Prevalence of Paid Sex and Associated Factors Among Women and Men Attending HIV Voluntary Counseling and Testing in Kinshasa, Democratic Republic of the Congo: A Prospective Cohort

Paid sex is associated with HIV and other sexually transmitted infections, which are highly prevalent in Sub-Saharan Africa (SSA). However, few data exist on this sexual practice among the general population in SSA, including the Democratic Republic of the Congo, where data on paid sex mainly comes from sex workers. In the DRC, most HIV Voluntary Counseling and Testing (VCT) centers do not discuss paid sex as a risk factor. Thus, we aimed to analyze the prevalence of paid sex, its associated factors and association with HIV among women and men attending HIV VCT at a reference hospital in Kinshasa. From 2016 to 2018, the Observational Kinshasa AIDS Initiative cohort analyzed the impact of HIV VCT on changes in HIV knowledge, attitudes, and sexual behaviors at follow-up. Participants aged 15–69 years were HIV tested and interviewed at baseline and at 6- and 12-month follow-ups. At baseline, participants were asked about their history of “ever” having had exchanged sex for money. At both follow-ups, the frequency of this practice was referred to as “the previous 6 months.” Descriptive, bivariate, and multivariate logistic regression analyses were carried out to evaluate the prevalence of paid sex, its associated factors, and the association between paid sex and HIV. Statistical analyses were performed with Stata 15.1. Among 797 participants at baseline, 10% of those sexually experienced reported having ever had paid sex (18% men and 4% women, p < 0.001). At 6 and 12-month follow-ups, 5% and 2%, respectively. Paid sex was significantly and independently associated with being male (aOR = 2.7; 95% CI = 1.4–5.2), working or studying (aOR = 2.8; 95% CI = 1.5–5.0), daily newspaper reading (aOR = 4.4; 95% CI = 1.7–11.2); daily/weekly alcohol consumption (aOR = 3.3; 95% CI = 1.8–6.1), first sexual intercourse before age 15 years (aOR = 2.3; 95% CI = 1.1–5.0), multiple sexual partners (aOR = 4.1; 95% CI = 2.2–7.7), and extragenital sexual practices (aOR = 2.4; 95% CI = 1.3–4.4). A high religiosity (daily/weekly church attendance and praying) was inversely associated with paid sex (aOR = 0.1; 95% CI = 0.0–0.4). The high prevalence of paid sex among people attending HIV VCT in Kinshasa, associated with other sexual and consumption risk behaviors, highlights the need to include paid sex among the risk factors mentioned in HIV prevention counseling. Supplementary Information The online version contains supplementary material available at 10.1007/s10508-024-02939-w.


Introduction
Currently, the human immunodeficiency virus (HIV) and other sexually transmitted infections (STIs) are considered an important public health problem in Sub-Saharan Africa (SSA).The last updated data from UNAIDS estimates show that in 2021 60% of the new HIV infections worldwide take place in this region, where 25.6 million infected people live.In Eastern and Southern Africa a median HIV prevalence of 6.2% has been estimated among adults (aged 15-49 years), with a much higher prevalence among key populations, such as sex workers (33.4%); in Western and Central Africa the median prevalence is 1.3 and 8.9% among sex workers (UNAIDS, 2022a).Additionally, SSA has the highest prevalence of the four curable STIs (T. vaginalis, C. trachomatis, N. gonorrhoeae, and T. pallidum) (WHO, 2021;Zhang et al., 2022), as well as HPV (Bruni et al., 2021) or HSV-2 (James et al., 2020;Looker et al., 2020;Silva et al., 2022), among others.
Data on the prevalence of this sexual practice in SSA mainly come from female sex workers.Much less information exists on the prevalence of purchase or sale of sex among the general population, which can be an important bridging group for HIV/ STIs transmission to the rest of the community (Adal, 2019;Baltazar et al., 2021;Chatsika et al., 2020;Döring et al., 2022;Maher et al., 2020;Mantell et al., 2022;Ruangtragool et al., 2022;Seidu et al., 2019;Willis, 2013).A systematic review carried out in 2018 evaluated in nearly half of the SSA countries the prevalence of sex trade among youth (12-26 years) of the general population excluding people from high-risk populations.Among young males the prevalence of having ever bought sex ranged from 14 to 60% and it was 7 to 12% for selling sex.Among female youth information on buying sex was scarce, but a prevalence around 7% was reported and the lifetime prevalence of sold sex ranged from 5 to 85% (Krisch et al., 2019).There are less data on paid sex among adults from the general population.A recent meta-analysis including data from 35 African national population-based surveys, found that 8% of sexually active men reported having ever paid for sex, with a 68% prevalence of condom use at last paid sex in those surveys carried out since 2010 (Hodgins et al., 2022).
In the Democratic Republic of the Congo (DRC), it is estimated that around 0.8% of the adult population in the country are sex workers (UNAIDS, 2022a).The HIV prevalence among them is 7.5%, compared with 0.6% in the general population (UNAIDS, 2022).Regarding sex exchange with sex workers, a study focused on HPV infection among women in Kinshasa showed that 5 and 3% reported that their partners visited sex workers occasionally or frequently, respectively, before their current union (Sangwa-Lugoma et al., 2011).The 2013-14 National Demographic Health Survey (DHS) collected data among men aged 15-49 years and found that 28% of men reported having ever paid someone for sex and 11% in the previous year.In Kinshasa, the capital city, the prevalence was 23 and 8%, respectively.No data on women are available nor information on the associated factors (MPSMRM et al., 2014).
The DRC is one of the poorest countries in the world, characterized by persistent sociopolitical instability.Within this context, sex work has become a common way to earn money, Fig. 1 A theoretical framework for the determinants of paid sex in SSA Africa, based on the Knowledge-Attitudes-Practice (KAP) model mainly among women and young girls (Apasa et al., 2018).However, there is often under-reporting of transactional sex as a result of the associated stigma, and additionally, valid and reliable data on this practice is often not registered in clinical practice or counseling guidelines and protocols (Kyegombe et al., 2021;Wamoyi et al., 2019;Steen et al., 2015;USAID, 2015).Paid sex is not usually discussed in HIV Voluntary Counseling and Testing (VCT) centers within the National HIV/AIDS Program of the DRC (PNMLS, 2020).Thus, from a public health point of view, and considering the complex nature of factors related to paid sex, it is very important to detect early in the community the prevalence of engagement in paid sex, as well as its determinants, in order to provide timely information about its possible health risks and prevention.
Considering that there are scarce data on paid sex in Kinshasa and on the associated factors determining this sexual practice, we aimed to (1) analyze the prevalence of paid sex among women and men attending HIV VCT at a reference hospital within the public healthcare system in Kinshasa, DRC; (2) evaluate the associated behaviors, misconceptions, and structural and sociocultural factors; and (3) analyze the association between paid sex and HIV.

Participants
From April 2016 to April 2018, people aged 15-69 years attending HIV VCT at Monkole Hospital in Kinshasa were offered to participate in the Observational Kinshasa AIDS Initiative (OKAPI) prospective cohort study.As previously described, this study analyses the impact of HIV VCT on changes in HIV knowledge, attitudes, and sexual behaviors after a 6-and 12-month follow-up period (Carlos et al., 2021).At baseline, people with a previous HIV-positive test as well as pregnant women were excluded.
People attending VCT at Monkole Hospital are representative of the general population of the Mont-Ngafula area, where the hospital is located, and other surrounding areas (i.e., Ngaba, Selembao, or Lemba).It is a semi-urban area in the outskirts of Kinshasa, near some university campuses in Kinshasa.Specific high-risk groups are not known to be particularly present in these communities.VCT participants are not reached for HIV screening and there are no specific sensibilization campaigns for high-risk groups in the area.
Participants' transportation to Monkole Hospital was paid on arrival to facilitate follow-up but no incentives were given for participation.

HIV Tests and Other STI Diagnoses
A blood sample was collected from each participant through venipuncture.Rapid diagnostic tests were used for HIV diagnosis, consistent with the local practices: first, a Determine® HIV-1/2 test was done and if it was positive, a Double Check Gold® and Unigold® rapid immunoassays were carried out.A dried blood spot (DBS) sample was also collected for additional external HIV analyses (i.e., subtyping and resistance analyses) and STI diagnosis.Additionally, participants self-reported if they had ever received an STI diagnosis.

Personal Interviews
After the participants accepted to in the study and were HIV tested, face-to-face personal interviews were held.Considering the high frequency of illiteracy in Kinshasa, local male and female interviewers were available to collect participants' data, mainly in French and in Lingala in a few cases.They were local nurses highly sensitized to HIV and sexual health as they were working in VCT at Monkole Hospital during the study time.All baseline and 6-and 12-month follow-up interviews took place in a private room.

Questionnaire
Interviewers used a pen and paper questionnaire to collect data.Sociodemographic data as well as information about HIV knowledge, beliefs, attitudes, behaviors, and exposure to community HIV information were collected (Appendix).The questionnaires were built ad-hoc for the OKAPI project, based on previous projects at Monkole Hospital (Carlos et al., 2015(Carlos et al., , 2016) ) and on previously validated surveys, including the HIV-Knowledge-27-Scale, designed specifically for the sub-Saharan African population (Ciampa et al., 2012).The questionnaires included mainly closed questions (often on a Likert scale).They were initially designed in Spanish and translated by back-translation into Congolese French.The duration of the questionnaire implementation at baseline and follow-up surveys was about 35 and 20 min, respectively.

Measures
Following the theoretical framework of the study (Fig. 1), the prevalence of the outcome variable, paid sex, was first evaluated, as well as its associated risk behaviors.Secondly, misconceptions and their sociodemographic and other structural determinants were analyzed.

Paid Sex
Information about paid sex was collected for both male and female participants.At baseline, participants were asked whether they had "ever" had any kind of paid sex (this could include purchase or sale).In both 6-and 12-month follow-up questionnaires, this question referred to the previous 6-month period.In all questionnaires, the possible answers were: "never," "seldom," "frequently," and "I don't want to answer."

Alcohol Consumption and Sexual Risk Behaviors
Participants were asked about alcohol consumption in a normal week ("How often do you consume an alcoholic drink (beer, wine, whisky) in a normal week?") with possible answers going from "never" to "more than twice daily." Data on different sexual behaviors frequent in Kinshasa, as shown in previous results from the OKAPI cohort as well as other previous studies at Monkole Hospital (Carlos et al., 2017(Carlos et al., , 2019)), were collected: age at first sexual intercourse, multiple sexual partners, condom use and intention of use, sexual violence, and oral and anal sex.

HIV Misconceptions
Misconceptions about HIV and its transmission are highly prevalent in Kinshasa, as previously shown by this research group (Carlos et al., 2015); thus, this variable was collected in the OKAPI cohort questionnaire, including believing that someone HIV+ cannot look healthy, HIV is a punishment from God, and HIV is transmitted by sorcery, a kiss on the mouth or mosquito bites.

Sociodemographic Factors
Data on different sociodemographic characteristics were collected, including sex, age, education, economic level, professional status, religion and religiosity (church attendance and praying), and media access.

Sample Size Calculation
Considering a z = 1.96 (for a 95% CI), an estimated paid sex prevalence of 23% (DHS, 2014), and an error margin of 0.03, we estimated a sample size of 756 participants, following the formula n = z 2 *p*(1p)/error 2 .We estimated that this sample size would allow us to include enough parameters to analyze our objectives, considering that to perform multivariate analyses about 10 people are needed for each quantitative variable/indicator of qualitative variable in the model (Hosmer & Lemeshow, 2013).
Before performing the statistical analyses, data cleaning, errors correction and consistency checking were carried out.The presence of missing data was also evaluated, being its prevalence very low.
A descriptive analysis was first carried out to evaluate the baseline and 6-and 12-month follow-up prevalence of paid sex, as well as participants' sociodemographic characteristics, HIV-related knowledge and perceptions, and other behaviors.All these descriptive analyses were further stratified by sex.
For all analyses concerning paid sex, a new dichotomous variable was created, "having had paid sex" ("ever" or "in the previous 6 months" if it was paid sex reported at baseline or follow-ups, respectively).The categories 'seldom' and "frequently" in the original variable were collapsed and classified as "having had paid sex" ("yes") and those participants answering "never" were classified as "no" paid sex.
All categorical variables were described as percentages.The normal distribution of the quantitative variables was analyzed using Shapiro-Wilk test and the median and interquartile range (IQR) were calculated for those variables not following a normal distribution.The prevalence of paid sex and participants' characteristics were compared between women and men using χ 2 or Fisher exact tests for categorical variables and Student t tests for quantitative variables.
After the initial descriptive analyses, crude logistic regression analyses were carried out to evaluate, among participants sexually experienced, the factors associated with reporting paid sex.Afterward, multivariate regression models were adjusted in order to control for any confounding.Only the significant variables in the crude logistic regressions were kept to be included in the subsequent multivariate models.A first cross-sectional analysis was carried out to evaluate the association between baseline factors and reporting "ever having had paid sex" at baseline.The association between baseline and some follow-up variables and paid sex reported at 6-or 12-month follow-up was then analyzed.An additional multivariate analysis was carried out to evaluate the association between paid sex and having an incident HIVpositive test.
All analyses were carried out with STATA version 15.1 (StataCorp, College Station, TX, USA).All p-values < .05were considered statistically significant.

Pilot Study
A pilot study was carried out before the beginning of the OKAPI project to test the questionnaire comprehension, measure the interview time, and check all the study protocols were appropriate.Some minor changes were made after the pilot study.

Results
All people invited to participate agreed to join the study (100% response rate).At baseline, 797 participants replied to the baseline interview and were HIV tested.As previously described, at 6-and 12-month follow-up, retention rates were 57% (N = 456) and 27% (N = 219), respectively (Carlos et al., 2021).

Participants' Sociodemographic Characteristics
At baseline, 58% of the study population were women (Table 1).The median age of the participants was 28 (IQR: 12) years.Most participants had a middle economic level.Among those who reported their professional status, nearly 80% were studying or working.Regarding the education level, 56% had not completed their degree but nearly 70% of participants reported attending university.As expected, this percentage of incomplete university studies was significantly higher among young participants (84%) than among adults (42%).Around half of the participants reported daily access to the Internet and over 90% daily use of their mobile phones.Regarding their religion, 42% were Christians (Catholics and protestants) and 51% belonged to Église de réveil.Overall, the vast majority had a high religiosity.The majority of participants had a partner at study time and lived with their partner(s) but only 18% of the respondents reported being married (a higher proportion among men).

Prevalence of Paid Sex
Regarding paid sex, among 728 participants sexually experienced at baseline, 10% (n = 73) reported having ever had paid sex.The prevalence was significantly higher among men (18%) compared to women (4%) (p < .001).Only 6 out of the 728 participants (0.75%) refused to answer the question about transactional sex.
At 6-month follow-up, the prevalence of paid sex in the previous 6-month period was 5% (n = 16) (8% among men and 3% among women, p = .04);77% of them had also reported paid sex at baseline.At 12-month follow-up, only 2% (n = 3) reported paid sex in the previous 6 months (2 and 1% among men and women, respectively, p = .66).Among those participants reporting having ever had paid sex at baseline, 18% said they had it "frequently."For those reporting follow-up paid sex in the previous 6 months, 13 and 0% reported frequent paid sex at 6-and 12-month follow-ups, respectively.
Among respondents ever having paid sex, nobody reported consistent condom use and 10% said they would never use a condom if they happened to have sex with a sex worker in the future.

Prevalence of Other Sexual and Consumption Risk Behaviors
Both men and women reported a similar frequency of sex under 15 years (11%) and a very high frequency of oral sex (59%) (Table 1).Men were significantly more likely to report multiple serial and concurrent sexual partners and higher condom use.
As shown in Table 1, one out of four men reported daily or weekly alcohol drinking, which was two times significantly more frequent than among women's consumption (p < 0.001).

HIV-Related Misconceptions
Around 20% of the par ticipants had HIV-related misconceptions and wrongly believed HIV is caused by witchcraft or God's punishment.
When the association between baseline variables and paid sex at 6-month follow-up was analyzed, the same factors remained significantly associated in the crude analyses.However, as a result of the low participants' sample size reporting paid sex at 6-month follow-up (n = 16), only alcohol consumption remained significant in the adjusted model (adjusted OR = 3.5; 95%CI = 1.1-11.0,p = .034)(Table 2).
The association with paid sex at 12 months was not analyzed as only 3 participants were reporting paid sex at that follow-up.

HIV Diagnosis and Reported STIs
Three percent of the study population got a new positive HIV diagnosis at baseline and 10% of the participants reported having been diagnosed with an STI in the last year.Overall, only 3% perceived a high HIV risk (Table 1).

Association Between Paid Sex and HIV
When the association between paid sex and a positive/ undetermined HIV test at baseline was analyzed, a nonsignificant association was found for men (aOR = 1.9; 95%CI: 0.6-5.6,p = .286)and a strong and significant association was present for women (aOR = 10.7;95%CI: 2.1-53.8,p = .005)(Table 2).

Prevalence of Paid Sex
Between 2016 and 2018, a high prevalence of paid sex was reported among men and women from the general population attending HIV VCT in Kinshasa.Paid sex was associated with other consumption and sexual risk behaviors.
Having ever had paid sex was reported by 18% of men and 4% of women.The prevalence of paid sex was lower at follow-up (there was a higher attrition of people engaged in paid sex).
Apart from the official data from the National Demographic and Health Survey (DHS) (Hodgins et al., 2022), few studies have evaluated the prevalence of paid sex among people from the general population in SSA (Baltazar et al., 2021;Kloek et al., 2022;Oldenburg et al., 2014).For the DRC the only previous data in the country on this exchanged sex was collected at the DHS in 2013-14 (DHS, 2014).The survey showed that 23% of men aged 15-49 years from Kinshasa reported having "ever" had paid sex and 8% when it was referred to the previous year.This prevalence is slightly higher than our 18% prevalence for men.This could be a result of the different areas in Kinshasa participating in the DHS, far more numerous and diverse than those included in OKAPI.Considering the different ages of the participants (15-49 years in the DHS and 15-69 years in OKAPI) we estimate this did not have an impact, as only a few men over 49 years reported paid sex in our study.
Based on other DHS results from different SSA countries, the mean prevalence of paid sex among men was 10%; the DRC was, after Madagascar, the country with the second highest frequency of paid sex among men.No data are available in the DHS for women.A study carried out among women in South Africa found that 21% reported having ever had sex with a non-primary male partner in exchange for material goods or money (Dunkle et al., 2004).In Zimbabwe, Ruangtragool et al. (2022) analyzed the prevalence of paid sex among men and women from the general population and found that 6% of men and 3% of women reported transactional sex in the last 12 months.

Factors Associated with Paid Sex
Consistently with our results, other studies in SSA have shown a higher frequency of paid sex among men (Adjei, 2017;Krisch et al., 2019;Wamoyi et al., 2019).Among women not self-identified as sex workers, paid sex is usually related to material needs, such as food, clothing, transport, items for their children or families, or even somewhere to sleep (Dunkle et al., 2004;Lusey et al., 2014).For men, as shown in different parts of the world and not just in African countries, transactional sex is often linked to an erroneous interpretation of masculinity (Deogan et al., 2021;Huysamen et al., 2015;Shumka et al., 2017).In the African culture where men's dominance over women is more present than in other cultures, it is even easier to find this behavior (Conroy et al., 2016;Duby et al., 2023;UNAIDS, 2022b).Thus, working with men on the particular reasons for having paid sex is still necessary and VCT sessions can be an opportunity for this health educational approach.
Being working or studying and reading a newspaper daily were independently associated with reporting paid sex.Both variables are related to a higher economic status.As shown in other studies, access to paid sex is more frequent among those people who can afford this extraordinary payment (Chikutsa et al., 2015;DHS 2014;Dunkle et al., 2004;Jewkes et al., 2011;Krisch et al., 2019;Lusey et al., 2014;Mbonye et al., 2022;Seidu et al., 2019).On the other hand, among women, working or studying has an inverse association with paid sex which can be probably explained by the fact that having a stable status prevents women from having sex for money and material goods they need.
In the crude analyses, access to other mass media different from the newspaper or daily/weekly use of the Internet were associated with a higher frequency of paid sex, but significance was lost in the adjusted analyses.These media can be a way in which paid sex can be promoted.However, public health preventive strategies, can be a good option for promoting behavioral changes.Thus, considering that almost all participants reported daily access to their mobile phones, strategies based on phone use could help reduce the prevalence of this risk behavior.
Reporting daily/weekly praying or religious service attendance was independently associated with a lower frequency of paid sex.Participants who belonged to the Église de réveil reported less paid sex than participants from other religions.This could be a result of under-reporting as they could feel they could not report something stigmatizing.However, people belonging to this religious group usually have a lower economic level and this finding is consistent with the previously shown fact that a higher economic level is associated with higher access to paid sex.On the other hand, a high religiosity remained inversely associated with reporting paid sex, after adjusting for economic status.As described for other healthy behaviors, the effects of religious practice can be a result of their positive emotional or social impact rather than a consequence of normative aspects specific to certain religions (Ahrenfeldt et al., 2023).
In the present study, alcohol consumption was associated with reporting paid sex, both at baseline and follow-up, and this association was much stronger for women.This is consistent with other studies from SSA countries (Krisch et al., 2019;Tran et al., 2019;Tumwesigye et al., 2012).The DRC is one of the African countries with the lowest gross domestic product and with lower alcohol consumption (Ritchie & Roser, 2018;WHO, 2018).However, similar to the official estimates, in our study population 24 and 12% of men and women, respectively, reported daily or weekly alcohol consumption.It has been widely explained that alcohol consumption is associated with sexual risk behaviors.A recent publication has shown that it is not just the type of drink or the drinking amount that increases the odds of sexual activity but also where you drink.For example, parties or bars are considered environments with an increased risk (Hone et al., 2023).Alcohol prevention strategies to reduce the associated risks in the Congolese population should consider the entire context of alcohol consumption.
As described in a previous OKAPI analysis (Carlos et al., 2021), there was a very low retention rate at the 12-month follow-up.Among other factors, alcohol consumption was inversely associated with retention.As people reporting paid sex were more likely to report alcohol consumption, this could explain part of the higher attrition among people reporting paid sex.
Another risk behavior associated with paid sex, both in the cross-sectional and the longitudinal analyses, was reporting multiple sexual partners, as shown in other studies from Africa (Baltazar et al., 2021;Krisch et al., 2019;Ssempijja, 2022).This effect was stronger for women.Both sexual risks are connected.As previously reported in this cohort, different sexual risk behaviors are associated with each other (Carlos S, et al., 2021).In the present study other sexual behaviors, such as reporting extragenital practices (oral or anal sex) or inconsistent condom use were associated with paid sex.The overlapping of different risk behaviors among the population, which increases the risk of acquiring or transmitting HIV and other STI, should all be considered in the counseling sessions (Waters & Dewsnap, 2022).

Paid Sex and HIV
As expected and described in the literature (Baltazar et al., 2021), participants reporting paid sex were more likely to have an HIV-positive test.This fact should lead the Congolese government and HIV organizations to support preventive campaigns that promote the avoidance of paid sex.Raising awareness of its risks in healthcare settings is necessary to reduce the incidence of HIV and other STIs.

Study Limitations
The present study has some limitations.First, paid sex and other risk behaviors are socially undesirable, and thus, misclassification bias could be present.Although professional interviewers (male to male, female to female), individual rooms, and anonymity were available at baseline and both follow-ups, to reduce this bias, we acknowledge that it is difficult to mitigate.If this bias was present, it would be non-differential and results would have been biased toward the null.Furthermore, the counseling received at baseline about the negative effects of sexual risk behaviors could influence the biased responses on sexual behaviors but also, the other way round, on reducing these risk behaviors.Secondly, the question used about exposure to paid sex was quite general and did not collect specific information about what they considered paid sex (ie.people reporting paid sex may have not had risky sex but other kinds of paid sex more related to pornography), the type of partners involved, or even the reasons for having paid sex.In this sense, the analyses adequately answer the study objectives (to evaluate the overall prevalence and associated factors).However, additional aspects need to be considered.Adding qualitative analyses in the future can also help to complete this information.Third, it needs to be highlighted the potential for bias in the fact that people reporting paid sex were less likely to be retained.Due to the result of the low retention rate at 12-month follow-up, no longitudinal analyses could be carried out to evaluate the association between paid sex at 12-month follow-up and participants' characteristics.However, they could be done for the 6-month follow-up and the same risk behaviors as at baseline showed to be associated with paid sex.

Study Strengths
Despite the mentioned limitations, this study has several strengths.First, this is the first study evaluating the prevalence of paid sex among people from the general population attending HIV VCT in Kinshasa, and the first time that data on women were collected.Second, both cross-sectional and longitudinal analyses were carried out to evaluate the different associations and all of them clearly showed the strong association between paid sex and male sex and with other risk behaviors such as alcohol consumption, multiple sexual partnerships, and extragenital sexual practices.Finally, the study included nearly 800 participants at baseline which allowed multivariate logistic regressions to be carried out, taking into account the possible effect of many sociodemographic, knowledge, and behavioral variables.

Conclusion
In conclusion, paid sex is a prevalent sexual practice among Congolese men and women from the general population and it is associated with alcohol consumption and other sexual behaviors.It needs to be included among the risk factors to be mentioned in the HIV counseling sessions in the Democratic Republic of the Congo.

Acknowledgements
The authors want to thank Monkole Hospital for their scientific and logistical support, and to the study participants who generously participate in this research.We also wish to thank the financial support of the Government of Spain, the Government of Navarra and the Institute for Culture and Society of the University of Navarra.Finally, we also thank Jennifer Simoni for her proofreading and editing contributions.
Funding Open Access funding provided thanks to the CRUE-CSIC agreement with Springer Nature.Funding was provided by the Government of Spain (Fondo de Investigación en Salud-FIS) (PI16/01908), the Government of Navarre (045-2015), and the Institute for Culture and Society of the University of Navarra (EASH).The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Data Availability
The underlying data set necessary for the replication of this study, as well as the main analyses syntaxis and the codebook of variables and labels, are available within Harvard Dataverse.The original questionnaires are available as supplementary material.Informed Consent A written informed consent was obtained from each participant or their parents/guardians if they were minors.

Conflicts of interest
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made.The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material.If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Fig. 2
Fig. 2 Factors associated with paid sex The authors have no conflict of interests to disclose.This study was supported by Government of Spain (Fondo de Investigación en Salud-FIS PI16/01908), the Government of Navarre (045-2015), and the Institute for Culture and Society of the University of Navarra.Ethical Approval The study was approved by the Research Ethics Committees of Monkole Hospital and of the University of Navarra (ref: 40/2015).

Table 1
Baseline characteristics of OKAPI cohort participants interviewed at different study times STI sexually transmitted infection *Data show the characteristics of the study population that changed from baseline to follow-up

Table 2
Factors associated with reporting paid sex at baseline and at 6-month follow-up