The Epidemiology of HIV Among MSM in Low- and Middle-Income Countries: High Rates, Limited Responses
Part of MSM, HIV and the Road to Universal Access: How Far Have We Come?
In many regions of the world, HIV first emerged among populations of men who have sex with men (MSM).1 More than a quarter-century later and in an increasingly broad range of countries, contexts, and development levels, male-to-male sexual contact remains an important route of HIV transmission. Men who have sex with men is a technical phrase intended to be less stigmatizing than culturally bound terms such as gay, bisexual, or homosexual. It describes same-sex behaviors between men rather than identities, orientations, or cultural categories. Therefore, the term MSM includes gay men, bisexuals, MSM who do not identify as gay or bisexual despite their behaviors, male sex workers, transgendered people, and a range of culture- and country-specific populations of MSM. MSM belonging to these diverse populations may have both individual and network-level risks and these groups may have diverging HIV epidemic dynamics.
It may be a long time before the differential risk status of these various populations can be accurately described, given the homophobia that is prevalent in many countries and the limited funding available for studies and programs targeting MSM. Nevertheless, recent data indicate that HIV prevalence among MSM is high and rising across these groups, and that these epidemics are no longer limited to the high-income countries in which they were initially described. These studies have demonstrated high HIV prevalence among MSM from a number of low- and middle-income country settings.2 In certain countries -- such as Thailand, Cambodia, and Senegal -- with relatively low and declining HIV prevalence among heterosexual populations but high prevalence among MSM, the data suggest concentrated HIV epidemics among MSM and a "dislinked" epidemic pattern.3-7 These men exhibit ongoing high HIV prevalence against a backdrop of declining general population epidemics. Despite these significant findings, this continues to be an understudied and underserved population.
Research among MSM in low- and middle-income countries has been limited by the criminalization and social stigmatization of their behavior, safety considerations for study participants, the hidden nature of this population, and lack of targeted funding. Thus, most available data evaluating determinants of HIV risk among MSM are derived from high-income countries. Available evidence from these countries suggests that structural risks -- social, economic, political, or legal factors -- are important in defining HIV risk for any one man. Individual-level acquisition risks have focused on the highest probability exposure: unprotected anal intercourse, and specifically on receptive anal intercourse.8 Use of "party" or "club" drugs has been associated with heightened sexual exposure risk among MSM, and, as with men who only report sex with women, HIV transmission in MSM is associated with genitourinary disease. However, high frequency of male partners and a high lifetime number of male partners are also relevant.7-9
There are strong data supporting the preventive efficacy of circumcision among heterosexual men, but among MSM there exist only limited observational data regarding the possible protective effect of adult circumcision on HIV acquisition.9-18 Being a black or minority ethnic man who has sex with men in high-income country settings is associated with a higher risk of HIV compared to white MSM.19 A critical review of the evidence examining the racial differential seen in the HIV epidemic among MSM suggested that the increased HIV prevalence seen among black MSM is most likely due to the fact that a lower proportion of them have been tested and know their HIV status, as well as to higher rates of sexually transmitted infections (STIs), which facilitate HIV transmission.20 These individual risk factors likely transcend geography, whereas the higher-level structural risk factors vary significantly between countries and continents. In particular, high-risk behaviors such as those common among male sex workers, transgendered people, and MSM who inject drugs, likely put all the members of a sexual network at increased risk of infection.7 A high prevalence of STIs increases the probability of HIV transmission within a network. At the community level, access to prevention services, voluntary counseling and testing, and antiretroviral treatment (ART) can help diminish risk within MSM communities. Finally, the more advanced an HIV epidemic is, the greater the risk to lower-order determinants of HIV infection.
In countries such as the U.S., HIV prevention and treatment efforts have been mainstreamed to target the general population. However, as of 2005, 72% of all HIV infections among men in the U.S. were related to MSM.21 MSM are the only vulnerable group with no significant decrease in transmission rates in the U.S. from 2001 to 2004.22 Active surveillance in Baltimore has demonstrated HIV prevalence as high as 46% among African-American MSM, 67% of whom were unaware of their status. HIV prevalence was 21% among white MSM, 18% of whom were unaware of their status. These data suggest that HIV continues to have a disproportionate impact on MSM in the U.S., and that these epidemics continue to grow.
To see if this disproportionate HIV burden also affected MSM in lower-income settings, in 2006 we examined a random sample of low- and middle-income countries and found some notable trends. First, it was difficult even to find studies of the prevalence of male-male sexual contact in lower-income settings, and second, where HIV data were available, prevalence was consistently high. To be able to draw more concrete conclusions, we completed a systematic review and meta-analysis of this topic in 2007. This review confirmed that HIV has spread widely in Asia, Africa, and Latin America, and that MSM are at increased risk of HIV infection compared to the general population of reproductive age adults, even in settings with high HIV prevalence.
The review evaluated 63,538 men from 38 countries and demonstrated an overall HIV prevalence among MSM of 12.8%. Studies from the 16 Latin American countries included 38,013 MSM with an overall HIV prevalence of 16.1%, compared to a general regional population HIV prevalence of 0.5%. Studies from 10 Asian countries included 19,142 men with an HIV prevalence of 11.4%, compared to a general regional prevalence of 0.1-0.3%. Though there were no published studies from the former Soviet Union and Eastern Europe, data were available from 12 countries in this region and demonstrated an HIV prevalence among 8,609 MSM of 1.2%, compared to a general regional prevalence of 0.9%. Limited data were also available from Africa, where studies from four countries included 2,353 MSM with an HIV prevalence of 13.0%, compared to a general regional prevalence of 5.0%.
Looking at the more widely available data from Latin America, there have been significant differences within the region in the responses to HIV epidemics among MSM, which have likely resulted in different epidemic dynamics, even between neighboring countries. For example, in Peru, MSM are included in national HIV surveillance programs, and targeted HIV prevention expenditures match the relative burden of disease among MSM. HIV prevalence among MSM sampled was 12.2% with an odds ratio (the probability of being HIV positive relative to the general population) of 22.6 times greater than the general population.23 This can be contrasted to Bolivia, where programming and spending on MSM as a proportion of total HIV prevention expenditures are less than one half of the proportion that MSM contribute to the country's HIV epidemic. In Bolivia, HIV prevalence among MSM was 21.5%, with an odds ratio of 178.8 above the baseline general population prevalence. These differences between countries are likely related to structural risk factors for HIV infection among MSM, rather than individual level risk factors. With some exceptions, Latin American countries have included indicators on MSM in their national AIDS strategies, but the high HIV prevalence speaks to the continued need for expansion within these programs to be in line with the relative HIV burden among MSM.
Similar to Latin America, HIV in Asia tends to be highly concentrated among subpopulations, including MSM. Many of these epidemics seem to be occurring separately from what is happening in the general population. Consequently, the regional average probability of being infected with HIV is 33.3 for MSM in Latin America, and 18.7 for MSM in Asia. In both cases, the HIV infection risk for MSM is much greater than for the general population. Given that Asia makes up the majority of the global population, it is surprising that, as of 2007, data were only available from 19,142 men from seven countries in the region. The vast and diverse Asian continent contains very different HIV epidemics, as can be seen by comparing countries such as Thailand and China. In Thailand, HIV prevalence among 3,236 MSM sampled was 24.6%, compared to a prevalence rate of 3.8% among 6,270 MSM in China. However, while the absolute risk among MSM of being infected with HIV was higher in Thailand, MSM in China were at higher risk of HIV infection compared to the general population. Specifically, while MSM in Thailand were approximately 20 times more likely than the general population to have HIV, MSM in China were more than 45 times more likely than the general population to be HIV positive. Thus, while programming has tended to focus on the absolute risk of HIV infection, it is also key to consider the relative risk of HIV infection among MSM in these settings. In Asia, prevention expenditures targeting MSM range from barely more than zero in parts of China to a high of 4% of all prevention expenditures in Thailand, highlighting the ongoing disparity between the burden of disease among MSM and the level of spending on prevention programs for this population.23
The high prevalence of HIV infection and high odds ratios among MSM are quite consistent across most individual countries and geographic regions, as well as all epidemic states (low-level, concentrated, and generalized). Eastern Europe appears to be an exception: data on MSM are scarce, and the region's HIV epidemics are primarily driven by injection drug use. Since an unknown but potentially significant number of MSM in this region may also be injection drug users, it may be difficult to estimate the attributable risk fraction -- the portion of the total burden of the epidemic that can be attributed to a particular cause -- for these differing behaviors. What is clear is the need for better characterization of the risks for MSM in this region and for the development of effective prevention programs to curb these epidemics.
Data regarding MSM in Africa are the sparsest in the world, but are beginning to emerge. One of the earliest studies was published in 2005 in Senegal, where 463 MSM from Dakar and four other urban communities demonstrated an HIV prevalence of 21.5%. STI prevalence among MSM was 4.8% for active syphilis, 22.3% for herpes simplex virus 2, 4.1% for chlamydia, and 5.4% for gonorrhea. A 2005 study of 713 receptive MSM from Khartoum, Sudan, revealed a mostly Muslim population with an HIV prevalence of 9.3%. The best-developed data have been generated in Kenya, with the support of the Kenyan National AIDS Council. Groups throughout the country have shown HIV prevalence as high as 43% among MSM.24-26 In Nigeria, a recent study hosted by the Ministry of Health characterized the risk of MSM across the country and found that their overall HIV prevalence was 13.5%, though prevalence varied significantly between sites of study.27 Specifically, HIV prevalence among MSM in Cross River was 2.8%, 11.7% in Kano, and 25.4% in Lagos. The combination of these studies suggests that even in the generalized HIV epidemics of sub-Saharan Africa, MSM are nearly four times more likely to be infected with HIV than the general population.
Data on HIV prevalence among MSM are currently being generated or analyzed in Botswana, Ghana, Ivory Coast, Malawi, Namibia, South Africa, and Zambia, among other countries. While these data are preliminary, they clearly demonstrate that MSM not only exist in Africa, they are also at risk for HIV infection and in need of targeted prevention programming.
Decreasing the relative burden of disease among MSM will require a concerted effort and a strategic approach. We suggest that any such strategy should include at least three main components: increased surveillance, enhanced research, and targeted prevention programs.
Surveillance is the ongoing collection, collation, and analysis of data and the timely dissemination of information to those who need it. Surveillance of MSM in lowand middle-income countries to date has been largely carried out through research, and only a few countries have included MSM in national surveillance programs. However, where available, prevalence data have demonstrated a consistent and disproportionately high burden of HIV among MSM. National surveillance systems should consider this high burden of disease and include MSM in countries where they are currently excluded. Methodologically sound surveillance can help determine and demonstrate the need for targeted HIV prevention expenditures from regional, national, and international funding agencies.
Prevention expenditures should be allocated according to evidence-based need. To this end, it is important to use research to generate data proving that HIV epidemics in low- and middle-income countries are real. Enhanced research can also inform the design of prevention strategies, and eventually serve to evaluate these prevention programs after their initiation.
Given documented high HIV prevalence among MSM, it is also vital to enact targeted and evidence-based prevention programs for these men. The goal of these programs is to decrease HIV transmission among men by increasing condom use during anal sex and employing other evidence-based biomedical interventions. We already know that these prevention strategies can work.
A recent systematic review and meta-analysis including 16,224 men in 38 experimental and observational studies demonstrated that compared to controls with no interventions, study groups reduced unprotected anal intercourse by 27%.28,29 In an additional 16 studies where MSM were given targeted prevention strategies, study groups decreased unprotected anal intercourse by 17%, compared to MSM who received standard HIV prevention measures.
Prevention strategies tend to work better when community-level rather than individual risks are targeted. These strategies functioned equally well independent of the proportion of minorities included. Globally, only 5-10% of MSM have access to programs such as these, with the majority taking place in high-income countries.30,31 However, studies of interventions targeting MSM in low- and middle-income settings have consistently demonstrated both the need for and effectiveness of these programs.32-35 Although prevention strategies targeting MSM have been shown to be effective across country income levels, the benefit of these interventions has been subject to decay over time, indicating that programs should be ongoing to preserve increased condom usage.
Effective arguments for improved HIV programming for MSM can be made both from public health and human rights perspectives. The data presented here make a clear case that MSM exist and are at risk for HIV infection throughout the world. Moreover, ignoring and stigmatizing high-risk population groups has never proven to be an effective tool in curbing HIV epidemics. From a human rights perspective, discrimination on the basis of sex, which includes sexual orientation, is prohibited by the International Covenant on Civil and Political Rights signed in 1966, of which most states, including all African countries, are signatories. In 1994, the United Nations held that sexual orientation was a status protected under this covenant from discrimination, with "sex" including "sexual orientation." Whether one gives more weight to the public health or the human rights argument, the conclusion is the same: It is time to comprehensively address the AIDS pandemic, and to do so effectively, all vulnerable populations -- including MSM -- should be included in HIV prevention programs.
Stefan Baral, M.D., M.P.H, and Frangiscos Sifakis, Ph.D., are at the Johns Hopkins Bloomberg School of Public Health; Farley Cleghorn, M.D., M.P.H., is at the Center for Health Futures at Constella Futures; Chris Beyrer, M.D., M.P.H., is at the Johns Hopkins Fogarty AIDS International Training and Research Program.
This article was provided by amfAR, The Foundation for AIDS Research. Visit amfAR's website to find out more about their activities and publications.