Is male sex workers(MSW) more susceptible to HIV than sex buyers(SB)?

Introduction: As both parties to commercial sex, MSW and SB are both at high risk of HIV infection. However, there are few studies on the comparison of HIV infection rates between these two groups of people. Methods: The method of snowball sampling was used to recruit research participants by means of baths, bars, and the Internet. The recruited participants should cooperate with the investigators to complete the questionnaire (including general demographic characteristics, high risky behaviors, status of HIV and AIDS infection, awareness of AIDS knowledge )and blood antibody / antigen tests / oral mucosal exudate antibody tests. The difference analysis was performed using the c 2 test, and the correlates of HIV infection were analyzed using logistic regression. Results: The nal analysis included 435 MSWs and 729 SBs with HIV infection rates of 7.1% and 2.6%, respectively. Compared with SB, the HIV infection rate of MSW was 2.775 (1.340 -5.746) times higher (after adjusting for age and marital status), 3.173 (1.512-6.659) times higher (after further adjusting the awareness rate of AIDS-related knowledge), 3.688 (1.711-7.950) times higher (after further adjusting the frequency of condom use in homosexual intercourse and commercial sex in the past six months). Conclusions: MSW is more likely to infect HIV than SB. Therefore, effective preventive measures should be focused on MSW.

incarcerated people) and their sexual partners accounted for more than half (54%) of global HIV infections in 2018 [1] . Among them, sex workers have been listed as one of the key groups of AIDS infection, which means that more and more people have begun to pay attention to the role of sex workers in the spread of HIV.
Male sex workers (MSW, a subgroup of sex workers) refer to people who exchange or sell sex in exchange for money or goods. The study found that MSW population has the characteristics of younger, more mobile, more sexual partners and higher proportion of unmarried men than ordinary MSM [2] . Therefore, (n = 205) ~ 12.80% (n = 47), which was at a low level [3] . By the end of 2015, the reported HIV infection rates for MSW in cities across China ranged from 0.88% ( n=113 ) [4] to 11.50% (n = 330). [5] So far, there is still a lack of persuasive data on the status of MSW infection in China.
When it comes to the harm of commercial sex, researchers tend to pay more attention to the MSW, but few studies mention the other party of commercial sex-sex buyers (SB). In fact, the harm of SB should not be ignored. A study in the United States compared the behavior and attitude perception between SB and MSW , but did not focus on their dangerous behavior and the prevalence of HIV [6] . While a domestic study compared the general characteristics and the HIV infection rate of SB and ordinary MSM [7] , however, the MSW is ignored and the results of the study are limited due to the limited sample size and the time of investigation. At present, there is a lack of in-depth study on the comparison between the MSW and SB. In order to solve this problem, our research group has conducted a study aimed at comparing the difference of the HIV infection rate between the MSW and SB and providing a reference for the control of the spread of AIDS.

Data Sources
The data of this research came from the Tianjin Bathhouse-based Health Center Program from 2011-2017. This project was funded by the President's Emergency Plan for AIDS Relief (PEPFAR) in the United States, which was the only joint pilot project in China by the China Centers for Disease Control and Prevention (CDC) AIDS Prevention Center, the United States CDC and Tianjin CDC, implementing by the Tianjin Shenlan Community-Based Organization and aimed to establish a site dedicated to MSM for HIV counselling and testing services. And the inclusion criteria for participants were self-reported as MSM and had (or one year ago) had same-sex male sex.
All participants must signed the informed consent and nished ngerprint veri cation before testing, and then entered the consultation room for professional one-on-one consultation and testing. The questionnaire survey was completed by the consultant using jargon to talk to participants. After the questionnaire was completed, blood antibody / antigen tests were performed on the participants who agreed to take the blood, and oral mucosal exudate antibody tests were performed on the participants who did not agree with the blood collection and the results of the quick test needed to wait 20 minutes. During the waiting period, the consultant conducted a risk assessment of the participants through the previous questionnaire survey and gave them targeted safety education and behavior recommendations.
Participants with negative test results will be given condoms and lubricants, meanwhile they will be encouraged to mobilize their peers to test. Participants with positive test results will also be given condoms and lubricants, what's more, their ID cards will be retained and their real names will be veri ed, and Western blot (WB) tests will be performed to con rm the diagnosis. For participants who tested positive for WB, their information was reported to the CDC where they were registered, and they were referred for standard treatment.
In addition, since 2018, through baths, bars, the Internet (QQ, WeChat, MSM dating software) and other channels, the project team began to recruit young men who have sex with men (YMSM), who also accepted the above questionnaire survey, ngerprint entry and HIV testing. Due to the particularity of the MSM, snowball sampling was used in the study.

Quality control
All the investigators are from the Tianjin Shenlan Community-Based Organization. They are MSM themselves, so they can use jargon when communicating with the MSM who came to consult, and all questionnaires were completed during the communication process. In order to ensure the privacy of the consultant, the consultation and testing were conducted in a one-to-one manner in a private room, and the nal test results are only known to the consultant himself. All above, the privacy of participants was fully protected. In addition, compared with ordinary people, investigators from Tianjin Shenlan Community-Based Organization are more professional and more likely to gain the trust of consultants. Therefore, the authenticity of the survey results is guaranteed.
Identi cation of study population According to the question: "Have you ever had commercial sex with the same sex in the past six months", those who answered yes were screened out. Then MSW and SB were de ned according to the occupation and the question "whether it is MSW". The occupation of MSW was de ned as MSW, otherwise de ned as SB. As for the question "whether it is MSW", if the answer was yes, then de ned as MSW, otherwise de ned as SB.

Data analysis
The number of cases (composition ratio) was used for statistical description of categorical variables.
Comparisons between groups of MSW and SB were performed using the c 2 test. Both univariate and multivariate analyses of HIV infection were performed using logistic regression analysis to obtain odds ratio (OR) and adjusted OR respectively. P <0.05 was considered the difference to be statistically signi cant. The above statistical analysis was achieved by SAS9.4.

Univariate analysis on factors associated with HIV infection
Univariate logistic regression analysis showed that the prevalence of HIV positive was correlated with marriage status, frequency of condom use in homosexual intercourse and commercial sex in the past six months, awareness rate of AIDS knowledge and identity of the participants (Table 3).  Every time ---*Model 1 was adjusted for age and marital status **Model 2 was further adjusted for AIDS knowledge score in addition to the variables in model 1 ***Model 3 was further adjusted for condom use in anal sex and commercial sex with the same sex in the last six months in addition to the variables in model 2

Discussion
In recent years, more and more researchers have begun to pay attention to the status of HIV infection in MSM. As a subgroup of MSM, MSW has also begun to attract attention. According to the UNAIDS report, there were approximately 1.7 million new HIV infections worldwide in 2018 [1] . Among infected people, gay men and other men who have sex with men accounted for an estimated 17% of new HIV infections globally, which was in a highly infected state. However, due to the hard-to-reach of MSW, it is di cult to carry out related research. As a result, there are fewer reports of HIV prevalence in MSW. In this study, a total of 50 HIV-infected individuals were identi ed, including 31 MSWs and 19SBs. The prevalence of MSW (7.1%) was signi cantly higher than that of SB (2.6%).
This study found that MSW (median age=26) was younger than SB (median age =37). This is similar to the ndings of Amanja Verhaegh-Haasnoot's study, which found that MSW was younger than ordinary MSM [8] . Besides that, similar ndings were found in research on female sex workers (FSW), which reported that the age of FSW was generally lower [9] . There may be several reasons. Psychologically, SB is more inclined to nd "young and beautiful" sexual partners. In terms of physiological functions, older people are not suitable to become MSW. And MSW is mostly composed of people with lower economic income [10] , while older people may have more stable jobs and incomes, so they do not need to make a living from it.
The heterosexual marriage rate of MSM in our study exceeded 25%, which was a high level in the countries currently reported [11] . This may be related to the cultural background of different countries.
Compared to other countries, the Chinese people's concept of marriage is not so open. People generally prefer to accept monogamy, so "unmarried people" are not promoted in China. And due to the widespread stigmatization of MSM in the world [12] , some MSM may use marriage as an umbrella for their homosexual sexual orientation [13] . Moreover, less than 10% of MSW were reported in heterosexual marriage in this study, and the proportion of SB was more than 90%.This is similar to the ndings of Weiming Tang's study [14] , which found that the marriage rate of MSW people was lower than that of ordinary MSM. The reason may be that the nature and content of MSW's work are contrary to the moral requirements in traditional Chinese culture, so that choosing to marry the opposite sex is not a responsible performance.
Besides, in homosexual and commercial sex, there were only a few people who reported never using condoms, and a high percentage of people use condoms each time, which is similar to the ndings of Bea Vuylsteke's research [15] . This may be related to the active publicity and education of relevant departments and the increase of awareness of disease prevention among MSM. Overall, the frequency of condom use in MSW was lower than that in SB, which is similar to the results of Stefan David Baral [3] . On the one hand, it may be because MSW does not have the right to choose whether to use a condom or not as the party providing services when commercial sex occurs [16] ; On the other hand, it may be due to the higher demand and attention to money and the lower attention to safety in MSW after long-term sex work.
We found that singles, low frequency of condom use in same-sex and commercial sex, low awareness rate of AIDS knowledge, and MSW identity were risk factors for HIV infection. After adjusting age, marital status, AIDS knowledge awareness rate, and the frequency of condom use when homosexual and commercial sex occurred in the past six months with a stepwise approach, the risk of MSW infection is still higher than SB, and the results are always statistically signi cant. Therefore, MSW can be considered as a risk factor for HIV infection compared to SB identity. This is similar to Heide Castañeda's research, which found that MSW was more susceptible to HIV infection than ordinary MSM [17] . The reasons may be as follows. First of all, compared with SB or other MSM population, MSW has more sexual partners and higher frequency of sexual intercourse [18] , which means more source and chance of transmission and higher risk of infection. Second, the use of condoms is currently recognized as one of the effective ways to prevent the spread of diseases (e.g, AIDS, STD). However, MSW does not have the right to choose to use condoms when they have commercial sex. They are more inclined to meet the requirements of SB to obtain more economic income [19] . Third, MSW faces more widespread discrimination and stigma, so they may have no access to services or be ashamed to seek services [20] . Fourth, pre-exposure prophylaxis ( PrEP ) [21] and post-exposure prophylaxis ( PEP ) [22] are currently considered as effective preventive and remedial measure before and after high-risk behaviors occurrence. But because these two drugs are expensive (for example, in China PrEP costs 1980 yuan / month) and most MSW is a low-income group, they do not have the economic strength to buy such expensive drugs. Therefore, before and after the occurrence of high-risk behavior, they have no way to reduce the risk of possible infection, which ultimately leads to a high level of infection in MSW. Limitation 1. Our study used snowball sampling with a degree of selection bias. However, due to the particularity of MSM, other methods of probability sampling are not applicable.
2. Due to the review of the situation in the last six months and one year in the questionnaire, it is inevitable that there will be some memory bias.
3. Unfortunately, due to limited survey conditions, we did not obtain samples of MSW and SB pairings.

Conclusions
In conclusion, MSW is more likely to infect HIV than SB, although they are both involved in commercial sex. Therefore, effective preventive measures should be focused on MSW.