SUMMARY
Background:
Limited assessments of structural-level factors of HIV affecting gay men and other men who have sex with men (MSM) have been conducted, especially in sub–Saharan Africa. Our objective was to examine HIV testing history, HIV status, and stigmas among MSM living in ten different countries with heterogenous legal environments.
Methods:
This study used pooled data from ten country-specific, cross-sectional studies done in 25 sites in Burkina Faso, Cameroon, Côte d’Ivoire, Gambia, Guinea-Bissau, Nigeria, Senegal, Eswatini, Rwanda, and Togo. MSM were recruited by respondent-driven sampling and were eligible if they met country-specific requirements for age, area of residence, and self-reported being assigned male sex at birth and having anal sex with a man in the past 12 months. Policy related to same-sex sexual behaviour for each country was categorised as not criminalised or criminalised. Countries were also categorised on the basis of recent reports of prosecutions related to same-sex sexual acts. Legal barriers were defined as those that legally prevented registration or operation of sexual orientation related civil society organisations (CSOs). Individual data on HIV testing history, HIV status, and stigma were collected via interviewer administered sociobehavioural questionnaires and HIV testing. Multilevel logistic regression with random intercepts was used to assess the association between policies, recent prosecutions, legal barriers to CSOs, and HIV-related factors with adjusted odds ratios (aORs) and 95% CIs.
Findings:
Between Aug 3, 2011, and May 27, 2020, we recruited 8047 MSM with a median age of 23 years (IQR 21–27). 4886 (60·7%) lived in countries that criminalise same-sex sexual acts. HIV prevalence among MSM was higher in criminalised settings than non-criminalised settings (aOR 5·15, 95% CI 1·12–23·57); higher in settings with recent prosecutions than in settings without prosecutions (12·06, 7·19–20·25); and higher in settings with barriers to CSOs than without barriers to CSOs (9·83, 2·00–48·30). HIV testing or status awareness was not associated with punitive policies or practices. Stigma was associated with HIV status but did not consistently vary across legal environments. Disparities in HIV prevalence between MSM and other adult men were highest in punitive settings.
Interpretation:
Structural risks including discriminatory country-level policies, prosecutions, and legal barriers may contribute to higher HIV prevalence among men who have sex with men. Taken together, these data highlight the importance of decriminalization and decreasing enforcement, alongside stigma reduction, as central to effective control for HIV.
INTRODUCTION
Globally, gay men and other men who have sex with men (MSM) are disproportionately affected by HIV with recent estimates suggesting MSM are about 25 times more likely to be living with HIV than other adult men.1 Even in countries across sub-Saharan Africa where the HIV epidemics are more generalized, estimates are as high as nearly one in two MSM to be living with HIV.2 For MSM, HIV risks are shaped by individual, network, social, and structural factors that include stigma and discrimination related to sexual behavior, orientation, or preferences. Stigma is a social process by which an individual or group is labeled and devalued based on a perceived characteristic, resulting in adverse experiences, limited opportunities, and suboptimal wellbeing in the context of unequal power.3 Stigmas can arise from limited understanding, tolerance, and acceptance of diverse identities, behaviors, or health conditions.4 Stigmas may manifest at the individual, community, and structural- or macro-levels, and through prejudice, discounting, discrediting, and discrimination.5
The role of individual and community-level stigmas as barriers to HIV outcomes has been well established,5,6 though assessments of structural stigmas and discrimination and their relationship to individual-level health outcomes have been limited. This is likely on account of the complexity of how structural stigmas manifest and the limited individual-level data from multi-country studies available to properly assess these factors alongside HIV outcomes.7 Structural stigma is a macro-social form of stigma acting at the level of policies, systems, and structures.5 At the structural or macro-levels, laws serve multiple roles. Laws may reflect the dominant norms of a given society and seek to shape a population’s behavior to meet those norms, but it also has an “expressive” function in which law shapes social norms.8 As such, laws and policies can be both a means through which social stigma is created or enforced and a means of remedying or preventing stigma and blunting its harms.9
Anti-Lesbian Gay Bisexual Trans (LGBT) legislation and enforcement functions at the structural-level to discriminate against individuals based on sexual orientation, behaviors, or preferences. An online survey conducted among MSM from 38 European countries found that in countries with more stigmatizing laws and policies, MSM were less likely to report same-sex attractions, sexual behaviors, and sexual identities than countries with less stigmatizing laws and policies.10 Non-disclosure of sexual behaviors has been shown to negatively impact HIV prevention, diagnosis, and treatment, and disclosure of sexual behaviors to healthcare providers is essential to informing appropriate care.11 A recent systematic review and meta-analysis of studies conducted in sub-Saharan Africa observed that levels of testing history and awareness of HIV status were lower in countries with the most severe legislation compared to countries with the least severe legislation.12 A cross-national global ecological analysis showed that countries with LGBT criminalizing laws and policies had lower levels of knowledge of HIV status and lower HIV viral suppression among people living with HIV.13 Despite these existing findings, the current evidence is limited to aggregate-level analysis as none of the existing studies have directly assessed the association between punitive legal environments and HIV outcomes utilizing individual-level data, especially in sub-Saharan Africa.
In settings with punitive policies like the criminalization of same-sex sexual acts, civil society organizations (CSO) may serve a protective function by enacting health services and advocacy to shape policies, funding, and programs for devalued communities such as MSM.14 For instance, the involvement of CSOs and communities of people living with HIV were central in scaling up the HIV response globally.15 However, many countries have legal barriers to the operation and registration of CSOs to support sexual and gender diverse efforts, impeding access to care and services among individuals at risk for and living with HIV.16
Recognizing stigma as a barrier to HIV prevention, diagnosis, and treatment, the Joint United Nations Programme on HIV/AIDS has established the goal of achieving zero discrimination by 2025 to support eliminating HIV as a public health threat by 2030. UNAIDS has also established the 10-10-10 targets which focus on removing social and legal impediments to support access and utilization of HIV services.17 However, there remains a critical need to characterize the mechanisms by which structural stigma and discrimination harm the HIV response in order to inform intervention strategies. In response, this study assessed the associations between discriminatory laws, prosecutions, and legal barriers with HIV prevalence and testing history among MSM across 10 countries in sub-Saharan Africa. We also aim to assess the association between stigmas and HIV and explore if these relationships vary based on the legal environments. Lastly, we assessed the association between punitive environments and HIV prevalence disparities between MSM and other adult men.
METHODS
Data collection and participants
This study used pooled data from ten country-specific, cross-sectional studies done in Burkina Faso, Cameroon, Côte d’Ivoire, Eswatini, The Gambia, Guinea-Bissau, Nigeria, Rwanda, Senegal, and Togo. Study teams led primary data collection in the ten countries. All data were cross-sectional and individuals were recruited and enrolled over the specified time periods for each study between 2011 and 2020. All recruitment was conducted via respondent driven sampling (RDS) independently across 25 sites within the ten countries.18 Recruitment chains were initiated by seeds in each site, who were individuals selected in collaboration with local community-based organizations to represent heterogeneity in demographic characteristics and geographic representation.
Participants were eligible if they met country-specific requirements for age, area of residence, and self-reported being assigned male sex at birth and having anal sex with a man in the last 12 months. All participants provided verbal or written informed consent depending on the approach determined for each respective country. Secondary data analysis is overseen by JHSPH IRB approval IRB00007006 and data collection was approved by country-specific ethical committees (appendix p1).
Study teams conducted interviewer-administered questionnaires in a private location with trained study staff. Testing for HIV, including pre- and post-test counseling, was conducted following country-specific national guidelines. HIV testing was conducted prior to administering the socio-behavioral questionnaires. Study staff led post-test counseling and review of HIV test results with participants after completion of the socio-behavioral questionnaire.
Measures
We used individual-level data from socio-behavioral questionnaires and HIV testing for these analyses. The primary dependent variable was HIV status based on HIV test results at the time of RDS participation. Additional HIV measures included self-reported history of ever having received an HIV test and awareness of HIV positive status among those living with HIV.
Individual-level stigmas were explored as exposures of interest. A total of 13 stigma items were administered consistently across countries. These items were combined into separate stigma scales for this analysis based on the results of an exploratory factor analysis and published previously (appendix p2).19 Stigma categories included stigma related to family and friends; anticipated/perceived healthcare stigma; and general social stigma which are self-reported experiences of stigma attributable to having sex with men.
Other individual-level measures explored as potential confounders include age, education, sexual orientation, marital status, disclosure of same-sex sexual relationships to family or friends, and disclosure of same-sex sexual relationships to a healthcare provider.
Legal environment measures included country-level same-sex sexual practice-related policies, which were defined and categorized based on work developed by the International Lesbian, Gay, Bisexual, Trans and Intersex Association (ILGA)20 and the HIV Policy Lab21. Criminalization of same-sex sexual practices was categorized as not criminalized where national law avoids criminalizing consensual same-sex sex acts; and criminalized in countries where national law criminalizes consensual same-sex sex acts. Prosecutions for consensual same-sex sexual acts was applied to countries where there were reports of prosecutions for consensual same-sex sexual acts within the last year. Categories were determined based on the year of data collection for each respective country. Lastly, legal barriers to the registration or operation CSOs was defined based on ILGA confirmed reports. Full definitions and sources are outlined in appendix p2–3.
Country-level measures includes HIV epidemic setting, HIV population prevalence, antiretroviral therapy (ART) coverage, and HIV disparity. HIV epidemic setting was defined for each country as either generalized or concentrated, based on UNAIDS and WHO definitions. HIV epidemic setting, HIV population prevalence, and ART coverage for each country was based on UNAIDS estimates. HIV disparity score was calculated as the difference between the HIV prevalence among MSM in each country-specific study sample and HIV prevalence among adult men in each country (appendix p3).
Statistical Analyses
The sample size was determined separately for each country-specific data collection. The sample size for each country was calculated based on the ability to estimate the HIV population prevalence in each setting. We pooled data across countries and sites. We did not apply RDS-adjusted weighting as MSM did not represent a single network, violating a key assumption of RDS.22,23 While results represent valid sample estimates, they may differ from population-level estimates given lack of full RDS-adjustment.18 Proportions of demographic characteristics, HIV status, HIV testing history, knowledge of HIV positive status, disclosure, and stigma were described using crude estimates.
The dependent variables of interest are HIV status, HIV testing history, and awareness of HIV positive status. The primary exposures of interest include policy, prosecutions, and legal barriers to CSOs. Potential confounders include age, education, marital status, sexual orientation, epidemic setting, HIV population prevalence, site, year of data collection, and recruitment seed.
We used multilevel logistic regression with random intercepts to estimate odds ratios and 95% confidence intervals (CIs). Final models were adjusted for age, education, marital status, sexual orientation, HIV epidemic setting, ART coverage, HIV prevalence, year of data collection, recruitment seed, site, and clustered by country. We ran separate multivariable models for each policy/prosecution/legal barrier exposure and outcome of interest to avoid potential collinearity (appendix p4).
Secondary exposures of interest are stigma related to family and friends, anticipated/perceived healthcare stigma, and general social stigma. Multilevel logistic regression with random intercepts was used to estimate the odds ratios and 95%CIs between stigma exposures and HIV status. Stigma exposure models were run separately due to potential collinearity between stigma scales. Final models were adjusted for age, education, marital status, sexual orientation, epidemic, HIV prevalence, year of data collection, seed, site, and respective disclosure variables when conceptually relevant. All models were clustered by country. Stigma models were not adjusted for the legal environment, as these measures were conceptualized as modifiers and not confounders (appendix p4).
To explore potential effect measure modification in which the relationship between stigma scales and HIV status might vary depending on the existence of particular policies, the relationship between stigma scale and HIV status was stratified by policies, prosecutions, and legal barriers to CSOs on same-sex sexual relationships. The Mantel-Haenszel test of homogeneity (MH) was used to assess differences between stigma and HIV across different legal barriers, using a significance level of p<0.05.
HIV prevalence disparity was assessed as a dependent variable with policy, prosecutions, and legal barriers to CSOs and exposures of interest. Multilevel mixed-effects generalized linear models were used to assess the exposures and HIV prevalence disparity outcome separately, and models adjusted for country, site, ART coverage, and HIV prevalence.
Statistical analyses were done using STATA 15.1 (College Station, Texas).
Role of funding source
The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
RESULTS
Data were collected between Aug 3, 2011, and May 27, 2020. 8047 MSM are represented in these analyses: 672 (8·4%) in Burkina Faso; 1323 (16·4%) in Cameroon; 1301 (16·2%) in Côte d’Ivoire; 326 (4·1%) in Eswatini; 114 (1·4%) in The Gambia; 451 (5·6%) in Guinea-Bissau; 1716 (21·3%) in Nigeria; 737 (9·2%) in Rwanda; 724 (9·0%) in Senegal; and 683 (8·5%) in Togo (appendix 2 p 1). Among participants, 3161 (39·3%) lived in countries where same-sex sexual acts were not criminalised; 4886 (60·7%) lived in countries where same-sex sexual acts were criminalised; 3877 (48·2%) lived in countries with recent prosecutions related to same-sex sexual acts; and 4761 (59·2%) lived in countries with legal barriers to CSOs. The median age of participants was 23 years (IQR 21–27; Table 1).
Table 1.
Demographic characteristics among men who have sex with men across ten countries in sub-Saharan Africa, stratified by policy related to same-sex sexual acts
| Total (n=8047) | Policy related to same-sex sexual relationships | Total (n=8047) | |
|---|---|---|---|
| Not criminalised (n=3161) | Criminalised (n=4886) | ||
| Age, years | |||
| <24 | 4941/7935 (62·3%) | 1993/3161 (63·0%) | 2948/4774 (61·8%) |
| 25–30 | 2163/7935 (27·3%) | 820/3161 (25·9%) | 1343/4774 (28·1%) |
| ≥31 | 831/7935 (10·5%) | 348/3161 (11·0%) | 483/4774 (10·1%) |
| Education level | |||
| None | 342/7968 (4·3%) | 137/3152 (4·3%) | 205/4816 (4·3%) |
| Some primary | 589/7968 (7·4%) | 203/3152 (6·4%) | 386/4816 (8·0%) |
| Primary completed or some secondary | 3324/7968 (41·7%) | 1772/3152 (56·2%) | 1552/4816 (32·2%) |
| Completed secondary or post-secondary | 3713/7968 (46·6%) | 1040/3152 (33·0%) | 2673/4816 (55·5%) |
| Sexual orientation | |||
| Gay or homosexual | 3871/8007 (48·3%) | 1733/3147 (55·1%) | 2138/4860 (44·0%) |
| Bisexual | 3994/8007 (49·9%) | 1313/3147 (41·7%) | 2681/4860 (55·2%) |
| Heterosexual | 133/8007 (1·7%) | 98/3147 (3·1%) | 35/4860 (0·7%) |
| Other | 9/8007 (0·1%) | 3/3147 (0·1%) | 6/4860 (0·1%) |
| Disclosure of sexual minority status * | |||
| Disclosure of sexual minority status to family | 1784/8033 (22·2%) | 758/3161 (24·0%) | 1026/4872 (21·1%) |
| Disclosure of sexual minority status to health-care providers | 2113/7333 (28·8%) | 548/2487 (22·0%) | 1565/4846 (32·3%) |
| Sexual minority stigmas * | |||
| Stigma from family and friends | 2422/8041 (30·1%) | 1193/3161 (37·7%) | 1229/4880 (25·2%) |
| Anticipated health-care stigma | 2064/8038 (25·7%) | 791/3161 (25·0%) | 1273/4877 (26·1%) |
| General social stigma | 4573/8046 (56·8%) | 1454/3161 (46·0%) | 3119/4885 (63·9%) |
| HIV * | |||
| History of HIV testing among all participants | 5895/8028 (73·4%) | 2379/3156 (75·4%) | 3516/4872 (72·2%) |
| Living with HIV | 1581/7958 (19·9%) | 264/3116 (8·5%) | 1317/4842 (27·2%) |
| Ever told of HIV positive status among those living with HIV | 629/1581 (39·8%) | 102/264 (38·8%) | 527/1317 (40·0%) |
| Currently on antiretroviral among those living with HIV | 333/1581 (21·1%) | 49/264 (18·6%) | 284/1317 (21·6%) |
Items reporting when value = yes and the values for reporting = no, it is not shown
1581 (19·9%) of 7958 participants with available data were living with HIV (Table 1). In countries without criminalisation, 264 (8·5%) of 3116 participants had HIV, and in countries with criminalisation, 1317 (27·2%) of 4842 participants had HIV (Table 1). HIV prevalence among study participants in criminalised settings was higher than among participants in non-criminalised settings (Table 2). HIV prevalence among study participants in countries with recent prosecutions was higher than in settings without recent prosecutions (Table 2). HIV prevalence among study participants living in countries with barriers to CSOs was higher than in settings without barriers (Table 2).
Table 2:
Policy, prosecutions, and legal barriers to CSOs related to same-sex sexual acts and HIV status among men who have sex with men in 10 countries in sub-Saharan Africa
| Polices and Legal Barriers | Living with HIV | |||||
|---|---|---|---|---|---|---|
| Total | Yes | |||||
| % | N | % | n/N | OR (95% CI) | aOR* (95%CI) | |
| Policy related to consensual same-sex sexual acts | ||||||
| Not criminalized | 39.3 | 3161 | 8.4 | 263/3315 | Ref | Ref |
| Criminalized | 60.7 | 4886 | 27.2 | 1317/4842 | 4.49 (2.05, 9.82) | 5.15 (1.12, 23.57) |
| Prosecutions for consensual same-sex sexual acts | ||||||
| No | 51.8 | 4170 | 9.3 | 385/4120 | Ref | Ref |
| Yes | 48.2 | 3877 | 31.1 | 1195/3837 | 4.92 (2.53, 9.52) | 12.06 (7.19, 20.25) |
| Legal barriers to civil society organizations | ||||||
| No | 40.8 | 3286 | 10.4 | 338/3243 | Ref | Ref |
| Yes | 59.2 | 4761 | 26.3 | 1242/4715 | 2.01 (0.66, 6.10) | 9.83 (2.00, 48.30) |
Adjusted for age, education, sexual orientation, marital status, year of data collection, country level population HIV prevalence among adults 15–49, epidemic setting, country-level ART coverage, recruitment seed, site, and clustered by country.
5895 (73·4%) of 8028 participants with available data reported ever having an HIV test (Table 1). Among 1581 participants living with HIV, 629 (39·8%) reported being aware of their HIV positive status (Table 1). Ever testing for HIV was not associated with policy, prosecutions, or barriers to CSOs (Table 3). Awareness of HIV positive status was associated with CSO barriers (Table 3).
Table 3.
Policy, prosecutions, and legal barriers to civil society organisations related to same-sex sexual acts and HIV testing history among men who have sex with men
| Total (8047) | Has ever tested for HIV among all study participants (n=5895)* | Has awareness of HIV positive status among those living with HIV (n=629)† | |||||
|---|---|---|---|---|---|---|---|
| n/N (%) | OR (95% CI) | aOR‡ (95% CI) | n/N (%) | OR (95% CI) | aOR‡ (95% CI) | ||
| Policy related to consensual same-sex sexual acts | |||||||
| Not criminalised | 3161 (39·3%) | 2379/3156 (75·4%) | 1 (ref) | 1 (ref) | 102/263 (38·8%) | 1 (ref) | 1 (ref) |
| Criminalised | 4886 (60·7%) | 3516/4872 (72·2%) | 0·70 (0·27–1·78) | 0·76 (0·25–2·26) | 527/1317 (40·0%) | 0·65 (0·19–2·21) | 0·78 (0·19–3·22) |
| Prosecutions for consensual same-sex sexual acts | |||||||
| No | 4170 (51·8%) | 3034/4160 (72·9%) | 1 (ref) | 1 (ref) | 127/285 (33·0%) | 1 (ref) | 1 (ref) |
| Yes | 3877 (48·2%) | 2861/3868 (74·0%) | 0·91 (0·35–2·38) | 1·03 (0·34–3·18) | 502/1195 (42·0%) | 0·88 (0·25–3·03) | 1·72 (0·43–6·79) |
| Legal barriers to civil society organisations | |||||||
| No | 3286 (40·8%) | 2408/3278 (73·5%) | 1 (ref) | 1 (ref) | 106/338 (31·4%) | 1 (ref) | 1 (ref) |
| Yes | 4761 (59·2%) | 3487/4750 (73·4%) | 1·08 (0·42–2·77) | 1·66 (0·52–5·40) | 523/1242 (42·1%) | 1·61 (0·48–5·50) | 4·02 (1·27–12·76) |
Adjusted for age, education, sexual orientation, marital status, year of data collection, country level population prevalence among adults 15–49, recruitment seed, site, and clustered by country.
2422 (30·1%) of 8041 participants reported stigma from family or friends, 2064 (25·7%) reported anticipated or perceived health-care stigma, and 4573 (56·8%) reported general social stigma. HIV was associated with stigma from family or friends, anticipated or perceived health-care stigma, and general social stigma (Table 4). The association between social and family stigma and HIV varied by existence of legal barriers to CSOs, and the association between anticipated or perceived health-care stigma and HIV varied by criminalisation status and recent prosecutions (Table 4).
Table 4:
Stigmas and HIV status among gay men and other men who have sex with men in 10 countries in sub-Saharan Africa
| Stigma | Living with HIV | ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Total | Yes | Total sample | Policy related to same-sex sex acts | Prosecutions for same-sex sexual acts | Legal barriers to civil society organizations | ||||||||||
| Not criminalized | Criminalized | No | Yes | No | Yes | ||||||||||
| % | n | % | n/N | OR (95% CI) | aOR* (95%CI) | MH p value* | aOR** (95%CI) | aOR** (95%CI) | MH p value* | aOR** (95%CI) | aOR** (95%CI) | MH p value* | aOR** (95%CI) | aOR** (95%CI) | |
| Stigma from family or friends | 30.1 | 2422 | 22.8 | 543/1580 | 1.66 (1.45, 1.89) | 1.37 (1.17, 1.60) | 0.42 | - | - | 0.53 | - | - | 0.0058 | 1.54 (1.20, 1.97) | 1.29 (1.07, 1.56) |
| Anticipated/perceived healthcare stigma | 25.6 | 2064 | 26.6 | 539/1579 | 1.52 (1.33, 1.73) | 1.49 (1.27, 1.74) | 0.0086 | 1.22 (0.91, 1.64) | 1.58 (1.33, 1.88) | 0.0015 | 1.19 (0.92, 1.54) | 1.69 (1.41, 2.03) | 0.46 | - | - |
| General social stigma | 56.8 | 4573 | 21.5 | 747/1580 | 1.91 (1.66, 2.20) | 1.49 (1.26, 1.78) | 0.19 | - | - | 0.087 | - | - | 0.073 | - | - |
Adjusted for age, year, orientation, education, marital status, disclosure of sexual behavior to relevant group^, year of data collection, country HIV prevalence, recruitment seed, site, and clustered by country.
Stigma from family or friends included disclosure to family or friends. Anticipated/perceived healthcare stigma included disclosure to healthcare provider. General social stigma included disclosure to healthcare provider.
The Mantel-Haenszel test of homogeneity (MH) was used to assess differences between stigma and HIV across different policy, enforcement, and legal barriers contexts, using a significance level of p<0.05. If the MH p value was not <0.01 then it did not suggest values differed by setting.
HIV prevalence difference between study participants and adult men in each country was higher in criminalised settings than non-criminalised settings (Table 5). HIV prevalence difference was higher in settings with recent prosecutions than settings without recent prosecutions and higher in settings with legal barriers to CSOs than settings without legal barriers (Table 5).
Table 5:
HIV prevalence disparity between gay men and other men who have sex with men and other adult men, and associations with legal environments across 10 countries in sub-Saharan Africa
| Polices and Legal Barriers | HIV prevalence disparity | ||
|---|---|---|---|
| HIV prevalence disparity score* | HIV prevalence ratio (PR) comparing HIV prevalence disparity score between policy/legal settings | ||
| Mean PD (sd) | PR** (95%CI) | aPR*** (95%CI) | |
| Policy related to consensual same-sex sexual acts | |||
| Not criminalized | 7.22 (3.57) | Ref | Ref |
| Criminalized | 24.82 (13.06) | 17.59 (17.13, 18.06) | 13.08 (12.73, 13.43) |
| Prosecutions for consensual same-sex sexual acts | |||
| No | 6.58 (4.07) | Ref | Ref |
| Yes | 30.09 (8.63) | 23.50 (23.21, 23.79) | 18.24 (18.08, 18.40) |
| Legal barriers to civil society organizations | |||
| No | 8.89 (5.69) | Ref | Ref |
| Yes | 24.13 (13.82) | 15.25 (14.74 15.74) | 16.58 (16.21, 16.95) |
HIV prevalence disparity score is defined as the prevalence difference (PD) between MSM and other adult men in each respective country.
Generalized linear model was used to assess the legal environment exposures and outcome of interest (HIV prevalence disparity score).
Adjusted for country, site, country level ART coverage, and country level population prevalence among adults 15–49.
DISCUSSION
We explored the relationships between legal environments, types of stigma, HIV testing, HIV status, and awareness of HIV status among MSM living in ten countries across sub-Saharan Africa. Punitive policies against same-sex sexual acts as well as recent prosecutions related to same-sex sexual acts were associated with prevalent HIV among MSM. Extending beyond the criminalization of same-sex sexual orientation, behaviors, and preferences – criminalization of CSOs supporting MSM was also associated with higher HIV prevalence among MSM in this study. Stigma from family and friends, within healthcare, and community were associated with HIV status among MSM but did not consistently change based on the legal environment, highlighting the potentially complex drivers and facilitators influencing stigmas affecting MSM. Lastly, HIV prevalence difference between MSM in this study and other adult men in each respective country was associated with criminalization, recent prosecutions, and legal barriers to CSOs.
Awareness of HIV positive status among MSM living with HIV in this study sample was low, and far from UNAIDS targets for HIV control. Although a high proportion of study participants reported having ever received an HIV test, awareness of HIV status study suggests that study participants may not be accessing HIV tests regularly. For people without HIV, HIV testing services provide opportunities to access prevention services allowing them to remain HIV free; and for people living with HIV, HIV testing facilitates access to ART, achievement of viral suppression, and reduction of onward transmission. Although HIV testing history was not associated with policy, prosecutions, or legal barriers to CSOs, this study suggests regular, routine testing and communication of positive results are needed across legal environments among MSM.
MSM living in settings where prosecutions had been recently documented had 12 times the odds of living with HIV than individuals in settings without recent prosecutions. Additionally, MSM living in criminalized settings had five times the odds of living with HIV compared to men living in settings without criminalization. These results support ecological analyses highlighting differential HIV measures based on criminalization status,13 and reinforce the harmful role of a punitive context on HIV.12 These results suggests that the existence of laws, even if the laws are not being actively enforced, may contribute to HIV risk among MSM – and that the burden may be greatest in settings with recent prosecutions. Therefore, while decreasing enforcement may prove helpful, removing laws and decriminalization may still be important in supporting HIV programming among MSM in sub-Saharan Africa.
This study highlights that CSOs likely play an important role in supporting health services and advocacy efforts for MSM, across criminalized settings.14 MSM in this study were more likely to be living with HIV in settings with legal barriers to CSO operation and registration. In settings with criminalization, access to legal and health services through CSOs remains an avenue to mitigate barriers to health and human rights.24 Senegal is an example where CSOs have made progress in achieving access to HIV services for MSM through advocacy efforts and direct care provision.25 Although same-sex sexual acts and marriage in Senegal remain criminalized, the ministry of health has prioritized efforts toward HIV services for MSM. In South Africa, the early HIV response among MSM was driven by existing social support networks and health services for MSM.26 Generally, CSOs are well positioned to achieve progress in rights-constrained environments.25 However, legal barriers to CSOs in the context of punitive policies against same-sex sexual acts provide little opportunities for advancements in HIV and human rights. In 2014, Nigeria increased punitive policies to further criminalize same-sex sexual acts, including prohibiting participation in organizations, service provision, or meetings that support gay people, and punishes attempts to enter civil unions or publicly show same-sex romantic relationships. This law resulted in an immediate effect on fear and avoidance of seeking health services.27
Stigma from family or friends, healthcare stigma, and general social stigma were associated with increased HIV prevalence among MSM in this study. The relationship between healthcare stigma and HIV status was higher in settings with recent prosecutions. However, the relationships between general social stigma with HIV did not differ by criminalization status, recent prosecutions, or legal barriers to CSOs. This suggests that stigma experienced by individuals within these domains and its influence on HIV remains a major factor regardless of the legal environment. It also suggests the need to address stigma that occurs at across levels such as intrapersonal-, social-, community-, organizational-, and structural-levels. This differs from studies assessing stigma and HIV across settings with criminalization policies of sex work, which suggest that state actors may be playing a large role in these dynamics.28 It may be that stigma affecting MSM is driven or facilitated by social context, cultural beliefs and norms more than legal frameworks. For example, despite the establishment of legal protections based on sexual orientation in South Africa in 1994, MSM is South Africa are not less likely to experience human rights abuses when compared to MSM in criminalized settings.29 This complex dynamic was observed in the United States as well, where although legal protections for MSM exist, the prevalence of stigma is high and not different from MSM across sub-Saharan African countries.30 Given this, interventions spanning across cultural, gender, religious, and legal spaces are likely needed to improve human rights among MSM. However, social acceptance should not be considered a prerequisite for decriminalization, rather, decriminalization can precede and coincide with community advocacy for improved social acceptance.
Lastly, the difference between HIV prevalence among MSM in this study and adult men in each respective country can serve as a measure of disparity between MSM and non-MSM populations. Punitive policy, recent prosecutions, and legal barriers to CSOs were all associated with higher HIV prevalence disparities in this study. Structural factors such as criminalization, active prosecutions and enforcement, and barriers to CSO may contribute to disparity in prevention, risk, and ultimately progress of achieving HIV control.
Limitations in this study should be considered. Importantly, all data are cross-sectional and cannot assess causality. RDS is a non-probabilistic-based sampling approach and may introduce selection bias. We adjusted for recruited seed in our analyses but did not apply RDS-weighted adjustments given the limited consensus on this approach for complex, multi-level models. Self-reported measures may be subject to recall bias and social desirability bias which may differ across settings. Measurement variance in the stigma measures across countries exists which may represent potential differential error in measurement by country. Data were collected over a period of nine years, and although we have adjusted for time in which data were collected to account for secular trends, this may not capture all the different ways in which the experiences of MSM and control of HIV may have changed over this period. Enforcement practices beyond arrests, program funding, and other external measures over time may have influenced stigma, HIV status, or HIV risk. It is possible that countries which take a punitive approach towards MSM may also take this approach towards other populations and overall may utilize less public health informed approaches for HIV control. This may result in latent unmeasured confounding which may influence the specific relationships observed in this study. Importantly, ART coverage reduces mortality among people living with HIV and can contribute to a higher prevalence. Therefore, higher HIV prevalence may not always represent poor HIV control. To account for this, we have adjusting for ART coverage and HIV prevalence in our analyses. It is possible that unmeasured confounders preceding both same-sex policies and country-level HIV epidemics may exist and feed independently into both, thus resulting in residual confounding. We cannot rule out the possibility of uncontrolled confounding, particularly as there may be unmeasured confounders which are associated with sexual diverse laws and/or stigma, as well as causally associated with HIV and not in the casual pathway between exposure and outcome. None of the countries in this study met the criteria for decriminalized status, or had protective policies in place, and therefore these legal contexts could not be assessed. Data collection for Eswatini, Burkina Faso, Togo categorization took place in years that are not captured in the HIV Policy Lab Database. ILGA State-Sponsored Homophobia Reports and online media were reviewed for recent prosecutions and did not report any prosecutions in these countries during these years. However, the process for identifying arrests was different from those identified in the HIV Policy Lab database and may be subject to misclassification. MSM living with HIV may experience intersectional stigma due to HIV status and sexual or gender diversity. However, this study did not assess the potential intersections of stigmas attributable to these characteristics. In the development of the disparity score, HIV estimates for adult men were compared to our study sample, which may not account for demographic characteristic of these populations. Importantly, most of the countries included in these analyses are in west and central Africa, with the exceptions of Rwanda and eSwatini. Therefore, these results may not be generalizable across all countries in SSA.
Across sub-Saharan Africa, punitive policies, enforcement practices, and legal barriers to CSOs were associated with a higher HIV prevalence among MSM in this study. These results provide individual empiric data demonstrating how structural risks and discriminatory policies may contribute to HIV prevalence among MSM across sub-Saharan Africa. This study highlights the potential impact of decreasing enforcement or decriminalization to optimize HIV prevention and treatment efforts. However, this study also highlighted the complexity of stigma affecting MSM– suggesting policy reform alone will not eliminate stigma but interventions addressing social and cultural drivers may support comprehensive stigma reduction.
Supplementary Material
Supplementary Table 1: Summary of data collection across 10 countries in sub-Saharan Africa
Supplementary Table 2: Country specific ethics approvals and inclusion criteria
Supplementary Table 3: Stigma Measures
Supplementary Table 4: Legal environment categories and definitions
Supplementary Table 5: Legal environment by country
Supplementary Table 6: HIV setting characteristics by country
Supplementary Table 7: Correlation of legal environment
Supplementary Table 8: History of HIV testing, HIV status, and awareness of HIV status among MSM by country
Supplementary Figure 1. Framework for legal environment, stigma, and HIV
RESEARCH IN CONTEXT.
Evidence before this study
Globally, gay men and other men who have sex with men (MSM) are disproportionately affected by HIV. The role of individual and community-level stigmas as barriers to HIV prevention and diagnosis has been well established and identified as key barriers to HIV control. However, the assessment of structural stigmas and discriminatory policies on individual-level health outcomes affecting MSM have been limited. We searched PubMed with the terms ““HIV” AND (“Sexual and Gender Minorities” OR “Homosexuality, Male”) AND (“Criminal Law” OR “Law Enforcement” OR “Policy” OR “Legislation as Topic” OR “Social Stigma”) with no language restrictions, for publications up to December 31, 2021. Anti-Lesbian Gay Bisexual Trans (LGBT) legislation have been associated with non-disclosure of same-sex attractions, sexual behaviors, and sexual identities among MSM. A recent systematic review and meta-analysis of studies conducted in sub-Saharan Africa observed that levels of testing history and awareness of HIV status were lower in countries with the most severe legal penalties related to same sex relations compared to countries with the least severe legal penalties. A cross-national global ecological analysis observed that countries with LGBT criminalizing laws and policies had lower aggregate levels of knowledge of HIV status and lower HIV viral suppression among people living with HIV. Despite these existing findings, the current evidence assessing criminalization of same-sex sexual behaviors affecting HIV outcomes among MSM is limited to aggregate-level analyses and is subject to ecological fallacies as none of the existing studies have directly assessed the association between punitive policies and HIV outcomes utilizing individual-level data.
Added value of this study
This study utilizes individual-level socio-behavioral and biological data from 8,049 MSM from across 10 countries in sub-Saharan Africa including Burkina Faso, Cameroon, Côte d’Ivoire, eSwatini, The Gambia, Guinea-Bissau, Nigeria, Rwanda, Senegal, and Togo to understand the relationship between country-level policies, prosecutions, and legal barriers with HIV testing, prevalence, and status awareness among MSM. This study demonstrates that MSM living in countries which criminalize and enforce discriminatory policies around same-sex relationships have the highest burden of HIV. MSM living in countries which have recent prosecutions related to same-sex sexual acts still have an elevated HIV prevalence compared to countries without recent prosecutions. Furthermore, the presence of legal barriers to CSOs registration and operation was associated with higher levels of HIV.
Implications of all the available evidence
This study contributes to the existing evidence on understanding the role of harmful policies, such as criminalization. Individual-level data from countries across sub-Saharan Africa demonstrated that structural risks including discriminatory country-level policies, recent prosecutions, and legal barriers may contribute to higher HIV prevalence among MSM. These results provide evidence that decriminalization of same sex relationships, as well as decreasing enforcement of criminalizing policies are central to effective HIV control.
Acknowledgements
We express our sincere appreciation to the participants of this study. In addition, we acknowledge the crucial role of the community groups that make great personal and professional sacrifices to serve the unmet health and advocacy needs of those most marginalized in the HIV response. We would also like to thank the data collection and study coordination teams across the different countries. The analysis and manuscript were made possible through effort funded through NIMH and NIAID, including Johns Hopkins HIV Epidemiology and Prevention Sciences Training Program (5T32AI102623-08); National Institute Of Mental Health of the National Institutes of Health under Award Number F31MH128079, R01MH110358, R01AI170249; and National Cancer Institute under award K07CA225403. This publication was made possible by the Johns Hopkins University Center for AIDS Research, an NIH funded program (P30AI094189). Data from the HIV Policy Lab is supported by a grant from USAID to Georgetown University and UNAIDS. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Original data collection for each individual study was funded through USAID, PEPFAR, and Global Fund to Fight AIDS, Tuberculosis and Malaria.
FUNDING STATEMENT:
National Institutes of Health
Footnotes
Declaration of interests
CL reports funding from NIMH and NIAID to institution. SM reports funding from NIMH to institution. SB reports funding from NIMH to institution. RGN reports funding from NCI to institution. TAC reports funding from US Army to institution. MK reports funding from USAID and UNAIDS to institution and serves as a Special Advisor to the United Nations Joint Programme on HIV/AIDS. GM reports that institution, amfAR, is completing a 3-year Viiv grant examining data reporting for key populations (inclusive of MSM). All other authors declare no competing interests.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Data availability statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.
References
- 1.Confronting Inequalities: Lessons for pandemic responses from 40 years of AIDS UNAIDS, 2021.
- 2.Keshinro B, Crowell TA, Nowak RG, et al. High prevalence of HIV, chlamydia and gonorrhoea among men who have sex with men and transgender women attending trusted community centres in Abuja and Lagos, Nigeria. Journal of the International AIDS Society 2016; 19(1): 21270. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Link BG, Phelan JC. Conceptualizing Stigma. Annual Review of Sociology 2001; 27(1): 363–85. [Google Scholar]
- 4.Goffman E Stigma: notes on the management of spoiled identity. Englewood Cliffs, NJ: Prentice Hall; 1963. [Google Scholar]
- 5.Stangl AL, Earnshaw VA, Logie CH, et al. The Health Stigma and Discrimination Framework: a global, crosscutting framework to inform research, intervention development, and policy on health-related stigmas. BMC medicine 2019; 17(1): 31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Rodriguez-Hart C, Musci R, Nowak RG, et al. Sexual Stigma Patterns Among Nigerian Men Who Have Sex with Men and Their Link to HIV and Sexually Transmitted Infection Prevalence. AIDS and behavior 2018; 22(5): 1662–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Link BG, Yang LH, Phelan JC, Collins PY. Measuring mental illness stigma. Schizophrenia bulletin 2004; 30(3): 511–41. [DOI] [PubMed] [Google Scholar]
- 8.Sunstein C On the expressive function of law. University of Pennsylvania Law Review 1996; 144(5): 2021–53. [Google Scholar]
- 9.Burris S Stigma and the law. Lancet (London, England) 2006; 367(9509): 529–31. [DOI] [PubMed] [Google Scholar]
- 10.Pachankis JE, Hatzenbuehler ML, Mirandola M, et al. The Geography of Sexual Orientation: Structural Stigma and Sexual Attraction, Behavior, and Identity Among Men Who Have Sex with Men Across 38 European Countries. Archives of sexual behavior 2017; 46(5): 1491–502. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Kokogho A, Amusu S, Baral SD, et al. Disclosure of Same-Sex Sexual Practices to Family and Healthcare Providers by Men Who Have Sex with Men and Transgender Women in Nigeria. Archives of sexual behavior 2021; 50(4): 1665–76. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Stannah J, Dale E, Elmes J, et al. HIV testing and engagement with the HIV treatment cascade among men who have sex with men in Africa: a systematic review and meta-analysis. The lancet HIV 2019; 6(11): e769–e87. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Kavanagh MM, Agbla SC, Joy M, et al. Law, criminalisation and HIV in the world: have countries that criminalise achieved more or less successful pandemic response? BMJ Glob Health 2021; 6(8). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Trapence G, Collins C, Avrett S, et al. From personal survival to public health: community leadership by men who have sex with men in the response to HIV. Lancet (London, England) 2012; 380(9839): 400–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Coutinho A, Roxo U, Epino H, Muganzi A, Dorward E, Pick B. The expanding role of civil society in the global HIV/AIDS response: what has the President’s Emergency Program For AIDS Relief’s role been? Journal of acquired immune deficiency syndromes (1999) 2012; 60 Suppl 3: S152–7. [DOI] [PubMed] [Google Scholar]
- 16.HIV Policy Lab. https://www.hivpolicylab.org/.
- 17.2025 Targets. 2022. https://aidstargets2025.unaids.org/ (accessed 24 August 2022).
- 18.Schonlau ML E Respondent-driven sampling. The Stata Journal 2012; 12: 72–93. [Google Scholar]
- 19.Augustinavicius JL, Baral SD, Murray SM, et al. Characterizing Cross-Culturally Relevant Metrics of Stigma Among Men Who Have Sex With Men Across 8 Sub-Saharan African Countries and the United States. Am J Epidemiol 2020; 189(7): 690–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.State Sponsored Homophobia 2014: A world survey of sexual orientation laws: criminalisation, protection and recognition. Geneva: International Lesbian, Gay, Bisexual, Trans and Intersex Association, 2014. [Google Scholar]
- 21.Kavanagh MM, Graeden E, Pillinger M, et al. Understanding and comparing HIV-related law and policy environments: cross-national data and accountability for the global AIDS response. BMJ Glob Health 2020; 5(9). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Salganik MJH, Douglas D. Sampling and Estimation in Hidden Populations Using Respondent-Driven Sampling. Sociological Methodology 2004; 34: 193–239. [Google Scholar]
- 23.Avery L, Rotondi N, McKnight C, Firestone M, Smylie J, Rotondi M. Unweighted regression models perform better than weighted regression techniques for respondent-driven sampling data: results from a simulation study. BMC Medical Research Methodology 2019; 19(1): 202. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Csete J, Cohen J. Health benefits of legal services for criminalized populations: the case of people who use drugs, sex workers and sexual and gender minorities. The Journal of law, medicine & ethics : a journal of the American Society of Law, Medicine & Ethics 2010; 38(4): 816–31. [DOI] [PubMed] [Google Scholar]
- 25.Diouf D HIV/AIDS policy in Senegal: a civil society perspective: Open Society Institute; 2007. [Google Scholar]
- 26.Mbali M The treatment action campaign and the history of rights-based, patient-driven HIV/AIDS activism in South Africa. Democratising development: The politics of socio-economic rights in South Africa 2005: 213–43. [Google Scholar]
- 27.Schwartz SR, Nowak RG, Orazulike I, et al. The immediate eff ect of the Same-Sex Marriage Prohibition Act on stigma, discrimination, and engagement on HIV prevention and treatment services in men who have sex with men in Nigeria: analysis of prospective data from the TRUST cohort. The lancet HIV 2015; 2(7): e299–306. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Lyons CE, Schwartz SR, Murray SM, et al. The role of sex work laws and stigmas in increasing HIV risks among sex workers. Nature communications 2020; 11(1): 773. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Zahn R, Grosso A, Scheibe A, et al. Human Rights Violations among Men Who Have Sex with Men in Southern Africa: Comparisons between Legal Contexts. PloS one 2016; 11(1): e0147156. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Stahlman S, Hargreaves JR, Sprague L, Stangl AL, Baral SD. Measuring Sexual Behavior Stigma to Inform Effective HIV Prevention and Treatment Programs for Key Populations. JMIR public health and surveillance 2017; 3(2): e23. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplementary Table 1: Summary of data collection across 10 countries in sub-Saharan Africa
Supplementary Table 2: Country specific ethics approvals and inclusion criteria
Supplementary Table 3: Stigma Measures
Supplementary Table 4: Legal environment categories and definitions
Supplementary Table 5: Legal environment by country
Supplementary Table 6: HIV setting characteristics by country
Supplementary Table 7: Correlation of legal environment
Supplementary Table 8: History of HIV testing, HIV status, and awareness of HIV status among MSM by country
Supplementary Figure 1. Framework for legal environment, stigma, and HIV
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.