. Int J STD AIDS 2012; 23: positive and HIV-negative men who have sex with men in
639–643. South Florida. J Urban Health 2010; 87: 480–485.
27. Zhou C, Ding X, Feng L-G, et al. Study on the preva- 37. Eaton LA and Kalichman S. Risk compensation in HIV
lence and associated factors of HIV and syphilis among prevention: Implication for vaccines, microbicides, and
1166 men who have sex with men. Modern Prevent Med other biomedical HIV prevention technologies. Curr
2011; 38: 815–820. HIV/AIDS Rep 2007; 4: 165–172.

Downloaded from std.sagepub.com at UNIV OF WESTERN ONTARIO on June 11, 2015