AIDS Behav (2013) 17:1288–1295 DOI 10.1007/s10461-012-0398-4 ORIGINAL PAPER Unprotected Receptive Anal Intercourse Among Men Who have Sex with Men in Brazil Gustavo Machado Rocha • Lı´gia Regina Franco Sansigolo Kerr • Ana Maria de Brito • Ines Dourado • Mark Drew Crosland Guimara˜es Published online: 17 January 2013 Ó Springer Science+Business Media New York 2013 Abstract The aim of this study was to assess factors Introduction associated with unprotected receptive anal intercourse (URAI) in a sample of MSM recruited by respondent dri- According to reports from the National Department of ven sampling in Brazil. Among 3,449 participants, 36.5 % STD, AIDS and Viral Hepatitis of the Brazilian Ministry of reported URAI. Final logistic model indicated that living Health (ND-STD/AIDS) [1], HIV prevalence is estimated with a male partner, illicit drug use, having stable part- in 0.6 % among the general adult Brazilian population. nership, having sex with men only, having few friends Nonetheless, the HIV epidemic is considered to be con- encouraging condom use, and high self-perceived risk for centrated with men who have sex with men (MSM) as the HIV infection were characteristics independently associ- main affected group, despite high prevalence among other ated with URAI. Intervention strategies should focus on the vulnerable populations such as injection drug users and role of anal sex practices on HIV transmission, address female sex workers in some regions of the country [1]. illicit drug use, stigma and expansion of HIV testing and Unprotected receptive anal intercourse is known to be the care. riskiest practice for acquisition of HIV infection during sexual intercourse, as the entrance of the virus in the host is Keywords HIV Á Sexual behavior Á Vulnerable enhanced by the fragility of the rectal mucosa and by the populations Á Brazil absence of local humoral immune protection. In a meta- analysis of studies that evaluated HIV risk transmission among serodiscordant couples [2], most of them among G. M. Rocha (&) MSM, the probability of transmission due to unprotected Campus Centro-Oeste Dona Lindu, Federal University of Sa˜o receptive anal intercourse was estimated in 1.4 % (95 % CI Joa˜o Del-Rei, Sebastia˜o Gonc¸alves Coelho, 400-Chanadour, 0.2–2.5 %) per-act, and in 40.4 % (95 % CI 6.0–74.9 %) Divino´polis, Minas Gerais 35501-296, Brazil e-mail: [email protected] per-partner. This could help explain the disproportionate HIV disease burden among MSM worldwide [3]. L. R. F. S. Kerr Recent data show decreasing rates of condom use by Department of Community Health, Federal University of Ceara´, MSM worldwide, possibly indicating that preventive Fortaleza, Brazil strategies may not be effectively working. In Brazil, a A. M. de Brito comparison of results from four different surveys (1995, Aggeu Magalha˜es Research Center, Oswaldo Cruz Foundation, 1998, 2002, and 2005) among MSM in the city of Fortaleza Recife, Brazil (400, 200, 401 and 406 participants, respectively), the I. Dourado proportion of risky sexual behavior varied from 31.4 to Institute of Collective Health, Federal University of Bahia, 54.6 % [4]. Reasons for the potential decrease in condom Salvador, Brazil use and the high rates of unprotected sex among MSM may include an increasing optimism related to antirretroviral M. D. C. Guimara˜es Department of Preventive and Social Medicine, Federal treatment efficacy, the emergence of the internet as University of Minas Gerais, Belo Horizonte, Brazil potential risky environment for sexual encounters, and 123 AIDS Behav (2013) 17:1288–1295 1289 structural deficiencies in the organization of health services Population and Procedures offered to MSM [3]. Many characteristics have been associated with incon- Respondents were MSM with at least 18 years old who lived sistent condom use by MSM, including sociodemographics in the following cities: Manaus, Recife, Salvador, Belo (e.g., low income and schooling), behavioral (e.g., alcohol Horizonte, Rio de Janeiro, Santos, Curitiba, Itajaı´, Brası´lia, and illicit drug use), personality traits (e.g., sensation- and Campo Grande. The cities were a priori defined by the seeking, impulsivity), contextual (e.g., poverty, homopho- ND-STD/AIDS taking into account regional, socioeco- bia and racism) and factors related to the social network nomic, and cultural diversity. Potential participants should (e.g., having supportive relationships). Fisher and Fisher have had at least one sexual relationship with another man in (1992) [5] proposed a comprehensive theoretical model in the twelve months preceding the interview. For this analysis, which information (e.g., HIV transmission and prevention only participants who reported a sexual contact with a man knowledge), motivation (e.g., perceived social support, in the previous six months were included. self-perception of vulnerability), and behavioral skills (e.g., Due to the difficulty of accessing MSM in surveys, alter- ability to negotiate with partner, to act publicity, to refuse native sampling techniques are necessary to obtain satisfac- to use drugs before sexual contacts) would be the major tory results. Therefore, Respondent Driven Sampling (RDS) determinants of safer sexual behavior. was used to obtain the desired sample in each center (previ- In Brazil, surveillance of the HIV epidemics is mostly ously defined between 250 and 350 participants per city). based on reporting of AIDS cases rather than on HIV RDS [7] is a sampling technique used to address hard to reach infection, except for specific groups such as pregnant populations, where recruitment is carried out by participants women. This potentially limits grasping the real extent of themselves using a dual incentive system, starting with pre- the epidemic in the general population and in several viously chosen participants, i.e., seed participants. In this vulnerable subgroups. In this regard, the ND-STD/AIDS study, the seeds were selected during a preliminary formative gave priority for monitoring and evaluation of the HIV/ research, when subjects of different age and socioeconomic AIDS epidemic [1], and recent studies were conducted classes were included. In each city each participant received among MSM, female sex workers and illicit drug users in three unique coupons, non-falsifiable, to distribute to their order to establish behavioral and prevalence baseline data peers. Individuals who came to the study site (usually HIV for future monitoring. Knowing the characteristics of sex- Testing and Counseling Centers) with a valid coupon and ual behavior among MSM and its determinants is crucial to who met the inclusion criteria were considered the first determine the burden of the problem and to plan new ‘‘wave’’ of the study. After carrying out research procedures, strategies for preventing HIV and other sexually transmit- each participating subject also received three coupons to ted diseases (STD) transmission. Thus, the purpose of this invite new acquaintances, repeating this process thereafter paper was to describe overall sexual behavior characteris- until the desired sample size was reached in each city. tics and to analyze correlates of unprotected receptive anal Data collection was conducted through a semi-structured intercourse in a sample of MSM in ten Brazilian cities. face-to-face interview, and it was composed of questions Moreover, the results can be used as a reference for future regarding sociodemographic data, sexual behavior and drug research to be developed to monitor the dynamics of sexual use, network and social context, health care, access to con- behavior in this population in Brazil. doms, HIV testing and information about STDs. Patients were also invited for HIV and Syphilis testing. Outcome and Explanatory Variables Methods For this analysis, the outcome of interest was defined as Study Design unprotected receptive anal intercourse (URAI), i.e., no condom use during all receptive anal intercourse in the six Cross-sectional analysis of data obtained from a national months preceding the interview. study of MSM in 10 Brazilian cities in 2008–2009 [6]. The Explanatory variables were evaluated in four main main objectives of the national study were to assess sexual groups: behavior, attitudes and practices of MSM and to estimate the prevalence of HIV and syphilis infections. The project 1. Sociodemographic: age, race, schooling, conjugal was approved by Ceara´ State University Ethical Commit- situation (single/living alone, married/living with a tee, the National Ethical Council (CONEP no. 14494), and female partner, or living with a male partner); by all institutions involved. All participants signed an 2. Behavioral: current alcohol intake (never/eventual and informed consent before answering the questionnaire. two or more times per week) and any illicit drug use, 123 1290 AIDS Behav (2013) 17:1288–1295 number of sexual partners in the six months preceding the interview and were included in this analysis. On the interview, gender (male, female, transvestite), and average, there were 15 (range 8–20) waves of recruitment type of sexual partnership (stable, casual or in each city, and approximately one-third of the individuals commercial); recruited by their peers returned to the project. Pooled HIV 3. Characteristics related to social context: receiving prevalence was estimated in 14.2 % (varying from 5.2 % encouragement from friends to use condoms, and self- in Recife to 23.7 % in Brası´lia) [6], while 36.5 % of the reported sexual identity, which was categorized as participants practiced unprotected receptive anal sex in the heterosexual, bisexual or homosexual/MSM/Gay; last six months, varying from 24.4 % in Itajaı´ to 40.1 % in 4. Characteristics related to health services: self-per- Curitiba. ceived risk of HIV infection, HIV transmission Most of the sample (57.5 %) consisted of MSM over knowledge, and prior HIV testing. Self-perceived risk 25 years old (24.1 % in Manaus, 33.4 % in Salvador, of HIV infection was assessed by a direct question 46.1 % in Campo Grande, 47.1 % in Itajaı´, 49.4 % in with four response options, and then classified into two Recife, 52.3 % in Belo Horizonte, 65.3 % in Curitiba, categories (no/little or moderate/high risk).