Worldwide, the incidence of HIV-1 has decreased by 38% since 2001.1 In Tanzania, HIV prevalence in the general population decreased significantly from 7.0% in 2004 to 5.1% in 2012.2 This success in HIV programming has been ascribed to the strengthening and scaling up of antiretroviral treatment and concerted efforts in behavioral intervention.3–5 Despite significant decrease in HIV infection due to efforts taken to address the epidemic, little attention has been given to population at higher HIV-infection risk. It has been urged that, to end the HIV epidemic more quickly and to sustain the success, more focus to key population at higher risk of HIV infection is needed with use of behavioral and biomedical approaches such as pre-exposure prophylaxis (PrEP).4,6
Men who have sex with men (MSM) are among the population subgroups which remain at heightened risk of HIV transmissions in Sub-Saharan Africa. Limited number of studies have revealed that the rate of HIV infection among MSM is 2–20 times higher than that of the general population.7–10 Moreover, geographical variations in the practice or risk behaviors because of differences in sociocultural and legal contexts have been reported across Africa, calling for population-specific studies.9–12
Most available studies in the region have used small sample size and sometimes lacking external validity and power.10,13,14 We conducted the largest so far in Africa, integrated biobehavioral survey among MSM in Dar es Salaam, Tanzania to explore their HIV prevalence and risk behaviors to provide population- and context-specific data to inform programming including possible PrEP.
This was a baseline cross-sectional study design using a respondent-driven sampling (RDS) method to recruit participants from various strata of MSM community in Dar es Salaam.
The project was conducted in Dar es Salaam, the largest city in Tanzania with a population of approximately 5 million people. The city has been among the areas in Tanzania where the prevalence of HIV infection has been consistently high. Currently, the prevalence of HIV in the general population in Tanzania stands at 5.1%, whereas that of Dar es Salaam is 6.9%.2 No published data on size estimation of MSM in Dar es Salaam but a consensus triangulation data indicated that Tanzania might have 49,000 MSM (range 41,000–71,000). Participants' recruitment took place from April to August 2014 and involved all the 3 municipalities of Dar es Salaam, namely, Kinondoni, Ilala, and Temeke.
Men were eligible to participate in this study if they were aged 18 years and above, currently living in Dar es Salaam (residents) and who may, at times or regularly, have sex with a man (or men). Participants were directly asked whether they have ever had sex with a man and whether they have had sex during the past 6 months preceding the study. Residence was based on having an address in Dar es Salaam and having lived in the city for the past 6 months preceding the survey.
The sample size was calculated in reference to the prevalence of HIV among MSM in Zanzibar of 12%.15 We hypothesized that the HIV prevalence among MSM in Dar es Salaam to be 25% due to higher background prevalence in the general population in Dar es Salaam. A sample size of 747 was adequate to achieve an 80% power to detect a difference (P1-P0) of 10% point with an actual significance level achieved by this test of 0.0495.
RDS, a method developed for the sampling of populations for which a sampling frame cannot be constructed, was used to recruit participants. RDS builds on a mathematical model which provides a theoretical basis for estimation of population proportions and their variances through statistical adjustment. When certain assumptions are met, RDS will asymptotically approach unbiased estimates of characteristics in the population under study. First-order Markov modeling allows calculation of sample weights and standard errors that adjust for the otherwise biased recruitment pattern.16
Five seeds representing different sociodemographic groups were identified particularly among members of our initial studies or other studies of MSM in Dar es Salaam.17,18 Each seed invited 3 other members of the same population to form a recruitment chain which was measured in waves. Those who accepted the invitation were in turn enrolled into the study, and also given 3 recruitment coupons to pass on to their acquaintances. This process was repeated until the desired number of participants was reached. Both the sample size and equilibrium were reached after 5 waves of recruitment chain.
Face-to-face interviews using a structured interview guide were conducted by trained interviewers (not MSM) using a pilot-tested survey tool. The interview guide collected information on sociodemographic characteristics, sexual practices, HIV risk perception, and alcohol and substance use. All interviews were conducted in Swahili, the language spoken by practically all Tanzanians. The interviews were collected in careful selected venues within the city in collaboration with MSM community. Blood specimen collection was performed after the interview and each participant was invited to a discussion about HIV/sexually transmitted infection (STI) and provided with sexual practice-relevant health education. Blood specimen collection for biological testing was collected from all consenting participants regardless of history of HIV and STI testing.
HIV serostatus was determined using Alere Determine HIV-1/2 assay (Alere Medical Co., Ltd., Matsudo Chi Shiba, Japan). Reactive samples were confirmed on a second rapid assay; Uni-Gold HIV-1/2 (Trinity Biotech Plc., Bray, Ireland). Discrepancies between the first and second test were resolved by Enzygnost HIV Integral II Antibody/Antigen ELISA (Siemens, Munich, Germany). Screening for the presence of hepatitis-B surface antigen (HBsAg) was performed using the SD Bioline rapid test (Standard Diagnostics, Inc., Yongin, Korea) and reactive samples were confirmed on a microparticle enzyme immune-assay (Abbott, Germany). Syphilis screening was performed using the venereal disease research laboratory (Omega Diagnostic, Alva, United Kingdom) assay and reactive samples were confirmed by treponemal particle hemagglutination assay (Omega Diagnostic). Herpes simplex virus 2 (HSV-2) serostatus was determined using HSV-2 IgG ELISA (Abbott Murex, Maidenhead, United Kingdom) and reactive specimens were tested further on HSV-2 IgM ELISA to identify active infection.
Data analysis for this study was performed using RDSAT software package for analysis of RDS data19 together with STATA version 14 for Windows. To control for selection probability for each participant, data were weighted according to network size by calculating weights as the inverse of the participants' network size.16,20 To reduce clustering and ensure that the whole sample was reflected in the analysis, we multiplied the weight by the sample size and divided it by the sum of the weights. Categorical variables were summarized by calculating proportions and mean values and standard deviations were used to summarize continuous variables. To identify independent association between HIV serostatus and various risk factors, logistic regression models were built. Variables with P < 0.2 in the bivariate analysis were included in the multivariable logistic regression model. It is important to note that analysis of risk factors emanate from cross-sectional study design, which inherently limits conclusion of causality. All analyses were 2-tailed and the significance level was set at 5%.
The study was ethically reviewed and approved by the Muhimbili University of Health and Allied Sciences Ethical Committee. All participants provided written informed consent for intervenes and blood sample collection. To maintain confidentiality and to enable effective collection of biological samples, interviews were conducted within the premises of the university referral laboratory. All participants were given appointments to come for collection of their test results from the same interviewers who provided the pretest counseling. Participants who had a positive STI result were treated and those with a positive HIV test result received posttest counseling were assisted to access health care services from a convenient and friendly health facility of their choice.
A total of 753 MSM with a mean age of 26.5 (SD ±6.6 years) participated in the study. Approximately half (51.8%) were youth (younger than 24 years), whereas 10% were older than 35 years. The majority (83.2%) were single at the time of the interview and only 6.3% were married. Almost a third (29.5%) of the respondents had children, the majority (72%) of whom had only 1 child.
Almost all (98.9%) had formal education, although 16.1% were primary or secondary school dropouts. A third (31.3%) of the study participants reported to be living with their parents or guardians, 34.4% were living alone, and 19.6% were living with other relatives (Table 1).
Practice of HIV-Related Risk Behaviors
Age at first sexual debut varied from 6 to 32 years with a median of 16 years (IQR 14–18 years). About a quarter (195; 26%) of the respondents reported to have had their initial sexual intercourse with a man or a woman before the age of 15 years, whereas about a third (239; 31.7%) had engaged in sex for the first time after reaching 18 years of age. Almost two-thirds (483; 64.1%) had their first sexual experience with a female sexual partner, whereas 39% (n = 294) said their first sexual intercourse was with another man. A significantly higher proportion of those who had their first sexual intercourse before the age of 15 reported that the first sexual partner was a male compared with those who had their first experience at a later age (P < 0.001).
A small proportion of the study participants (118; 15.7%) had taken any precaution against HIV/STI infection during their first sexual intercourse and the lower the age at first sex, the less likely the person was to have used any protective measure against HIV/STI infections. In this population, 62.7% (n = 472) of the participants reported to have taken an HIV test before and of these, 92.9% (n = 438) attended posttest counseling and got their results. However, participants were not inquired about their test results.
Sexual position has a bearing on the risk of HIV infection among MSM. In our study population, about two-thirds (60.8%) of the participants reported to take the insertive position, 31.8% assumed the receptive position, and only 7.4% reported both insertive and receptive positions.
Among the 84% (n = 632) who reported ever having sex with a woman, 84.1% had 3 or more lifetime partners, 8.2% had 2 partners, and 7.7 had only 1. Among the 79.2% (n = 596) of men who had sex with a woman within the last 3 months, 27.9% had 3 or more female partners, 26.4% had 2 partners, 45.7% had only 1. Overall, 65.8% (n = 495) of men reported having sex with both male and female partners in the last 3 months. Among men who had sex with at least 1 female partner in the last 3 months, 25.0% had sex with 3 or more male sexual partners, 28.2% had 2, and 38% reported only 1.
Condom use with both male and female partners was not a common practice. Less than half reported that they (for insertive) or their partners (for receptive) used a condom during their last sex with a male partner. Condom use during the last sexual encounter was more common with male sexual partners (340, 45.5%) than with female sexual partners (177, 27.5%).
Although 89.1% (n = 671) of MSM reported to have ever used lubricants during their last anal sex, KY jelly (water-based lubricant) was the most common type used (32%), followed by Vaseline (26%), saliva (10%) and the remaining used other different types of non–water-based lubrications.
Involvement in group anal sexual intercourse was reported by 27.5% of the study participants and slightly more than 7% had been involved in such sexual activity in the last 3 months preceding the study.
Nearly 2 thirds (66%) used alcohol. Beer was the most common drink used (99.5%) followed by hard drinks (31.1%) and wines (6.6%). Among those who used alcohol, 40.9% did so the last time they had sex and almost a third (32.6%) said they had been drunk the last time they had sex.
Of the 753 participants interviewed, 646 (85.7%) provided blood for HIV, HSV-2, HBV, and syphilis testing. Blood specimen was collected from all participants regardless of history of HIV testing. Although the most common reason for not consenting for blood testing was fear of results, there was no difference in sociodemographic characteristics between those who consented and those who did not. The prevalence of HIV infection was 22.3% [95% confidence interval (CI): 18.5 to 26.2] and the HSV-2 prevalence stood at 40.9% (95% CI: 36.3 to 45.6). The syphilis prevalence was 1.1% (95% CI: 0.9 to 2.0) and the prevalence of HBV was 3.25% (95% CI: 2.5 to 7.0).
HIV infection and HSV-2 were highly prevalent among MSM who assumed the penetrated position during anal sex followed by those who practiced both penetrated and penetrating positions (Fig. 1).
HIV prevalence was statistically significant higher among those aged 25 years and above (349; 26.3%) and among men reporting to have no children (493; 25.9%) (Table 2).
Predictors of HIV Infection Among MSM in Dar es Salaam
We examined for independent determinants of HIV infection among MSM in Dar es Salaam using multivariable logistic regression modeling. Men aged 25 years and above had a significantly higher odds of HIV infection as compared with those aged less than 25 years. Not having children and low perceived risk of HIV infection was associated with increased odds of being HIV seropositive [adjusted odds ratio (aOR), 2.4, 95% CI: 1.4 to 4.2 and aOR, 2.6, 95% CI: 1.2 to 5.3, respectively].
Although men who were assuming both positions during anal sex (insertive or receptive) had 3 times the probability of being HIV seropositive (aOR, 3.4, 95% CI: 1.4 to 9.7), those who practiced receptive anal sex were almost 9 times more likely to be HIV seropositive as compared to those practicing insertive (aOR, 8.7; 95% CI: 1.2 to 5.3).
The results also indicated that MSM who were having sexual relationship with women had 8 times higher odds of being HIV seropositive (aOR, 8.0, 95% CI: 4.1 to 15.6) and those cohabiting with women having almost 6 times the probability of testing positive for HIV (aOR, 5.5, 95% CI: 1.6 to 8.4).
In this population of MSM, men who were engaging in group sex (aOR, 3.8, 95% CI: 1.6 to 8.4) and those infected with HSV-2 (aOR, 4.1, 95% CI: 2.6 to 6.5) were about 4 times more likely to be HIV positive then their counterpart. Moreover, not using water-based lubricants (aOR, 2.6, 95% CI: 1.0 to 4.5) and having a history of genital ulcers (aOR, 4.1, 95% CI: 1.1 to 7.2) significantly increased the odds of HIV infection among MSM in Dar es Salaam (Tables 2 and 3).
This was the first survey in Africa that recruited the largest number of MSM (N = 753) in a single city in such a short time despite the anticipated sociocultural and legal barriers.
A typical man who is having sex with another man in Dar es Salaam was a young, not married, completed primary education, who lives alone without a child and having sexual relationship with women. These findings are similar to what has been reported from other studies within and outside Tanzania.9,13,15,17,21
HIV prevalence among MSM in Dar es Salaam increased by age and was 4 times and about 2 and a half higher than that of the general population in Tanzania and the population in Dar es Salaam, respectively.22 However, these estimates are relatively low than what was reported in another study conducted 4 years ago in Dar es Salaam (22.3% versus 30.2%).14 The differences between these 2 estimates cannot entirely be explained by a possible decrease in infection rates but potential selection bias and low study power of the previous study (only recruited 200 men of the calculated 310 MSM needed in each city which was also based on a high prevalence estimates from Kenya) could have played a role. Despite the difference, MSM in this city continue to carry the largest burden of HIV infection calling for concerted efforts to scale up prevention, care, and treatment to be able to achieve the intended zero new infection.
Practice of high HIV risk behaviors including group sex was common among MSM in Dar es Salaam and this is supported by other studies elsewhere.7,9,22 Nearly half (346/753; 46%) of MSM in this population did not use condom during the last anal sex and of those who reported engaging in group sex (n = 178), 42% (75/178) reported to have had unprotected anal sex the last time they participated in group sex. The high rates of unprotected sex were similar to what has been reported earlier in the region and elsewhere.13,23 In addition, use of water-based lubricants was rare in this population, with only 32% reporting such use during the last anal sex. This reported estimate is lower than what has been reported elsewhere. Data from other studies in the country elude that higher cost, lack of knowledge on the importance of use of these lubricants, and availability could explain such a low use.9,11,24
MSM in this study reported to have high sexual contact with women (bisexuality behaviors), with 66% of them reporting having a female sexual partner at the time of the study. Contrary to what has been reported in Kenya, having sex with a women or married/cohabiting with one was associated with increased likelihood of testing seropositive for HIV in this study.9 However, these results were similar to our study involving 406 men conducted in Dodoma region, central Tanzania as well as a recent study in Malawi.24,25 Because of stigma associated with homosexuality in Tanzania and Africa at large, some MSM would be engaging sexually with women to avoid such stigma. Studies have shown that MSM who opt for such camouflage are more likely to have less self-efficacy and self-esteem for HIV prevention.12,24 Our study also indicates that a substantial proportion of MSM who engage in sexual relationship with women reports multiple sexual partnership with both men and women.11,26 These risk behaviors could partly explain the high rate of HIV infection in this group of MSM.
The link between MSM and women in the general population has an implication in the spread of the HIV epidemic and hence the development of preventive interventions.24 Targeting both populations in HIV programming is therefore crucial in ending the HIV epidemic in the country.
Although a substantial proportion of MSM in this study reported to have tested for HIV infection, approximately half (353/753; 47%) had moderate-to-low perceived risk of HIV infection. Perceived risk is an important aspect of behavioral practices and this low perceived risk may have contributed into the observed high risk behaviors and infection rates in this population.17
Biological risk factors for HIV infection could be playing a major role in the observed high HIV rates in this population. The prevalence of HSV-2 was very high and together with history of genital ulcers during the past 12 months, they were associated with increased odds of HIV infection. These results indicate that preventive intervention and HSV-2 suppressive therapy as well as PrEP should be included in the existing MSM intervention package in Tanzania.4
The results of this study should be interpreted in light of a number of limitations. First, sensitive data such as those related to sexual behaviors are prone to desirability bias and this may have underestimated some of the association presented. Second, this was a cross-sectional study design and limit temporal relationship. However, most of the factors identified have been proved by more robust studies. Last, respondent-driven sampling may, to some extent, affect the external validity of this survey. However, robust analytical methods were used to address some of the recruitment biases expected.
Results of this study highlight the fact that although HIV infection is at decrease in the general population, the rates are still high among MSM. Low perceived risk of HIV infection, practice of risk behaviors, and high prevalence of HSV-2 infection play a major role in HIV transmission in this population. HIV epidemic among MSM is highly linked to that of the general population. Intensification of ongoing intervention and use of newer biomedical approach such as HSV-2 suppressive therapy and PrEP are highly needed to address this risk group and facilitate government efforts to achieve zero new infection.
1. UNAIDS. The Gap Report. Switzerland, Geneva: United Nations Programme on AIDS; 2014.
Commission for AIDS (TACAIDS), Zanzibar AIDS Commission (ZAC), National Bureau of Statistics (NBS), Office of Chief Government Statistician (OCGS), and ICF International. Tanzania HIV
/AIDS and Malaria Indicator Survey 2011–2012. Dar es Salaam, Tanzania
: TACAIDS, ZAC, NBS, OCGS, ICF International; 2013.
3. McCoy SI, Kangwende RA, Padian NS. Behavior change interventions to prevent HIV
infection among women living in low and middle income countries: a systematic review. AIDS Behav. 2010;14:469–482.
4. Grant RM, Lama JR, Anderson PL, et al. Preexposure chemoprophylaxis for HIV
prevention in men who have sex with men
. N Engl J Med. 2010;363:2587–2599.
5. Mastro TD, Sista N, Abdool-Karim Q. ARV-based HIV
prevention for women - where we are in 2014. J Int AIDS Soc. 2014;17(3 suppl 2):19154.
6. Fauci AS, Marston HD. Ending AID-is an HIV
vaccine necessary? N Engl J Med. 2014;370:495–498.
7. Smith AD, Tapsoba P, Peshu N, et al. Men who have sex with men
/AIDS in sub-Saharan Africa. Lancet. 2009;374:416–422.
8. Wade AS, Kane CT, Diallo PA, et al. HIV
infection and sexually transmitted infections among men who have sex with men
in Senegal. AIDS. 2005;19:2133–2140.
9. Sanders EJ, Graham SM, Okuku HS, et al. HIV
-1 infection in high risk men who have sex with men
in Mombasa, Kenya. AIDS. 2007;21:2513–2520.
10. Lane T, Osmand T, Marr A, et al. The Mpumalanga Men's Study (MPMS): results of a baseline biological and behavioral HIV
surveillance survey in two MSM communities in South Africa. PLoS One. 2014;9:e111063.
11. Park JN, Papworth E, Kassegne S, et al. HIV
prevalence and factors associated with HIV
infection among men who have sex with men
in Cameroon. J Int AIDS Soc. 2013;16(suppl 3):18752.
12. Wirtz AL, Jumbe V, Trapence G, et al. HIV
among men who have sex with men
in Malawi: elucidating HIV
prevalence and correlates of infection to inform HIV
prevention. J Int AIDS Soc. 2013;16(suppl 3):18742.
13. Nyoni J, Ross MW. Condom use and HIV
-related behaviours in Urban Tanzanian men who have sex with men
: a study of beliefs, HIV
knowledge sources, and partner interactions. AIDS Care. 2013;25:223–229.
14. Michael R, Nyoni J, Ahenaku H, et al. High HIV
seroprevalance, rectal STIs and risky sexual behaviour in men who have sex with men
in Dar es salaam and Tanga, Tanzania
. BMJ Open. 2014;4:e006175.
15. Dahoma M, Johnston LG, Holman A, et al. HIV
and related risk behavior among men who have sex with men
in Zanzibar, Tanzania
: results of a behavioral surveillance survey. AIDS Behav. 2011;15:186–192.
16. Heckathorn D. Respondents driven sampling. A new approach to the study of Hidden populations. Social Probl. 2008;44:174–199.
17. Mmbaga E, Dodo MJ, Leyna GH, et al. Sexual practices and perceived susceptibility to HIV
infection among men who have sex with men
in Dar Es Salaam, Mainland Tanzania
. J Int AIDS Soc. 2012;S1:012.
18. Moen K, Aggleton P, Leshabari MT, et al. Not at all so hard-to-reach: same-sex attracted men in Dar es Salaam. Cult Health Sex. 2012;14:195–208.
19. Volz E, Wejnert C, Cameron C, et al. Respondent-Driven Sampling Analysis Tool (RDSAT) Version 7.1. Ithaca, NY: Cornell University; 2012.
20. Wejnert C. An empirical test of respondent-driven sampling: point estimates, variance, degree measures, and out of equilibrium data. Social Methodol. 2009;39:73–116.
21. Baral S, Trapence G, Motimedi F, et al. HIV
prevalence, risks for HIV
infection, and human rights among men who have sex with men
(MSM) in Malawi, Namibia Botswana. PLoS One. 2009;4:e4997.
22. Price M, Rida W, Mwangombe M, et al. Idetifying a risk populations in Kenya and South Africa: HIV
incedence in cohorts of men who reports sex with men, sex workers and youth. J Acquir Immune Defic Syndr. 2012;59:185–193.
23. Moen K, Aggleton P, Leshabari MT, Middelthon AL. Situating condoms in sexual lives: experiences of same-sex-attracted men in Tanzania
. Int J Sex Health. 2013;23:185–197.
24. Wirtz AL, Trapence G, Kamba D, et al. Geographical disparities in HIV
prevalence and care among men who have sex with men
in Malawi: results from a multisite cross-sectional survey. Lancet HIV
25. Mmbaga EJ, Moen K, Makyao N, et al. HIV
and sexually transmitted infections among men who have sex with men
in Dodoma Municipality, Tanzania
: a cross sectional study. Sex Transm Infect. 2017;93:314–319.
26. Maleke K, Makhakhe N, Peters RP, et al. HIV
risk and prevention among men who have sex with men
in rural South Africa. Afr J AIDS Res. 2017;16:31–38.