A cross-sectional analysis, nested within a cohort study was conducted at Evandro Chagas National Institute of Infectious Diseases-INI-Fiocruz (formerly known as the Evandro Chagas Clinical Research Institute-IPEC), Rio de Janeiro, Brazil. The parent cohort study was designed to evaluate the prevalence and incidence of anal HPV infection and intraepithelial anal lesions among HIV infected and uninfected men [13]. In summary, men were recruited by trained staff either while in the clinic waiting room or by phone. Inclusion criteria for the parent cohort study were: being aged 18 years or older, not having anal cancer or related treatments (surgery, radiotherapy, and chemotherapy) and a willingness to sign the informed consent form. The exclusion criteria were: the use of immunomodulators agents, such as prednisone (dosage > 10 mg/day), interleukin or interferon. HIV positive men were already under care at the INI HIV Clinic, and HIV negative men attending the clinic for HIV prevention services. At the enrollment visit, information on demographic, sexual behavior, substance use as well as history of prior STDs were collected using Audio Computer Assisted Interview (ACASI) and specimens for STD diagnostic testing were collected. The study population for the present analysis were men who have sex with men (MSM), herein defined as men that reported having at least 1 male sexual partner within the past 12 months (regardless of having female partners).
From August 2, 2010 to June 30, 2012, 391 MSM were enrolled into the parent cohort study. Of these, 72 (18.4 %) were excluded for not having a male sexual partner(s) within the last 12 months, and 27 (6.9 %) were excluded because of missing information regarding their sexual partners within the past 12 months (Fig. 1). In total, 292 MSM were included in this analysis. To determine the prevalence for each specific STD, we considered all participants who had available results for each of the STDs under evaluation, thus explaining the different denominators. The outcome “Having one or more STD at study entry” was defined as having a positive testing for at least one of the following STDs: CT (rectal, urethral, or urine), NG (rectal, urethral, or urine) or syphilis, at the study baseline visit. If a participant tested positive for at least one of the STDs under evaluation, regardless of having missing data on other STD tests, his data contributed to the analysis of the outcome “Having one or more STD at study entry”. Thus, 275 men were included in the analysis of the associated factors of having at least one STD (Fig. 1).
STD diagnosis
Rectal CT and NG infection were diagnosed using APTIMA Combo 2 assay (Gen-Probe/Hologic San Diego, CA). Testing was processed at the Johns Hopkins STD Laboratory, in Baltimore, Maryland, USA. All indeterminate results for rectal CT/NG were repeated using the same tests on the same sample. If the repeated test was conclusive, the results were reported accordingly. If remained indeterminate, the result was reported as negative. Urethral CT and NG infection were diagnosed using urine samples on the Abbott RealTime platform and the NG/CT Amplification Reagent Kit (Abbott Molecular, Des Plains, IL); these samples were processed at the INI-Fiocruz Laboratories.
The rapid plasma reagin (RPR) test was performed for syphilis screening; positive results were confirmed using a microhemagglutination assay for Treponema pallidum (MHA-TP). Titers equal to or higher than 1/8 and a positive MHA-TP constituted a syphilis diagnosis. HIV infection was diagnosed according to the Brazilian HIV diagnosis algorithm (www.aids.gov.br). HIV-negative participants were tested at the baseline visit and also at each follow-up visit. Both the INI-Fiocruz Laboratories and the Johns Hopkins University STD Research Laboratory successfully participate in the College of American Pathologists (CAP) External Quality Assurance (EQA) proficiency testing panels for all relevant testing associated with this study.
Measures
STD-related symptoms included urethral/anal discharge, anal or genital nodules, anal or genital ulcers, spontaneous anal pain, tenesmus and anal pruritus. Individuals were considered symptomatic when at least one of the above-referenced symptoms was reported.
Demographic variables included age, self-reported skin color (further classified as white or non-white) and schooling (years of formal education).
The number of male and female sexual partners within the last 12 months was determined using the following questions: “During the past 12 months, how many men and how many transvestite/transsexual/transgender(s) did you have sex with?” and “During the past 12 months, how many women did you have sex with?” The latter question was categorized as “none” or “at least one female partner”.
Stable partner in the past 3 months was defined as “Yes” if the participant considered at least one sexual partner within the last 3 months to be a stable partner (husband, wife, boyfriend or girlfriend).
HIV-positive male sexual partner in the past 3 months was determined using the following questions: a) “In the past 3 months, how many HIV-positive male partners did you have insertive anal sex with?” and b) “In the past 3 months, how many HIV- positive male partners did you have receptive anal sex with?”.
Each study participant provided information, if known, on the HIV serostatus of his sexual partners. Independent HIV testing was not performed on the partners. Alcohol and stimulant use before or during sex were determined using the following questions: “In the past 3 months, were you drunk or high before or during sex?" and “In the past 3 months, did you use either inhaled or intravenous illicit drugs before or during sex?” [14].
Commercial sex within the past 3 months was defined as the exchange of sex for money/other favors and/or seeking commercial sex workers. This was determined using the following questions: “In the past 3 months, did you have sex for money, drugs or other favors?” and “In the past 3 months, did you look for commercial sex workers?”.
Anal sexual practices with male partners during the past 3 months were evaluated using the following questions: a) “In the past 3 months, how many HIV-negative male partners did you have insertive anal sex with?”; b) “In the past 3 months, how many male partners, with an unknown HIV-serostatus did you have insertive anal sex with?”; c) “In the past 3 months, how many HIV-negative male partners did you have receptive anal sex with?”; and d)“In the past 3 months, how many male partners, with an unknown HIV-serostatus did you have receptive anal sex with?” The answers were categorized as ‘only insertive’, ‘only receptive’ or ‘both’.
Unprotected anal intercourse during the past 3 months was defined as any positive answer to the following questions: “In the past 3 months how many: a) HIV-negative male partners did you have insertive anal sex with, without using condoms? b) male partners, with an unknown HIV-serostatus did you have insertive anal sex with, without using condoms? c) HIV-negative male partners did you have receptive anal sex with, without using condoms? and d) male partners, with an unknown HIV-serostatus, did you have receptive anal sex with, without using condoms?”.
Statistical analysis
For those with results available we have described the overall prevalence of CT, NG and syphilis. These results are sorted out based on HIV status. The overall prevalence for having one or more STD at study entry’ outcome and the confidence interval for the proportion [15] were calculated and are also presented sorted to HIV status using Chi-square test for comparison. Generalized linear models using logarithmic linkage and Poisson distribution with robust variance were used to estimate the prevalence ratio between selected variables and the outcome [16]. Age, schooling, skin color, number of male sexual partners within the previous 12 months, having female partners, having a stable partner, having an HIV-positive sexual partner, having been high from alcohol and having used stimulants before/during sex, reporting commercial sex, position during anal sex, unprotected anal sex with male partner (s) in the last 3 months and HIV status were evaluated in the univariate analysis. Variables that were associated with an STD according to the univariate analysis (P < 0.25) were entered into the initial multivariate model. Multi-colinearity was tested using generalized colinearity diagnostics (GVIF). Variables were kept in the final multivariate model if (a) the P value < 0.05 or (b) it was a confounder, e.g., when removed, a change equal or higher than 10 % in the prevalence ratio of any other variable of the model was observed [17]. We forced the variable age (a priori) into the multivariate models. The STATA/SE 10.1 software was used to perform the analysis.
Ethics
The study was approved by the IPEC-FIOCRUZ IRB (CAAE 0044.0.009.000-09); all study participants signed an informed consent prior to cohort enrollment.