Article Text

Download PDFPDF

Original article
High levels of unprotected anal intercourse and never testing for HIV among men who have sex with men in Nigeria: evidence from a cross-sectional survey for the need for innovative approaches to HIV prevention
  1. Lung Vu1,
  2. Katherine Andrinopoulos2,
  3. Waimar Tun3,
  4. Sylvia Adebajo4
  1. 1Population Services International, Washington, DC, USA
  2. 2Tulane University, School of Public Health and Tropical Medicine, New Orleans, Louisiana, USA
  3. 3Population Council, Washington, DC, USA
  4. 4Population Council, Abuja, Nigeria
  1. Correspondence to Dr Lung Vu, Population Services International, 1120 19 Street, NW, Suite 600, Washington, DC 20036, USA;{at}


Objectives To describe sexual risk behaviour, correlates of unprotected anal intercourse (UAI) and never testing for HIV and its implications for HIV prevention interventions among men who have sex with men (MSM) in Nigeria and other similar contexts.

Methods A cross-sectional survey was administered to 712 MSM in Abuja, Ibadan and Lagos, recruited through respondent-driven sampling (RDS). Levels of sexual risk behaviour and never having tested for HIV prior to the survey were calculated using weighted data for each city and unweighted data for the pooled sample. Correlates of UAI and never testing for HIV were determined using multiple logistic regression.

Results The risk for HIV and STI among MSM in Nigeria is high, with 43.4% reporting UAI at last sex, 45.1% never having been tested for HIV and 53.9% reporting exchange of sex for resources in the past 6 months. Correlates of UAI in multivariate analysis included living in Ibadan, marriage or cohabitation with a woman, identification as bisexual, not having tested for HIV and being HIV-positive. Correlates of not having tested for HIV in multivariate analysis included living in Ibadan, young age, less education, unemployment and report of UAI.

Conclusions HIV testing is low and associated with UAI. Findings merit targeted and innovative approaches for HIV prevention for MSM, especially access to HIV self-testing. Attention to social and structural determinants of health-seeking and sexual risk behaviour is also needed, including the criminalisation of homosexuality and social marginalisation of MSM.

  • AIDS

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


In Nigeria the prevalence of HIV among men who have sex with men (MSM) is disproportionately high compared with the national HIV prevalence, indicating a need for increased programming targeted to this population. Data from a 2007 integrated biological and behavioural, and surveillance survey (IBBSS) among MSM using respondent-driven sampling (RDS) indicated an HIV prevalence of 17.4% in Lagos, 9.3% in Kano and 1.1% in Cross River.1 A more recent survey also using RDS and serological testing reported an HIV prevalence among MSM of 34.9% in Abuja, 11.3% in Ibadan and 15.2% in Lagos.2 In comparison, the national HIV prevalence in Nigeria is estimated at 4.1%.3

The pattern of a disproportionately high prevalence of HIV among MSM within a generalised epidemic setting is consistent across African countries, including those in sub-Saharan Africa with an exceedingly high national HIV prevalence.4 However, until recently, MSM in generalised epidemic settings have not received the attention warranted to them for a number of reasons including the illegality and stigma associated with male-to-male sexual behaviours, the lack of data and the relatively small size of this population. Rather, the bulk of all HIV/AIDS spending on research and prevention services has been geared towards reducing heterosexual transmission.5 ,6 This is also true in Nigeria,7 although the trend is slowly reversing in many African countries due to increased surveillance data demonstrating the significance of the MSM population in the epidemiology of HIV.4 ,8–10 In Nigeria the 2009 Modes of Transmission study projected that 10.3% of all new infections in the country were contributed by MSM.11

This paper examines sexual risk behaviour and HIV testing to inform HIV prevention programmes among MSM in Nigeria and other similar contexts. HIV testing and counselling (HTC) provides an opportunity for prevention education, an entry point for treatment and care, and is an integral component of the national HIV strategy in Nigeria.12 It is also a critical component of combination prevention strategies, widely promoted in high prevalence settings. However, in Nigeria only 34% of MSM report ever having had an HIV test in their lifetime and only 24% report HTC in the past year.1 The findings may also shed light on the way that sexual identity intersects with class and socioeconomic factors that would affect the health and use of services by MSM.


Study design and setting

A cross-sectional survey was conducted among 712 MSM living in Abuja (n=194), Ibadan (n=210) and Lagos (n=308), recruited between August and September 2010 using RDS. The survey was administered to participants either through audio computer self-interviews (n=342 or 48%) or face-to-face interviews (n=370) based on random assignment of the interview mode. The survey elicited information about HIV risks and utilisation of HIV prevention services. Participants were also tested for HIV. The results and details of the HIV testing process have been presented in a separate publication.2 The survey instrument was adapted from a previous study and tested before the field data collection.13

Participants and recruitment

MSM were defined as men aged ≥18 years living in the cities of Abuja, Ibadan and Lagos who reported oral or anal sex with another man in the past 12 months. RDS is an adaption of snowball or chain referral sampling where peers recruit their peers into the study.14 RDS limits the number of peers that a study participant can recruit to minimise the oversampling of certain subgroups. To initiate the recruitment of participants, ‘seed’ participants were purposively selected based on a diverse set of characteristics (age, socioeconomic status and linkages to other MSM). Seeds recruited three of their peer MSM and each successive peer recruited three peers until the desired sample size was reached. In addition, RDSAT data analysis was used to produce population-based estimates.

Statistical methods

Univariate analysis was used to describe population characteristics and key risk behaviours using weighted data by city and unweighted pooled data. Weighted descriptive analysis was conducted using RDSAT while unweighted descriptive analysis was conducted using STATA. Logistic and multiple logistic regressions were used to ascertain factors associated with unprotected anal sex and not having tested for HIV prior to the survey using the pooled data. Unadjusted and adjusted ORs with 95% CIs are reported. Because the data from the three cities were combined, use of RDSAT weights in multivariate analysis was omitted (since the three networks were distinct and not linked, RDSAT weights were not applied to the pooled data). All logistic regressions were conducted using STATA V.12.


Participant characteristics

Table 1 presents the characteristics of the study population. The participants were young with a median age of 23 years (range 18–55). The majority had secondary education (65%) or tertiary education (29%); 6.5% had completed only primary education or less. Few participants lived in their own house (7.8%), with the majority living in a rented room (44.7%) or their parents’ house (42.7%). One-third of respondents (33.4%) were not earning any money. Over half of the men in the sample identified their sexual orientation as bisexual (55.4%). A slightly smaller proportion identified as gay or homosexual (42.0%), and a minority reported that they were straight (2.6%). Over a quarter of the participants lived with or was married to a woman (27%).

Table 1

Characteristics of the study population

Comparison of population-based estimates for the three cities revealed slight differences in demographic variables. However, there were no statistically significant differences across the cities in the percentage reporting unprotected anal intercourse (UAI) or multiple partnerships. Never testing for HIV was more prevalent among participants in Ibadan than in Abuja. The prevalence of HIV was higher among participants in Abuja (34.9%) than among those in Ibadan (11.3%) and Lagos (15.2%).

Description of sexual risk behaviour and HIV testing

Sexual risk behaviour was high, including UAI at last sex with a male partner (43.4%) and multiple male sex partners in the past 2 months (61.7%). A total of 37.6% of participants reported active bisexual activity (sex with both men and women in the past 2 months). Of those who had sex with women, 67% of the last sexual intercourse with women was unprotected. Slightly more than half (53.9%) reported exchanging sex to receive money, goods or favours in the past 6 months (53.9%). Almost half of MSM (45.1%) had never been tested for HIV. Of those who reported a previous HIV test, 40% had their recent test at clinics that provide male-friendly services in the three cities, followed by government and private clinics (31% and 25%, respectively).

Correlates of UAI

Table 2 describes factors associated with UAI at last sex. Increased odds of UAI in bivariate analysis were found among participants living in Ibadan compared with Abuja (OR=1.8; 95% CI 1.1 to 2.9), participants married or cohabitating with a woman (OR=1.5; 95% CI 1.1 to 2.2), those who identified as bisexual or straight compared with homosexual (OR=1.4; 95% CI 1.04 to 2.0), participants who reported sex with men exclusively in the past 2 months (OR=1.5; 95% CI 1.1 to 2.0) and among participants who had never tested for HIV (OR 2.1; 95% CI 1.6 to 2.9). Having completed tertiary education demonstrated a protective effect, with this group being 40% less likely to report UAI at last sex compared with participants with less formal education (OR=0.6; 95% CI 0.4 to 0.8).

Table 2

Factors associated with unprotected anal intercourse (n=712)

Results from multivariate logistic regression reveal similar associations, with increased odds of UAI at last sex among participants who lived in Ibadan (AOR (adjusted OR)=1.8; 95% CI 1.1 to 2.9), those who were married or cohabitating with a woman (AOR=1.5; 95% CI 1.02 to 2.2), those who self-identified as bisexual or straight (AOR=1.5; 95% CI 04 to 2.1), those who had never tested for HIV (AOR=1.9; 95% CI 1.3 to 2.7) and those who tested positive for HIV during the survey (AOR=1.8; 95% CI 1.2 to 2.8).

Correlates of never testing for HIV

Table 3 describes bivariate and multivariate analysis of variables in relation to having ever tested for HIV. In the bivariate analysis, participants who lived in Ibadan (OR=2.5; 95% CI 1.6 to 3.7) and Lagos (OR=2.0; 95% CI 1.4 to 2.9) were more likely to have not tested for HIV than those who lived in Abuja. Increased odds of never testing for HIV were also found among younger participants aged 18–25 years compared with the older age group aged 26–50 years (OR=2.5; 95% CI 1.8 to 3.6), participants completing secondary education or less compared with person with more formal education (OR=3.0; 95% CI 2.1 to 4.2), participants who were married or cohabitating with a woman compared with participants who were single (OR=1.7; 95% CI 1.2 to 2.4), those who were unemployed (OR=2.2; 95% CI 1.6 to 3.0), those who reported UAI at last sex (OR=2.2; 95% CI 1.6 to 2.9) and participants who were HIV-negative (OR=2.3; 95% CI 1.6 to 3.4).

Table 3

Factors associated with never testing for HIV (n=708)

Results from multivariate analysis reveal similar associations with increased odds of never testing for HIV among participants in Ibadan compared with Abuja (AOR=1.4; 95% CI 1.1 to 2.9), the younger age group (AOR=1.8; 95% CI 1.1 to 2.7), those with less education (AOR=2.2; 95% CI 1.4 to 3.4), those who were unemployed (AOR=1.9; 95% CI 1.3 to 2.7), those who reported UAI at last sex (AOR=2.0; 95% CI 1.4 to 2.8) and those who were HIV-negative (AOR=1.7; 95% CI 1.1 to 2.6) . Interview modes were not significantly associated with either HIV testing or UAI.


Principal findings

The results from this study describe levels of sexual risk behaviour and factors influencing UAI and HIV testing among MSM in three cities of Nigeria. In this study, 43.4% of participants reported UAI at last sex. The level of UAI reported by participants is consistent with reports of low consistent condom use among MSM in other West African countries15 ,16—for example, in Senegal, UAI at last sex with a man was reported at 23.2% among a sample of 501 MSM16 and, in a convenience sample of MSM in Cameroon, 57% reported UAI in the previous 6 months.15 Participants who lived in Ibadan, who were married or cohabitating with a woman, self-identified as bisexual or straight, were HIV-positive and who had not tested for HIV prior to the survey were more likely to report UAI.

A substantial proportion of MSM in our sample (43.4%) had never been tested for HIV prior to the survey. This finding is problematic given the efforts that have been made through the Men's Health Network of Nigeria to provide access to HTC for high-risk men in these three cities. However, this finding is consistent with reports from other studies among MSM in Africa that indicate testing rates ranging from 17.7% up to 65.2%. HIV testing reported among samples of MSM include 17.7% ever testing in Kano, Nigeria17; 19.2% in Zanzibar, Tanzania18; 25.3% in Mombasa, Kenya19; 26.7% in Cross River, Nigeria17; 35.2% in Malawi20; 40.9% in Lagos, Nigeria17; 54.5% in Lesotho21; 59.4% in Namibia20; 60.5% in Tanzania22; and 65.2% in Nigeria.23 This summary suggests a wide variation in testing across countries and sites within countries, which may reflect differences in levels of access to and delivery of testing services, stigma in these societies or variations in sampling. Our findings add to this empirical work by demonstrating a need to reach more MSM with HTC services using innovative approaches tailored to the population and specific context. In this regard, it is noteworthy that the proportion never having tested for HIV was higher in Ibadan, as was the frequency of bisexuality and UAI. Ibadan is a traditional indigenous and old city, while Lagos is a more modern highly cosmopolitan megacity and Abuja is a young city made up of mostly recent migrants. MSM in Abuja and Lagos have better prospects of meeting other MSM and seeking testing services because these cities have more MSM-targeted interventions and HTC centres. HIV and prevention strategies may require different approaches based on the social context and health system in each city.

Bisexuality is an important characteristic of MSM populations across African contexts.24 In this study, 55.4% of participants self-reported their sexual orientation as bisexual. Research from South African MSM suggests that internalised homonegativity is higher among MSM with a bisexual orientation, and that this has negative consequences for health-seeking behaviour including not testing for HIV and UAI.13 Cognitive approaches towards behaviour change in relation to sexual risk and health-seeking behaviour for HIV must be tailored towards an individual's stage in development of sexual identity and overall perception of self, in a context where homosexual orientation is incongruent with larger group norms.

The results of this study indicate that MSM at highest risk for HIV based on the report of sexual risk behaviour are not being reached with HIV testing services. Participants who had never been tested for HIV were more likely to report UAI. There was no difference in HIV testing between participants who reported exchange of sex for resources and those who did not. This is a gap in HTC coverage among a group at ‘double’ risk for HIV through both MSM exposure and sex work. Increased attention among the general population is being paid to testing modalities that might reach higher risk persons (eg, provider-initiated, outreach, couples testing, community-based approaches and self-testing).12 This same attention should be devoted to increasing HIV testing among higher risk MSM. For example, HIV self-testing ensures privacy and convenience that may by particularly important for MSM concerned about disclosure of sexual orientation when seeking HIV services. Studies in some African countries have found HIV self-testing highly accepted and feasible.25–28

Slightly more than half (53.9%) of participants reported receiving resources for providing sex in the last 6 months. High levels of commercial sex are also noted in other studies of MSM in countries in Western and Central Africa (24.4% in Lagos, Nigeria; 36.0% in Kano, Nigeria; 35% in Cross River, Nigeria,17 61.7% in Nigeria23; 20.1% in Senegal29) and East and Southern Africa (27.7% ever in Lesotho21; 74% in Mombasa, Kenya19; 29.3% in Botswana; 37.3% in Namibia; 62.6% in Malawi20; 12% in Cape Town, South Africa30; and 84% in Zanzibar, Tanzania18). A high prevalence of transactional sex indicates the need to address the increased risk of sex work among MSM, and the vulnerabilities that arise from being MSM in this context that may lead to engaging in sex exchange.

Persons with less education and no employment were less likely to test. These barriers to accessing services are not unique to MSM, and rather demonstrate that sexual orientation intersects with other individual characteristics related to class and economics in determining health service utilisation. Thus, while it is important to consider that psychosocial barriers to testing uptake are highly relevant for MSM (eg, stigma), other commonly noted access barriers including cognitive (knowledge about health benefit of HIV testing), economic and geographical barriers should also be considered.12 In fact, increased intensity of HTC services and a focus on reducing these barriers is important in Nigeria overall, as only an estimated 11.7% of women and men aged 15–49 were tested for HIV and received their result in the last 12 months.11

Persons who were younger were less likely to have ever tested for HIV in our sample, similar to results from a study in Tanzania among MSM which found that younger participants were less likely to have ever tested.22 This finding makes intuitive sense, as older age implies cumulative exposure and access to testing over a longer period of time for HTC. It is important to note, however, that overall the age of our sample was young, with a mean age of 25 years (median 23, range 18–55). Younger age representation among this sample of MSM is not unique to our study. The average age for MSM samples from African countries in the literature reviewed also ranged from 22 to 27 years.17 ,19–23 ,30 ,31 The failure to engage older MSM may be due to the use of social network referrals that may be biased towards younger age groups, who also might be more likely to socialise together and be more open about their sexual orientation. It is also possible that older men are less likely to identify as MSM. This challenge is also an important barrier to engaging older MSM programmatically given the cumulative risk of HIV in older MSM. Targeting of services towards older men should be a focus of future research and HIV prevention and testing strategies.


One limitation of this study is the possibility of bias in self-reporting of sensitive sexual risks in a homophobic and hostile environment such as Nigeria. Another limitation is that data were pooled across three cities and RDSAT weights were not used in the regression analyses. However, we have adjusted for differences across study sites in the regression analyses.

It is a challenge to draw comparisons across research available among MSM in Africa given the inherent differences in social geographies. Sampling and analysis methods used in the available literature vary substantially. The current study attempted to address this issue by using RDS sampling methodology, presenting weighted and unweighted data, and presenting data across three different cities representing different contexts in Nigeria.

Conclusions and implications

The level of sexual risk behaviour including UAI, multiple partnerships and sex work among MSM in Nigeria is high and highlights the need for targeted and innovative HIV prevention approaches for this group. The criminalisation of homosexuality in countries like Nigeria will continue to prevent MSM from accessing the basic prevention package including HIV testing. Thus, access to HIV prevention for MSM should be viewed as an urgent public health need and human right. Promotion of HIV testing among this group is warranted, and should focus on increasing testing among persons who are younger, report UAI and are unemployed with less formal education.

Key messages

  • HIV-related risks among men who have sex with men (MSM) in Nigeria are high, including unprotected anal intercourse , multiple sexual partnerships, bisexual sex, transactional sex and never testing for HIV.

  • Targeted and innovative approaches for HIV prevention for MSM are needed. Promotion of HIV testing among this group is warranted.

  • The criminalisation of homosexuality prevents MSM from accessing HIV services. Access to HIV prevention for MSM should be viewed as an urgent public health and human rights need.

  • Future studies on how sexual identity intersects with individual characteristics related to class and socioeconomic factors in determining health and health service utilisation are needed.


The authors thank the participants, field supervisors, laboratory technicians and interviewers. Special thanks go to Andrew Karlyn, Meredith Sheehy, Sandra Johnson, Emeka Nwachukwu, Apera Iorwakwagh, Oliver Anene, Denis Akpona, Segun Sangowawaand Akin Toyose.



  • Handling editor Jackie A Cassell

  • Contributors LV and KA conceived, analysed and drafted the manuscript. LV, WT and SA designed the survey and coordinated the field data collection. All authors interpreted the findings and approved the final manuscript. All authors had full access to the data and take responsibility for the integrity and accuracy of the data analysis.

  • Funding The study was funded by the UK Department for International Development (DFID) through the Enhancing Nigeria's Response to HIV/AIDS (ENR) led by the Society for Family Health (SFH Nigeria). The findings and recommendations of this study are those of the authors and do not necessarily reflect the views of the funder.

  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval Ethical approval was obtained from the Institutional Review Boards of the Population Council and the Nigerian Institute of Medical Research.

  • Provenance and peer review Not commissioned; externally peer reviewed.