Introduction

Black men who have sex with men (MSM) are heavily affected by the HIV/AIDS epidemic in the United States. The HIV prevalence among non-Hispanic black MSM is estimated by the Centers for Disease Control and Prevention (CDC) to be 28 % compared to 18 % for Hispanic MSM and 16 % for non-Hispanic white MSM [1]. More than half (59 %) of black MSM who are HIV-seropositive are not aware of their infection [1] and therefore are not accessing HIV care and treatment or aware of their ability of transmit HIV to others. Black MSM account for approximately 35 % of new HIV infections among all MSM compared to 46 % for white MSM and 19 % for Hispanic MSM. HIV incidence among young black MSM is of particular concern. Forty-eight percent of new infections among MSM aged 13–29 years were among black MSM [2].

There is a growing body of literature that examines the elevated rates of HIV among black MSM. In 2006, Millett published a seminal literature review that examined 12 hypotheses to explain the high rates of HIV among black MSM. This review concluded that the disparities in HIV could not be explained by high-risk sexual behavior, a lack of disclosure of sexual identity, or drug use. He found insufficient or conflicting evidence in research that examined: genetic susceptibility, circumcision, duration of infectiousness, HIV status of partners, incarceration and anal douching. Millett found that differences in HIV rates could be partially explained by high prevalence of sexually transmitted diseases (STDs) and unrecognized infection among black MSM. A 2007 meta-analysis by Millett quantified many of the conclusions from his 2006 review. The meta-analysis found that compared to white MSM, black MSM reported less substance use, fewer sex partners, and were less likely to be on ART. Black MSM were more likely to have a STDs. The meta-analysis found no differences by race in unprotected anal intercourse, commercial sex work, HIV testing history, or sex with a known HIV-positive partner [3].

This paper augments Millett’s 2006 and 2007 articles by reviewing the current literature on behavioral, biomedical, structural, social contextual, psychosocial, and social network factors that affect HIV rates among black MSM. The paper examines the status of the current research in each of these six areas and will make recommendations for future research. The paper also provides up-dates on the 12 hypotheses put forward by Millett by examining new research in these areas and explores the following topics not covered in Millett’s 2006 and 2007 papers: serosorting, quality of HIV care, stigma and discrimination, peer support, mental health, and social networks. This review aims to answer the question “What factors help to explain elevated rates of HIV among black MSM in the United States?”

Methods

We searched three databases: PubMed, Scopus, and Google Scholar. We did the literature review in two stages. First we searched all articles that included black or African American and MSM and HIV. This electronic database search resulted in a large number of articles (over 4,000). We then limited these articles to include only studies that focused on our six areas of inquiry. We searched the following terms for each area: behavioral (drug use during sex, crack cocaine use, and serosorting); biomedical (circumcision, STDS, and STIs); structural (access to care, HIV care, ART, HAART, patient-provider communication, HIV quality of care); social contextual (stigma, discrimination, internalized homophobia, internalized heterosexism, medical mistrust, social isolation, and incarceration); psychosocial (peer support and mental health); and social network (sexual mixing, partner characteristics, and social networks) factors. At this stage, many articles were excluded based on their titles and/or abstract because they were irrelevant to our review. Based on title and abstract, we identified 71 articles which we assessed for eligibility. Of these 71, 39 were included in our review. We focused primarily on studies published after 2006 to avoid repetition and redundancy. However, we included literature published prior to 2006 when it was necessary to provide an accurate and complete assessment and when focusing on a topic area not covered by Millett’s literature reviews. (Of the 39 articles included in the review, seven were published before 2006.) We excluded studies that were not conducted in the United States, Europe, or Australia because of potential cultural differences between study populations. We also excluded dissertations, editorials, letters, and commentaries. Data from each article were extracted, organized into summary tables to highlight study population, sample size, study design, and key findings. All but three studies included in this review were quantitative. For quantitative studies, we included results from bivariate analyses but emphasized multivariate analyses when available (Table 1).

Table 1 Studies of black MSM and HIV risk
Full size table

Results

Behavioral Factors

Millett’s 2006 literature review found that high-risk sexual behavior and substance abuse could not explain the high rates of HIV among black MSM [4]. Recent research confirms these findings [3, 5]. However, Millett’s review noted the need for further research on substance use during sex. A recent study suggests that lower-income black MSM were more likely to use crack during sex and less likely to use methamphetamine during sex compared to lower-income non-Hispanic white MSM [6]. We found three studies that assessed substance use during sex among samples of black MSM. Tieu et al. [7] found that HIV-positive black MSM were more likely to use substances at last sex than HIV-negative men (48.8 vs. 22.0 %, p ≤ 0.0001). A study of black MSM in Boston found that drug use during sex was associated with unprotected anal sex [8, 9]. Millett found that undiagnosed seropositive black MSM were more likely than uninfected black MSM to have used alcohol or drugs during most recent unprotected anal intercourse (UAI) with a male partner (35.2 vs. 24.2 %, p ≤ 0.05) but that drug or alcohol use during UAI was not associated with unrecognized HIV infection among a sample of black MSM [10]. Thus, initial research suggests that substance use during sex does not explain the disparity in HIV by race. Additional research is needed to confirm these findings. Future research needs to ask about specific substances to understand differences by race.

Serosorting is limiting unprotected sex to partners of the same serostatus and strategic positioning is assuming either the insertive or receptive role during unprotected sex based on HIV status. The effectiveness of these harm-reduction techniques is widely contested. Research has shown that both HIV-negative and HIV-positive MSM guess their partners status which would undermine a protective effect [11]. While several studies suggest that serosorting offers limited protection from HIV [11–13], others have concluded that serosorting [14, 15] and strategic positioning [15] might be effective. A study of self-reported HIV-negative black and Latino men found that the odds of testing HIV-positive were significantly higher among men who did not limit unprotected anal sex to HIV-negative partners compared to men who only reported HIV- negative partners (serosorters) and that black MSM were less likely to serosort than Latino men [15]. Research suggests that black MSM are less likely to have unprotected anal sex with presumed HIV-negative men, more likely to report UAI with partners of unknown HIV status, and less likely to endorse serosorting beliefs compared to white MSM [16]. Together this research suggests that black MSM are less likely to serosort than white MSM. Given inconclusive evidence around the effectiveness of serosorting, it is unclear if this difference could help to explain racial disparities in HIV.

Biomedical Factors

Circumcision

The results from three randomized control trials in Africa of heterosexual couples suggest that adult male circumcision reduces HIV acquisition for the insertive male partner [17–19]. The applicability of these results to MSM is being explored. A meta-analysis of 15 studies published in 2008 found an insignificant protective effect among MSM who primarily engaged in insertive intercourse and concluded that there was insufficient evidence that male circumcision protects against HIV infection or other STDs [20]. A review conducted in 2010 found weak and inconclusive evidence that circumcision reduces HIV and STDs among MSM [21]. A third review conducted in 2011 found that the pooled effect estimate for HIV acquisition from 20 studies was not statistically significant [OR 0.86 (95 % CI 0.70–1.06)]. However, a sub-group analysis of seven studies found significant effects for men reporting an insertive role [OR 0.27 (95 % CI 0.17–0.44)] and no effects for men in the receptive role [OR 1.02 (95 % CI 0.63–2.29)] [22].

Several US studies suggest that blacks are historically less likely to be circumcised than whites [23–25]. However, hospital data from the past decade suggests that black and white newborns are equally likely to be circumcised [26]. We found one study that looked at circumcision status and HIV among minority MSM. The study found that black MSM were more likely to be circumcised than Latino MSM, circumcision was not protective against HIV overall and that this relationship did not differ by sexual position [27].

The research on the preventative effective of circumcision does not suggest that the preventative effect of circumcision would be strong enough to curb the epidemic among MSM. Research on circumcision among black MSM is scarce but does not support the hypothesis that differences in rates of circumcision could help explain the high rates of HIV among black MSM.

Sexually Transmitted Diseases

There is strong biological evidence that having a STD increases risk for acquisition and transmission of HIV. STDs increase susceptibility to HIV infection through ulcers and inflammation and increase transmission through viral shedding [28]. There is substantial evidence that black MSM are more likely than white MSM to have STDs [28–31]. In addition, HIV-positive black MSM are more likely to report a recent STD than HIV-negative black MSM [32]. The trends in STD prevalence by race observed by Millett in 2006 and 2007 are still apparent and help to explain the persistently elevated rates of HIV among black MSM.

Structural Factors

Access to Health Care

One possible explanation for the elevated HIV prevalence among black MSM is lower quality health care. While several studies published before 2006 indicated no differences among MSM by race in hospitalization, emergency room (ER) visits, inpatient visits, source of HIV medical care, perceived ability to access HIV medication, or insurance [33–35]. A Chicago-based study found that areas where black MSM reside typically have low HIV service densities and that while HIV prevalence clusters spatially, HIV services do not. These findings point to lower availability of HIV prevention services for black MSM [36].

However, black and white MSM are equally likely to have been seen by a health care provider in the past 12 months [5] and are equally likely or more likely to have been offered an HIV test by their health care provider [5, 37]. Black MSM and white MSM also are equally likely to have been tested for HIV. CDC surveillance indicates that in the past year 61 % of black, 60 % of Hispanic, and 62 % of white MSM were tested for HIV [38]. Despite similar proportions of testing, black MSM had lower CD4 counts at AIDS diagnosis than MSM of other races and shorter survival times [39].

HIV Care and Treatment

Recent research shows that use of antiretroviral therapy (ART) prevents the sexual transmission of HIV among heterosexual HIV-serodiscordant couples [40–44]. Therefore access to ART is important not only for the health and wellbeing of the HIV-positive individual but also for HIV prevention. Both Millett’s 2006 and 2007 articles found that black MSM are less likely to be on ART than white MSM [3, 4]. A recent study confirmed this, finding that black MSM were less likely than white MSM to have been to a health care provider for their HIV infection within 3 months of diagnosis and to be on ART [37]. However, a study of HIV-positive MSM attending an STD clinic in Washington from 2004 to 2008 found that black and white MSM were equally likely to be on ART in multivariate analyses [45].

Studies suggest that blacks are less likely than whites to adhere to ART [46, 47] or to be retained in care [48]. In his review, Millett found the research on ART adherence to be inconclusive. For example, a study by Halkitis et al. [33] of MSM in New York City and San Francisco found no difference in adherence by race while analyses of the Multicenter AIDS Cohort Study (MACS) found that blacks were less likely to be adherent [49, 50]. Recent analysis of the MACS data from 2002 to 2006 found that black MSM were more likely to be non-adherent than white MSM [AOR 1.37 (95 % CI: 1.05–1.79) p = 0.020] [51]. The study also found that among black MSM increased cost of personal expenditure for prescription medication, financial difficulty meeting expenses, crack use, and having a new skin rash were associated with not adhering to ART [51].

In summary, HIV-positive black MSM might be less likely to access care and treatment services, to be on HAART, and to be adherent to HAART and this could help to explain disparities in HIV prevalence by race. Additional research is needed to understand barriers to care and treatment among black MSM and to further investigate differences in retention and adherence by race.

Quality of HIV Care: Patient Provider Dynamics and Medical Mistrust

Negative encounters within medical institutions and interpersonal barriers might help to explain poorer outcomes of HIV care among black MSM. Medical mistrust is high among African American populations [52–54] and is a barrier to accessing HIV care and treatment services [55, 56]. Trust in physicians is associated with HIV-related outpatient clinic visits, fewer ER visits, acceptance of ART, increased use of ART, and ART adherence [55–58]. The quality of patient-provider relationships is associated with greater receipt of ART, ART adherence, and viral suppression [59–61]. Racial concordance between patient and provider influences perceived quality of care, receipt of preventative care [62] and time of receipt of HIV therapy [63]. However, the role that trust and provider relationships plays in explaining differences in HIV by race is not clear. A recent study found that while patient provider cultural differences are negatively associated with perceived quality of care and trust, these factors did not account for racial/ethnic disparities in HIV care [64].

The research on patient provider dynamics and medical distrust among black MSM is limited. Medical distrust among black MSM appears to be high [65–67] and influences health care utilization, HIV testing, patient-provider communication and ART adherence [66]. A recent study of black men who have sex with men and women found that service providers’ attitudes towards black MSMW are affected by negative beliefs towards bisexual behavior. These findings underscore the need for black MSMW to have access to safe and non-judgmental HIV services [68].

Research on black MSMs’ experiences within the HIV medical system is limited. More research is needed in this area to (a) better understand the relationship between medical distrust and HIV rates among black MSM and (b) to better understand and address barriers to black MSMs’ use of care and treatment services.

Social Contextual Factors

Stigma and Discrimination

Non-stigmatizing attitudes towards same-sex behavior are important for HIV reduction. Research has found that same-sex partnership laws, including same-sex marriage, led to statistically significant reductions in syphilis [69]. A United States-based study found that tolerance for same-sex behavior was associated with lower HIV rates and found evidence that this was due to more low-risk men entering the pool of partners and a decrease in risk behaviors [70].

Stigma undermines the health of MSM. Stigma limits access to health care and health insurance; contributes to poor mental health, stress, and social isolation; affects the ability to have long-term stable relationships; and contributes to substance use, risky sex, and suicide [71]. The effect of stigma might be particularly acute for black MSM because they face multiple forms of discrimination, including discrimination due to race, sexual orientation, gender identity and HIV status. In addition, many black MSM are exposed to heterosexist community and familial norms [72–74]. A study of US attitudes towards homosexuality found blacks were 1.15 times as likely to report that homosexuality was always wrong as whites and that black MSM were twice as likely to report that homosexuality was always wrong compared to white MSM (57.1 vs. 26.8 %, p = 0.003) [75]. A study of discrimination among black MSM found that 30.6 % reported discrimination based on race, 39.0 % felt discrimination based on sexual orientation, and 28 % reported that their family did not approve of their same-sex behavior [76].

While there is a strong theoretical framework to support the hypothesis that increased exposure to stigma and discrimination is associated with HIV risk among black MSM [77, 78], we located a small number of studies that quantified this relationship. A 1996 study found that black men reported higher frequency and cumulative severity of lifestyle hassles than white men and that HIV- positive gay minority men were more prone to psychological stress than white HIV positive MSM [79]. A small number of studies have assessed the relationship between stigma and HIV risk behavior. A study of young black men found no association between discrimination and unprotected insertive or receptive anal intercourse [76]. A study of HIV positive young black MSM found that individuals exposed to higher levels of HIV stigma reported more unprotected sex while under the influence of alcohol or drugs [80] (Table 1).

Internalized heterosexism has been defined as the internalization of negative messages about homosexuality [81] while internalized homophobia as been defined as the internalization of society’s homophobic attitudes [82]. In the literature these similar constructs have been referred to jointly as internalized heterosexism/internalized homophobia and abbreviated as IH [81]. A literature review of the effect of IH on HIV risk behavior among MSM was inconclusive. The review identified studies that support a relationship between IH and HIV risk behavior as well as studies that found no relationship or an inverse relationship [81]. Newcomb found that the study of IH suffers from methodological challenges such as a wide variation in operationalization of IH and his meta-analysis of 31 studies of IH and risky sexual behavior found a small to moderate effect size [82]. The majority of current research on IH has focused on non-minority samples and more research is needed to understand if IH impacts HIV risk among black MSM [81, 82].

One potential result of low levels of tolerance and high levels of stigma and discrimination is social isolation. Social isolation may affect HIV risk behavior. A study of black MSM in Massachusetts found that social isolation was positivity associated with unprotected sex with a serodiscordant partner [8]. To summarize, there is a strong theoretical basis for a relationship between stigma, discrimination and HIV among black MSM. However, we found only a few articles that quantitatively assessed the effect of stigma and several of these studies had very modest effect sizes. The current research on IH is inconclusive and suffers from methodological shortcomings. IH is a distal construct that includes multiple dimensions. This provides challenges for measurement and operationalization. Measures should be validated and tested for cultural appropriateness. There is a need for studies that assess the intersection of multiple types of stigma on HIV and on the use of care and treatment services among black MSM.

Incarceration

In his 2006 review article, Millett concluded that there was insufficient evidence with regard to the role that incarceration might play explaining high rates of HIV among black MSM. While he found evidence that black MSM were more likely to have been incarcerated than MSM of other races, he only located one article which tested the association between incarceration and HIV among black MSM and this study did not support the hypothesis that incarceration increased risk for HIV [83]. We found three recent articles that were relevant. A study of MSM using national surveillance data found that incarceration was not associated with HIV infection in a mixed race sample [37]. A study by Mangus et al. [5] found that black MSM were more likely than MSM of other races to have ever been to jail, prison, or a detention center but that, among black MSM, having ever been to jail, prison, or juvenile detention was not associated with HIV seropositivity. A CDC study of HIV transmission among male inmates of a state prison in Georgia found that black race was associated with seroconversion during incarceration [84]. However, this study was not limited to MSM. Additional studies are needed among black MSM to assess the relationship between incarceration and HIV.

Psychosocial Factors

Peer Support

Studies of peer support among black MSM suggest strong peer norms for condoms use and that supportive peer norms for condom use predict less-frequent risky behavior [76, 85, 86]. The Brothers Y Hermanos study found that MSM with more supportive relationships had lower odds of testing HIV-positive and that this relationship was partially explained by a greater likelihood of being tested for HIV and a lower likelihood of engaging in risky sexual behaviors [87]. Additional resiliency research is needed to identify strengths and assets of black MSM, such as supportive relationships and peer support for condom use, to inform interventions that aim to reduce the HIV burden among black MSM.

Mental Health

Mental health factors, including depression, disproportionately affect black MSM [88, 89]. An exploratory study of the mental health of black MSM concluded that heterosexist and homophobic expectations of sexual orientation and norms of hegemonic masculinity (male behavior that is patriarchal and characterized by aggressiveness and strength) negatively affect the mental well being of black MSM [90]. However, research from the 1990s suggests that among black MSM depression is not related to HIV serostatus [91, 92]. We found no recent studies that assessed the relationship between mental health and HIV and only one study that looked at mental health and sexual risk behavior among black MSM. The study found that depressive symptoms were associated with serodiscordant unprotected anal sex with a casual male partner and having a recent STD in the past year [93]. Thus, research suggests that depression is elevated among black MSM and that depression may be related to HIV risk behaviors in this population. However, it is not clear that mental health plays a role in explaining HIV disparities.

Sexual Identity and Disclosure of Same Sex Behavior

Research has found that black MSM are less likely than MSM of other races to disclose their same-sex behaviors to others [94] or to identify as gay or homosexual [37, 85, 95]. HIV infection is more likely among MSM who identify as gay compared to MSM who identify as non-gay [96, 97]. This relationship may be true for black MSM as well. An analysis of black MSM found that undiagnosed seropositive men were more likely to be gay-identified than to be non-gay identified [10].

Research also suggests that minority MSM are less likely to disclose their same-sex behavior to their health care providers [98]. MSM who disclose same-sex behavior to a health care provider are more likely to be tested for HIV [98–100]. However, research does not suggest that a lower level of disclosure to health care providers among black MSM compared to MSM of other races has translated into differences in frequency of HIV testing by race [4].

Network Factors

The social networks of black MSM might help to explain the elevated HIV prevalence in this population. For example, researchers have speculated that high background prevalence of HIV, high rates of undiagnosed seropositivity, and high rates of STDs within racially assortative sexual networks of black MSM might allow for rapid transmission of HIV. Social network studies of heterosexuals have found that where initial prevalence of disease was high, assortative mixing by race increased disease within that racial group [101]. Several studies have found that black MSM are more likely to have partners who are also black [102–104] which suggests a high level of racial assortative mixing within this group. However, the role sexual networks play in HIV among black MSM is not clear. A study of black young men in Mississippi found no association between partner race and HIV infection [105]. A recent study by Millett found that UAI with a black partner was not associated with unrecognized HIV infection among black MSM [10]. However, a study by Maulsby found that among a mixed-race sample of men who have sex with men only, having an black male partner was associated with increased risk for undiagnosed seropositivity [106]. Research suggest high levels of assortative mixing by race among black MSM but current research is inconclusive as to the contribution of partner race in explaining elevated rates of HIV among black MSM.

Another sexual mixing pattern that might explain elevated rates of HIV among black MSM is disassortative mixing by age. If younger black MSM are engaging in unprotected sex with older MSM, a population with an elevated HIV prevalence, this could lead to an increase in HIV risk for younger MSM. A DC-based study found that having an older partner was not associated with HIV positivity among black MSM or white MSM [5]. A study of CDC surveillance data found that black and white MSM with newly diagnosed HIV infection did not differ in age of most recent partner and that age of partner was not associated with newly diagnosed HIV among a mixed race sample [37]. A study of young black MSM found that men with partners older than 25 years were at increased risk for HIV infection [AOR 5.5 (95 % CI: 1.8–17.3)] [105]. A study of black and Latino MSM aged 18–35 years found that having a partner that was at least 4 years older increased odds of having an unrecognized infection (AOR 2.51, 95 % CI 1.18–5.34) [107]. Two studies support the hypothesis that partnerships with older men might help to explain elevated rates of HIV among black MSM [105, 107] and two studies do not support this hypothesis [5, 37]. Additional research is needed in this area.

If black MSM are more likely to have partners of unknown or positive HIV status than white MSM this might help to explain differences in HIV rates by race. White and black MSM are equally likely to have sex with a known HIV-positive partner (0.77, 95 % CI 0.34–1.74) [3]. A small number of articles have looked at difference by race in having a partner of unknown HIV status. One study found no differences by race in likelihood of having a partner of unknown HIV status [108] and a second found that black MSM were less likely than white MSM to have partners of unknown HIV status [109]. A third recent study found that among newly diagnosed HIV-positive MSM, black MSM were more likely than white MSM to report that their last partners had an unknown HIV status (53 vs. 41 %) and less likely to report having HIV positive partners (5 vs. 10 %). In multivariate analyses, having a partner with a positive HIV status [AOR 1.9 (95 % CI: 1.2–3.0)] and having a partner with an unknown HIV status [AOR 1.4 (95 % CI: 1.1–1.7)] were positively associated with newly diagnosed HIV infection [37]. Thus, research in this area is inconclusive and additional research is needed to understand if having a partner with unknown HIV status contributes to the disparity in HIV by race.

Discussion

To better understand the factors that contribute to increased rates of HIV among black MSM, this review assessed the status of the current literature in six areas. Current research confirms Millett’s findings that the elevated rates of HIV among black MSM cannot be explained by differences in HIV risk behavior [5]. In addition, circumcision does not appear to play a role in differences in HIV by race [27].

Like Millett, we found that research in the area of incarceration was inconclusive and that additional research is needed to understand if incarceration affects HIV status among black MSM. Additional research is also needed to better understand the social impacts of incarceration among black MSM and how incarceration affects the sexual networks of black MSM. Our review found a need for further research in the related areas of stigma, discrimination, social isolation, and mental health. Areas for future inquiry include teasing out the potential roles of social isolation and mental distress as mechanisms through which stigma affects health and wellbeing.

Peer support for condom use decreases the likelihood of HIV sexual risk behavior among black MSM [76, 85–87]. This represents an area of strength. Though there is abundant evidence of resilience among MSM, most HIV scientific literature focuses on deficits [110]. There is a need for studies among black MSM to identify the resiliencies that are most effective at combating HIV so that interventions can be tailored to promote these strengths. Rates of STDs and rates of undiagnosed HIV seropositivity continue to be higher among black MSM [28–31, 111]. In addition, black MSM who are aware of their HIV infection are less likely to be on HAART [33, 37, 112, 113]. These three factors increase risk for HIV transmission [28, 114–116] and help to explain why we see higher rates of HIV among black MSM when compared to white MSM.

Evidence with regard to differences by race in retention in HIV care and non-adherence are inconclusive. Additional research is needed to determine if retention in HIV care and non-adherence differ by race and to identify factors associated with HAART use, adherence and retention among black MSM. Medical mistrust and patient provider dynamics impact HIV care and treatment. Additional research is needed to understand how these factors affect HIV testing, treatment and adherence among black MSM.

There is a compelling public health need for STD and HIV programs for black MSM. Improved access and linkage to STD screening and treatment is necessary to reduce STD incidence among black MSM. There is an urgent need for HIV programs that identify, link, and retain HIV-positive black MSM who are not in care into high-quality HIV care and treatment. To be successful programs must address the external root causes of delayed HIV care and dropping out of care, such as homelessness, joblessness, substance abuse, mental health, stigma, and discrimination based on race, gender, and/or sexual orientation. If black MSM are going to be successfully linked and retained in HIV care and treatment, interpersonal barriers to quality HIV care must also be addressed by ensuring that HIV care providers are culturally competent [63, 64, 67, 68].

There has been much speculation that some of the difference in HIV rates by race can be explained by differences in social networks. Our review looked at three characteristics of partners that could impact HIV transmission at the network level: race of partner, age of partner and partner HIV status. We found that evidence on the role partner characteristics play in explaining HIV disparities by race to be inconclusive.

This review has many limitations. First, it does not include meta-analyses to test specific hypotheses. Second, though we included findings from multivariate analyses wherever possible, in some cases we were limited to bivariate findings. Third, the number of studies that looked at HIV as an outcome was limited and thus our review also included analysis of studies that assessed the relationship between our factors of interest and HIV sexual risk behaviors. Fourth, we searched three data bases using a discrete list of terms. Searching other databases and using other terms might have identified additional studies which could have been included in this literature review.

Finally, the studies included in this review face a variety of methodological shortcomings which impact the quality of the data reviewed. The majority of the studies included in this review were cross-sectional and therefore temporality and causality could not be inferred. The studies reviewed used a variety of sampling strategies. Studies that used convenience samples are subject to the greatest amount of selection bias and limited generalizability. Studies employing venue-based and respondent driven sampling (RDS) strategies are also subject to potential biases. With RDS the non-random selection of seeds could affect the characteristics of the study population. With venue-based sampling, venues and participants are randomly selected but the study population is only generalizable to individuals attending venues similar to the ones sampled. The majority of studies gathered behavioral data via self-report which is subject to social desirability bias, especially when the study is interviewer-administered rather than computer-assisted. In addition, many of the studies reviewed had small sample sizes of black MSM which could limit the type of analyses conducted and increase Type II errors.

Despite these limitations, we feel that our findings are important for future research. Black MSM are one of the populations most affected by HIV and additional research is needed to better understand factors that contribute to this disparity. High rates of undiagnosed seropositivity, STDs, and decreased access to and use of HIV care and treatment place black MSM at greater risk. Culturally competent HIV and STD testing and treatment programs are needed for black MSM.