Transactional sex, condomless anal sex, and HIV risk among men who have sex with men

Abstract

To understand the HIV epidemic among men who have sex with men who engage in transactional sex (MSM-TS) in Paris, France, we sought to examine the association between engagement in transactional sex and HIV risk behaviors among MSM in Paris, France. Users of a geosocial-networking application in Paris were provided an anonymous web-based survey (N = 580), which included questions about transactional sex and behavioral risk factors for HIV along with sexually transmitted infection (STI)/HIV status. Multivariate analyses showed that engagement in transactional sex was associated with condomless receptive and insertive anal intercourse (adjusted relative risk [aRR] = 1.34, 95% confidence interval [CI] = 1.04–1.72 and aRR = 1.41, 95% CI = 1.04–1.91, respectively). MSM-TS were more likely to have engaged in substance use before or during sex (aRR = 1.35, 95% CI = 1.13–1.62), to have participated in group sex (aRR = 1.37, CI = 1.13–1.62), and to have had an STI during the last year (aRR = 1.68, 95% CI = 1.16–2.45). Transactional sex was not associated with HIV status. MSM-TS in Paris engaged in higher HIV risk behaviors, however, did not have higher rates of HIV infection. Sexual health interventions should continue to target MSM-TS; however, future studies should characterize the social, cultural, and structural factors that interact with individual behaviors to elevate HIV risk for MSM-TS.

Full text

Original research article
Transactional sex, condomless anal sex,
and HIV risk among men who have sex
with men
Ofole Mgbako
1,2
, Su H Park
2
, Denton Callander
3
,
Dustin A Brinker
4
, Christopher Kuhner
4
, Adam W Carrico
5
,
H Jonathon Rendina
6
and Dustin T Duncan
2
Abstract
To understand the HIV epidemic among men who have sex with men who engage in transactional sex (MSM-TS) in Paris,
France, we sought to examine the association between engagement in transactional sex and HIV risk behaviors among
MSM in Paris, France. Users of a geosocial-networking application in Paris were provided an anonymous web-based
survey (N¼580), which included questions about transactional sex and behavioral risk factors for HIV along with
sexually transmitted infection (STI)/HIV status. Multivariate analyses showed that engagement in transactional sex was
associated with condomless receptive and insertive anal intercourse (adjusted relative risk [aRR] ¼1.34, 95% confidence
interval [CI] ¼1.04–1.72 and aRR ¼1.41, 95% CI ¼1.04–1.91, respectively). MSM-TS were more likely to have engaged
in substance use before or during sex (aRR ¼1.35, 95% CI ¼1.13–1.62), to have participated in group sex (aRR ¼1.37,
CI ¼1.13–1.62), and to have had an STI during the last year (aRR ¼1.68, 95% CI ¼1.16–2.45). Transactional sex was not
associated with HIV status. MSM-TS in Paris engaged in higher HIV risk behaviors, however, did not have higher rates of
HIV infection. Sexual health interventions should continue to target MSM-TS; however, future studies should charac-
terize the social, cultural, and structural factors that interact with individual behaviors to elevate HIV risk for MSM-TS.
Keywords
HIV, men who have sex with men, transactional sex, France
Date received: 3 September 2018; accepted: 13 December 2018
Introduction
The global HIV epidemic among men who have sex
with men (MSM) is an important public health concern
around the world given the disproportionate burden of
infection this group faces. The World Health
Organization reports that MSM make up 49% of
new HIV infections in Western and Central Europe
and North America.
1
Given the broad categorization
of MSM, a more nuanced approach is necessary to
assess the HIV risk of MSM subgroups.
Previous research has suggested that transactional
sex poses elevated risk for HIV among MSM.
Transactional sex is defined as the exchange of sex for
money or some other commodity, an expression that
recognizes diverse configurations of contemporary sex
work among men from routine employment, to casual
encounters, to other needs like food or shelter.
2
One
1
Department of Internal Medicine, School of Medicine, Columbia
University, New York, NY, USA
2
Department of Population Health, Spatial Epidemiology Lab, School of
Medicine, New York University, New York, NY, USA
3
The Kirby Institute, University of New South Wales Sydney,
Sydney, Australia
4
School of Medicine, New York University, New York, NY, USA
5
Department of Public Health Sciences, School of Medicine, University of
Miami, Miami, FL, USA
6
Department of Psychology, Hunter College, City University of New
York, New York, NY, USA
Corresponding author:
Ofole Mgbako, Columbia University Medical Center, 630 West 168th
Street, P&S Box 82, New York, NY 10032-3784, USA.
Email: fofie.mgbako@gmail.com
International Journal of STD & AIDS
2019, Vol. 30(8) 795–801
!The Author(s) 2019
Article reuse guidelines:
sagepub.com/journals-permissions
DOI: 10.1177/0956462418823411
journals.sagepub.com/home/std

meta-analysis of 33 studies in 17 countries found that
MSM who engage in transactional sex (MSM-TS) had
overall higher HIV prevalence than those who do not.
3
However, conflicting empirical evidence exists around
HIV among MSM-TS depending on local context. One
Nigerian study, for example, found a slightly lower rate
of HIV among MSM engaged in transactional sex than
those who were not while a study from Mumbai, India
found that men selling sex had higher rates of HIV.
4,5
By contrast, a large study of gay and bisexual male sex
workers in Australia found that their rates of HIV and
other sexually transmitted infections (STIs) were no
different than their nonsex-working peers.
6
These stud-
ies are illustrative examples of the diverse epidemiology
that exists among MSM-TS globally, underscoring the
need for local, contextualized approaches to research
and interventions among MSM-TS.
7
Among MSM in France, diagnoses of HIV have
remained persistently stable and even increased slightly
in recent years, contrasting with declines observed
among other populations.
8
These trends demand
renewed attention to subpopulations that might be at
increased risk for HIV infection. Although some his-
torical literature exists on MSM-TS in France, more
recent accounts are needed.
9
This need is especially
great given the disparate HIV and STI prevalence esti-
mates among MSM-TS in other parts of Europe.
8
To address this gap and explore the role of transaction-
al sex in sexual risk and HIV infection rates among
MSM in France, we undertook a behavioral survey
to assess risk behaviors, such as condom use in recep-
tive and insertive anal intercourse, substance use before
or after sex, participation in group sex, as well as STI
and HIV status among a group of Parisian MSM-TS.
We sought to understand whether transactional sex is
associated with sexual health-related risk practices and
whether transactional sex is associated with HIV and
other STIs in this group.
Methods
We employed broadcast advertisements on a popular
geosocial networking application (‘app’) for MSM, tar-
geting individuals in the Paris (France) metropolitan
area in October 2016. Users were shown an advertise-
ment with text encouraging them to complete an anon-
ymous web-based survey.
10
In English the ad read,
‘Looking to improve your health, and the health of
those in your community? Share your thoughts with
us on gay and bisexual men’s health and have a
chance to win e65! Click more to get started!’ The
advertisement was placed during three consecutive 24-
hour weekday periods and was presented to users upon
opening the app for the first time during the broadcast
period. Upon clicking, users were directed to a
webpage to complete the survey, which included 52
items in total. As an incentive for participation, the
advertisement described that users who completed the
survey would be entered in a raffle to win e65 (approx-
imately $70 US dollars at the time of survey
administration).
Users were given the option to take the survey either
in English or French. The vast majority (94.3%) took
the survey in French. The survey took an average of
11.4 min for users to complete. At the end of the
recruitment period, 5206 users had clicked on
the advertisement and reached the landing page of
the survey, 935 users provided informed consent and
began the survey, and 580 users provided informed
consent and completed the survey (62.0% completion
rate and 11.1% overall completion rate). All protocols
were approved by the New York University School of
Medicine Institutional Review Board prior to
data collection.
Variables
The survey collected information on participant demo-
graphics, including age, sexual orientation, country of
birth, employment status, and relationship status.
It also asked questions about sexual practices, includ-
ing engagement in group sex, drug use before or during
sex, condomless anal intercourse, engagement in trans-
actional sex, as well as HIV and STI status
(Appendix 1).
Statistical analysis
The study participants’ characteristics, sexual risk
behaviors (including condomless anal intercourse, sub-
stance use before or after sex, engagement in group
sex), and HIV/STI status according to engagement in
transactional sex were analyzed descriptively (i.e. fre-
quencies and percentages) and compared using Chi
square test. In the multivariable analyses, we used a
modified Poisson regression model to assess the associ-
ation between transactional sex and condomless anal
intercourse, substance use before or after sex, engage-
ment in group sex, and HIV/STI status, adjusting for
sociodemographic characteristics.
11
In addition to gen-
eral engagement in transactional sex, we conducted a
sensitivity analysis to assess the association between
recent engagement in transactional sex (within last
year) and HIV risk behaviors. The results are presented
as adjusted relative risks (aRRs) with 95% confidence
intervals (CIs). All statistical analyses were carried out
using Stata version 14.0 (Stata Corp., College Station,
TX, USA). All tests were two-sided, with a level of
significance set at 0.05.
796 International Journal of STD & AIDS 30(8)

Results
Table 1 presents the characteristics of participants by
engagement in transactional sex. Of the 580 partici-
pants, 14.0% (n ¼81) reported they had engaged in
transactional sex. 44.5% (n ¼36) of MSM-TS were
below 30 years of age compared to 30.7% (n ¼151)
of MSM who do not engage in transactional sex
(MSM-NTS). Additionally, 51.9% (n ¼42) of MSM-
TS were employed, while 70% (n ¼345) of MSM-NTS
were employed.
Among MSM-TS, 63.0% (n ¼51) engaged in con-
domless receptive anal intercourse, whereas 48.1%
(n ¼237) of MSM-NTS reported engaging in condom-
less receptive anal intercourse (p <0.01). 32.1%
(n ¼26) of MSM-TS reported that they had been diag-
nosed with an STI in the past year compared to 20.7%
(n ¼102) of MSM-NTS (p <0.05). Moreover, 66.7%
Table 1. Sample characteristics by engagement in transactional sex (N¼580).
Transactional sex
Total, N (%) Yes No p
a
Overall 580 (100) 81 (14.0) 493 (85.0)
Age (years) 0.045
18–24 84 (14.5) 20 (24.7) 64 (13.0)
25–29 103 (17.8) 16 (19.8) 87 (17.7)
30–39 180 (31.0) 17 (21.0) 162 (32.9)
40–49 139 (24.0) 20 (24.7) 110 (24.1)
50 54 (9.3) 7 (8.6) 47 (9.5)
Sexual orientation 0.362
Gay 487 (84.0) 70 (86.4) 417 (84.6)
Bisexual 69 (11.9) 7 (8.6) 61 (12.4)
Born in France 0.238
Yes 450 (77.6) 60 (74.1) 389 (78.9)
No 113 (19.5) 20 (24.7) 93 (18.9)
Employment status 0.003
Employed 388 (66.9) 42 (51.9) 345 (70.0)
Unemployed 84 (14.5) 19 (23.5) 65 (13.2)
Student 81 (14.0) 17 (21.0) 64 (13.0)
Current relationship status 0.394
Single 378 (65.2) 51 (63.0) 326 (66.1)
Relationship with a man 172 (29.7) 28 (34.6) 144 (29.2)
Condomless receptive anal intercourse 0.014
0 partners 340 (58.6) 39 (46.9) 300 (60.9)
1 partners 226 (39.0) 42 (51.9) 184 (37.3)
Condomless insertive anal intercourse 0.046
0 partners 371 (64.0) 43 (53.1) 327 (66.3)
1 partners 193 (33.3) 34 (42.0) 158 (32.1)
Any condomless anal intercourse 0.006
0 partners 269 (46.4) 26 (32.1) 242 (49.1)
1 partners 289 (49.8) 51 (63.0) 237 (48.1)
Substance use during sex 0.004
No 272 (46.9) 27 (33.3) 244 (49.5)
Yes 293 (50.5) 54 (66.7) 239 (48.5)
Participation in group sex <0.0001
No 198 (34.1) 14 (17.3) 184 (37.3)
Yes 378 (65.2) 67 (82.7) 309 (62.7)
HIV status 0.262
Negative 444 (76.6) 60 (74.1) 384 (77.9)
Positive 58 (10.0) 11 (13.6) 47 (9.5)
Any STI 0.022
No 451 (77.8) 55 (67.9) 391 (79.3)
Yes 129 (22.2) 26 (32.1) 102 (20.7)
STI: sexually transmitted infection.
a
Chi square analysis.
Mgbako et al. 797

(n ¼54) of MSM-TS reported substance use during sex
and 82.7% (n ¼67) reported participation in group sex,
both higher than their MSM-NTS counterparts
(p ¼0.004 and p <0.0001, respectively).
The results of the multivariate analyses are pre-
sented in Table 2. Engagement in transactional sex
was positively associated with condomless receptive
anal intercourse and condomless insertive anal inter-
course in these multivariate models (aRR ¼1.34, 95%
CI ¼1.04–1.72 and aRR ¼1.41, 95% CI ¼1.04–1.91,
respectively). Compared with MSM-NTS, MSM-TS
were more likely to have used a substance before or
during sex (aRR ¼1.35, 95% CI ¼1.13–1.62), to have
participated in group sex (aRR ¼1.37, CI ¼1.13–1.62),
and to have had an STI during the last year
(aRR ¼1.68, 95% CI ¼1.16–2.45), after adjusting for
age, sexual orientation, country of origin, and employ-
ment and relationship status. Engagement in transac-
tional sex was not associated with HIV-seropositive
status (aRR ¼1.53, 95% CI ¼0.84–2.78).
Furthermore, MSM-TS were more likely to report
more frequent episodes of condomless receptive and
insertive anal intercourse (aRR ¼2.17, 95%
CI ¼1.08–4.36 and aRR ¼2.77, 95% CI ¼1.40–5.51,
respectively), as seen in supplemental Table 1. In
terms of STIs (supplemental Table 2), MSM-TS were
more likely to report infection with Chlamydia
(aRR ¼2.53, CI ¼1.41–4.56) and herpes simplex virus
(HSV) (aRR ¼2.63, CI ¼1.10–6.28).
A sensitivity analysis (supplemental Table 3) showed
that MSM-TS within three months were more likely to
have engaged in condomless receptive anal intercourse
(aRR ¼1.34, 95% CI ¼1.03–1.76), to have had an STI
(aRR ¼2.07, 95% CI ¼1.32–3.23), and have engaged
in substance abuse (aRR ¼1.50, 95% CI ¼1.20–1.88).
MSM-TS but not in the past three months were more
likely to have engaged in both insertive (aRR ¼1.57,
95% CI ¼1.08–2.27) and receptive (aRR ¼1.37, 95%
CI ¼1.07–174) condomless anal intercourse, and to
have participated in group sex (aRR ¼1.48, 95%
CI ¼1.32–1.67).
Discussion
For MSM in Paris, transactional sex was associated
with higher STI rates and HIV risk behaviors. The
MSM-TS in this study were younger with higher rates
of unemployment than MSM-NTS. Interestingly, while
being involved in transactional sex was associated with
certain high-risk practices and translated into higher
rates of STIs, a similar relationship was not observed
with HIV infection.
These findings hint at the complex interplay between
transactional sex, risk, and HIV infection among MSM
in Paris. Our findings build upon previous studies from
other parts of the world that demonstrate an associa-
tion between transactional sex and HIV risk behaviors
among MSM. A recent U.S. Boston-based study
revealed that transactional sex among MSM was sig-
nificantly associated with sex under the influence of
drugs or alcohol.
12
In a study in the French Antilles
and French Guiana on MSM-TS, transactional sex was
associated with drug use and a greater number of sex
partners.
13
Furthermore, supplemental Table 1 shows
that transactional sex was associated with a higher
number of episodes of condomless anal intercourse. It
is unclear whether this represents multiple sexual
encounters with the same partner or with different
partners; however, the number of condomless expo-
sures among MSM-TS likely is a major driver of
higher STI diagnosis.
Our findings also build upon studies showing the
positive association between MSM-TS and STI diag-
nosis, fitting with the trend of a higher STI burden
among MSM in France.
3,14
In our study, Chlamydia
and HSV were specifically associated with transactional
sex: prevalence was 9.48 and 4.31%, respectively. To
put the Chlamydia prevalence in context, the landmark
IPERGAY trial reported an overall Chlamydia preva-
lence over 20%.
15
The low overall STI prevalence as
compared to the IPERGAY trial may be a result of
self-reported STI status and lack of testing in our
study. For the same reason, due to lack of testing it
is impossible to know whether Chlamydia and HSV
infection rates were actually more prevalent than
other STIs or simply reported at a higher rate among
MSM-TS. Thus, it is unclear whether transactional sex
itself serves as a risk factor for Chlamydia infection.
Assuming the rates of Chlamydia were actually higher,
it is possible there were higher rates of asymptomatic
carriage of Chlamydia in our sample of MSM-TS, who
Table 2. Multivariate association (aRRs)
a
between transactional
sex with condomless anal intercourse, HIV status, STI status,
substance use before/during sex, and participation in group sex.
Transactional sex
aRR (95% CI)
Condomless receptive anal intercourse 1.34 (1.04, 1.72)*
Condomless insertive anal intercourse 1.41 (1.04, 1.91)*
Condomless anal intercourse 1.36 (1.12, 1.65)**
Substance use before or during sex 1.35 (1.13, 1.62)**
Participation in group sex 1.37 (1.20, 1.56)**
HIV-positive 1.53 (0.84, 2.78)
STI status 1.68 (1.16, 2.45)**
aRR: adjusted risk ratio; CI: confidence interval; STI: sexually transmit-
ted infection.
a
Adjusted for age, sexual orientation, origin (born in France), employ-
ment, and relationship status.
*p <0.05; **p<0.01.
798 International Journal of STD & AIDS 30(8)

were notably younger with higher rates of condomless
anal intercourse. A recent study in France exploring
the prevalence of asymptomatic STIs at different ana-
tomic body sites among MSM showed a significant
prevalence of Chlamydia in all three sites assessed
(pharynx, rectum, and urine) and a particularly high
rate of rectal chlamydia in MSM reporting receptive
anal intercourse.
16
This supports our finding of greater
association of Chlamydia compared to other STIs.
Higher Chlamydia rates may also be an indication of
irregular utilization of STI testing and other healthcare
services. Asymptomatic HSV carriage may also explain
the high rate of HSV among MSM-TS as well; howev-
er, more research among MSM subpopulations in
France is necessary to understand the predictors of
HSV infection.
The lack of an association between transactional sex
and HIV-seropositive status may be due to low overall
prevalence of HIV in this sample (10%) or lower rates
of self-reported HIV among MSM-TS, both of which
may mask a higher HIV prevalence than MSM-NTS.
Furthermore, while we did not include the participants
who responded that they did not know their HIV status
in our analysis, the 12.4% of MSM in Paris who
responded in this way may include many who are
HIV-positive. One of the drivers of continued high
HIV incidence among MSM is the lack of knowledge
of one’s HIV status, and this likely holds true among
MSM-TS. There may also be protective behaviors
among MSM-TS in Paris such as higher use of pre-
exposure prophylaxis (PrEP). A recent study showed
a high level of awareness of PrEP among MSM-TS in
Paris and a higher likelihood to use PrEP than MSM-
NTS; however, data on actual rates of PrEP use among
MSM-TS are lacking.
17
Other protective behaviors
may include increased adherence to antiretroviral ther-
apy leading to greater community viral suppression, as
well as MSM-TS having sex with regular partners who
are known to be HIV-negative or who are tested
more frequently.
Future research understanding the relationship
between transactional sex and HIV risk should assess
the varied motivations of MSM-TS and how different
expressions of paid sex may relate to different practi-
ces, sexual subcultures, and infection rates. For exam-
ple, male sex workers face punitive laws and
criminalization throughout the world in ways MSM-
TS who do not identify as sex workers may not under-
stand. Thus, there is likely further nuance around risk
behaviors and HIV incidence within this group of
MSM-TS.
Furthermore, the cultural and legal context of any
assessment of transactional sex and HIV risk is also
important. A recent global meta-analysis revealed an
association between a lower average HIV prevalence
among MSM-TS in countries that actively enact laws
to protect the rights of sexual minorities, such as
Canada, Argentina, and Ecuador.
18
In France, while
same-sex relations have been legal since the 1980s,
transactional sex has always been criminalized.
19
In
this context, it will be important to understand the
local sociocultural factors in Paris that influence rates
of HIV infection among MSM-TS, as criminalization
may deter some from seeking HIV prevention and care.
Future studies should explore the racial/ethnic back-
grounds and socioeconomic and employment status of
the MSM population in France in order to gauge how
these factors intersect with sexual risk among MSM-TS.
It is notable that 24.7% of our sample of MSM-TS were
born outside of France and 23.5% were unemployed.
Previous research has found that for sub-Saharan
African migrants living in Paris, transactional sex
between men and women has been associated with
HIV risk due to immigration hardships and to a lack
of stability in housing and in relationships.
20
Understanding the mechanisms for this risk among
MSM is essential.
Limitations
First, our sample was recruited from a single geosocial-
networking app raising the potential for selection bias.
Self-report bias is also a concern for self-reported
measures of sexual risk and HIV/STI status. The
survey did not confirm HIV/STI status by serological
testing. Also, our study sample was mostly gay-
identified, single, and born in France, further reducing
the generalizability of the results. As such, our results
provide limited insight into MSM-TS who have other
sexual identities such as straight or bisexual, are in
relationships, or are recent immigrants. Furthermore,
the motivations behind engaging in high risk behaviors
and partner-type characteristics of MSM-TS may have
further contextualized our findings.
Conclusion
Our results provide a meaningful contribution to the
literature surrounding MSM, transactional sex, and
HIV risk. Sexual health promotion interventions
should target MSM-TS but such interventions must
also account for the nuances of contemporary sex
work including its diverse manifestations. Qualitative
studies should explore factors that motivate sexual risk
behaviors and assess how these relate to transactional
sex, including attention to how sexual adventurism
could mediate men’s entry into the world of sell-
ing sex.
21
Mgbako et al. 799

Acknowledgements
We thank the translators and participants of this study who
contributed to the project. We thank Noah Kreski and Jace
Morganstein for assisting in the development, translation,
and management of the survey used in the current study.
Declaration of conflicting interests
The authors declared no potential conflicts of interest with
respect to the research, authorship, and/or publication of
this article.
Funding
The authors disclosed receipt of the following financial sup-
port for the research, authorship, and/or publication of this
article: This work was supported by Dustin Duncan’s New
York University School of Medicine Start-Up Research Fund
ORCID iD
Ofole Mgbako https://orcid.org/0000-0002-8955-4303
Dustin A Brinker https://orcid.org/0000-0002-5255-2531
H Jonathon Rendina https://orcid.org/0000-0002-
0148-2852
References
1. UNAIDS. Global AIDS response progress reporting
(GARPR), http://www.unaids.org/en/resources/docu
ments/2016/Global-AIDS-update-2016 (2016, accessed 1
August 2018).
2. McMillan K, Worth H and Rawstrone P. Usage of the
terms prostitution, sex work, transactional sex, and sur-
vival sex: their utility in HIV prevention research. Arch
Sex Behav 2018; 47: 1–11.
3. Oldenburg CE, Perez-Brumer AG, Reisner SL, et al.
Transactional sex and the HIV epidemic among men
who have sex with men (MSM): results from a systematic
review and meta-analysis. AIDS Behav 2015;
19: 2177–2183.
4. Bamgboye EA, Badru T and Bamgboye A. Transactional
sex between men and its implications on HIV and sexu-
ally transmitted infections in Nigeria. J Sex Transm Dis
2017; 2017: 1–7.
5. Narayanan P, Das A, Morineau G, et al. An exploration
of elevated HIV and STI risk among male sex workers
from India. BMC Public Health 2013; 13: 1059.
6. Callander D, Read P, Prestage G, et al. A cross-sectional
study of HIV and STIs among male sex workers attend-
ing Australian sexual health clinics. Sex Transm Infect
2016; 93: 299–302.
7. Minichiello V, Scott J and Callander D. A new public
health context to understand male sex work. BMC
Public Health 2015; 15: 282.
8. European Centre For Disease Prevention and Control.
HIV/AIDS surveillance in Europe, http://www.euro.
who.int/__data/assets/pdf_file/0019/324370/HIV-AIDS-
surveillance-Europe-2015.pdf (2015, accessed 16
June 2018).
9. da Silva LL. Travestis and Gigolos: male sex work and
HIV prevention in France. In: Aggleton P (ed.) Men who
sell sex: international perspectives on male prostitution and
HIV/AIDS. Philadelphia, PA: Temple University Press,
1999, pp.41–60.
10. Duncan DT, Park SH, Goedel WC, et al. Perceived
neighborhood safety is associated with poor sleep
health among gay, bisexual, and other men who have
sex with men in Paris, France. J Urban Health 2017;
94: 399–407.
11. Zou G. A modified poisson regression approach to pro-
spective studies with binary data. Am J Epidemiol 2004;
159: 702–706.
12. Perry NS, Taylor SW, Elsesser S, et al. The predominant
relationship between sexual environment characteristics
and HIV-serodiscordant condomless anal sex among
HIV-positive men who have sex with men (MSM).
AIDS Behav 2016; 20: 1228–1235.
13. Klingelschmidt J, Parriault MC, Van Melle A, et al.
Transactional sex among men who have sex with men
in the French Antilles and French Guiana: frequency
and associated factors. AIDS Care 2017; 29: 689–695.
14. Ndeikoundam N, Viriot D, Fournet N, et al. Les infec-
tions sexuellement transmissibles bacte
´riennes en France:
situation en 2015 et e
´volutions re
´centes. Bull Epide
´miol
Hebd 2016; 41-42: 738–744.
15. Molina JM, Capitant MD, Spire B, et al. On-demand
preexposure prophylaxis in men at high risk for HIV-1
infection. N Engl J Med 2015; 373: 2237–2246.
16. Philibert P, Hacene K, Penaranda G, et al. High preva-
lence of asymptomatic sexually transmitted infections
among men who have sex with men. J Clin Med 2014;
3: 1386–1391.
17. Mgbako O, Park SH, Mayer KH, et al. Transactional
sex and preferences for different PrEP modalities
among men who have sex with men (MSM). J Sex Res
2018; 56: 650–658.
18. Oldenburg CE, Perez-Brumer AG, Reisner SL, et al.
Human rights protections and HIV prevalence among
MSM who sell sex: cross-country comparisons from a
systematic review and meta-analysis. Glob Public Health
2018; 13: 414–425.
19. Legifrance. LOI no. 2016-444 du 13 avril 2016 visanta
`
renforcer la lutte contre le syste
`me prostitutionnel eta
`
accompagner les personnes prostitue
`es, https://www.legi-
france.gouv.fr/affichTexte.do?cidTexte.JORFTEXT
000032396046&categorieLien.id (2016, accessed 15
July 2018).
20. Desgrees-Du-Lou A, Pannetier J, Ravalihasy A, et al. Is
hardship during migration a determinant of HIV infec-
tion? Results from the ANRS PARCOURS study of sub-
Saharan African migrants in France. AIDS 2016;
30: 645–656.
21. Prestage G, Jin F, Bavinton B, et al. Sex workers and
their clients among Australian gay and bisexual men.
AIDS Behav 2014; 18: 1293–1301.
800 International Journal of STD & AIDS 30(8)

Appendix 1. Expanded description of
study variables
We assessed engagement in group sex events with the
question, ‘Have you ever had group sex (sex with three
or more people during a single sexual encounter)?’
Response options were as follows: ‘Yes, in the last
three months,’ ‘Yes, but not in the last three months,’
and ‘No.’ For analyses, we retained the original three-
level variable as well as dichotomized it to be ‘Yes’
(indicating any group sex) and ‘No’ (indicating no
group sex). We assessed drugs use by asking ‘How
many times in the past three months have you or
your partners used alcohol or drug before or during
sex?’ Response options were 0 times, 1–2 times, 3–5
times, 6–9 times, and 10 or more times. We dichoto-
mized it to be ‘0 times’ and ‘1 or more times.’
Participants were also asked about condomless anal
intercourse. Participants indicated the total number of
partners with whom they had had condomless insertive
anal intercourse and condomless receptive anal inter-
course in the preceding three months. These count var-
iables were transformed into categorical variables with
two categories (0 partners and 1 or more partners). The
study separately assessed condomless insertive anal
intercourse, condomless receptive anal intercourse,
and any condomless anal intercourse including
condomless insertive anal intercourse or condomless
receptive anal intercourse.
We assessed engagement in transactional sex with
the question, ‘Have you ever exchanged sex for
money, drugs, food, or shelter?’ Response options
were as follows: ‘Yes, in the last three months and I
used a smartphone application to do so,’ ‘Yes, in the
last three months and I did not use a smartphone appli-
cation to do so,’ ‘Yes, but not in the last three months;
I did use a smartphone application to do so,’ ‘Yes, but
not in the last three months; I did not use a smartphone
application to do so,’ and ‘No.’ For the purposes of
these analyses, a composite variable comprising all
‘Yes’ responses was created to indicate any transaction-
al sex versus none.
Participants were also asked to self-report their HIV
status (negative/positive/unknown); we chose to
exclude the 12.4% of men who did not know their
status. Participants were also asked to report diagnoses
of other STIs in the 12 months prior to participation,
specifically gonorrhea, Chlamydia, syphilis, herpes sim-
plex virus, human papillomavirus, and hepatitis C.
In addition to examining specific STIs, a composite
variable was created to indicate any recent STI diagno-
sis (yes/no).
Mgbako et al. 801