Introduction

Blacks in the USA typically have strong associations with religion and religious institutions. A recent national survey of religious behaviors and beliefs found that, relative to other racial and ethnic groups, Blacks are more likely to report a formal religious affiliation (Glick and Golden 2010; Pew 2003, 2008; Pitt 2010). A high majority (85 %) of Blacks reported that religion is very important to them, and 60 % of Black individuals surveyed reported weekly or “regular” church attendance (Pew 2003, 2008). Among the Black individuals surveyed in a nationwide probability sample who reported no formal religious affiliation, 60 % indicated that religion was somewhat or very important in their lives (Pew 2008). Comparatively, 56 % of the general population reported that religion was very important to them, and 39 % reported weekly or regular church attendance (Pew 2003, 2008).

In addition to religious differences among racial groups, persistent differences concerning the acceptance of homosexuality appear to exist among different racial communities. Blacks tend to report more negative attitudes regarding homosexuality than Whites and other racial/ethnic groups (Glick and Golden 2010; Pitt 2010). In a recent survey, 85 % of Blacks endorsed homosexuality as sinful and reported generally unfavorable views regarding homosexual males (Pew 2008). Similarly, Glick and Golden (2010) found that 72.3 % of Blacks surveyed endorsed homosexuality as “always wrong,” compared to 51.8 % of Whites. The study also found that this attitude toward homosexuality among Blacks was largely unchanged since the 1970s, but among White respondents, negative perceptions of homosexuality declined from 70.8 to 51.8 %. As described below, these attitudes concerning religion and homosexuality and higher levels of homophobia experienced by Black MSM versus other MSM may be a contributing factor to the increased incidence of HIV/AIDS risk behaviors and support the need for continuing research (Glick and Golden 2010; Jeffries et al. 2012; Pitt 2010).

Cognitive Dissonance and Sexual Risk Among Black MSM

The strong association of many Blacks with predominately Black religious institutions likely influences their attitudes regarding sexual norms (Pew 2008; Pitt 2010; Stokes and Peterson 1998; Yip 2002). Due to the negative attitudes of many Black churches regarding sex in general, and homosexuality in particular (Miller 2007; Wilson and Miller 2002), many Black MSM report finding themselves in conflict with and rejected by organized religious communities (Fullilove 2006; Miller 2007; Ross et al. 2008; Wilson and Miller 2002; Woodyard et al. 2000). Such conflicts and incongruities between key life areas can lead to a disconnection between religious identity and homosexual behavior. It is possible that Black MSM internalize the homonegative (judgmental) messages often present in their religious and larger community contexts, giving rise to dissonant self-perceptions that foster HIV-related risk (Glick and Golden 2010; Wilson and Miller 2002; Woodyard et al. 2000).

Due to stigma some black MSM may be less likely to be long-term monogamously partnered, less likely to arrive at high-risk locations with prevention supplies, and/or less likely to engage in safer sex practices, such as condom use (Millett et al. 2006; Murray et al. 2007). All of these risk behaviors are associated with increased risk of STIs, which are proxy measures for HIV risk behavior engagement and also independently may increase risk of HIV transmission (Beatty et al. 2004; CDC 2009).

Several studies also have found that the internalized barriers of stigma, shame, low self-esteem, and homophobia create an atmosphere of silence around MSM sexual behaviors and decrease the probability of Black MSM discussing high-risk sexual behaviors. This culture of silence can lessen the impact of HIV prevention education interventions (Jeffries et al. 2012; Martin and Knox 1997; Malebranche 2003). Some studies also have suggested that the internalization of religiously influenced homonegative messages contributes to lower self-efficacy regarding preventive sexual practices among this population (Martin and Knox 1997; Noar et al. 2009; Stokes and Peterson 1998). Thus, it is important to examine the association between religiosity and spirituality and high-risk sexual behavior among Black MSM.

Religiosity and Spirituality Among Black MSM

Although religion and spirituality as concepts often overlap and much research has found significant associations between the two (Horn et al. 2005; Ironson et al. 2006), “religion” typically refers to the formal set of beliefs and practices affiliated with an acknowledged religious authority and includes such overt visible acts as praying before meals, attending worship services, and reading sacred texts (Davis et al. 2003; Coyle 2002). “Spirituality” typically denotes a personal experience or feeling of connectedness with a higher being and typifies intrinsic unseen qualities appropriated to ones relationship with an otherness, higher power, or God. Spirituality can also include the search or path for transcendent meaning (Davis et al. 2003; Coyle 2002; Miller 2007).

Despite the unique role of religiosity and spirituality among Black MSM, research has not examined the relationships between these constructs and risk behaviors. The present study was exploratory in nature and was aimed at increasing understanding of population views and needs to enhance HIV prevention and intervention programs targeting Black MSM.

Methods

Description of the Data Sample

The present study used a subset of data from the 2005 Brothers y Hermanos Study (ByHS) conducted by the Centers for Disease Control and Prevention (CDC). Several factors make the ByHS an appropriate choice to address the specific aims of this project. First, the ByHS study was rigorously designed and assessed a broad range of relevant areas. ByHS used an audio computer-assisted self-interview (ACASI) that allowed private responding to solicit that most unbiased responses on sensitive questions (e.g., HIV risk, UAI, substance abuse). Second, the ByHS data set provides a very large sample (N = 1140) of Black MSM recruited from large urban areas. This sample will provide adequate statistical power to examine the specific aims. Third, the ByHS data set provides comprehensive data on HIV risk behaviors, HIV testing, and perceptions of HIV risk. The level of risk data obtained in this federally funded study exceeds that possible in most primary data collection by unfunded researchers. Fourth, the ByHS data set contains excellent quality data, with few missing data points. Fifth, the ByHS data set assesses many of the covariates of HIV risk identified in the literature (e.g., comprehensive substance use assessment, depression/mental health, social support). Sixth, the ByHS data set includes questions on both religiosity and spirituality, although no studies using these data have been conducted from this data set.

ByHS collected a sample of 1141 Black MSM (601 from New York City and 540 from Philadelphia). Eligible ByHS participants were born male, were over age 18, and were sexually active with another male within the past 12 months (i.e., oral or anal sex, mutual masturbation with another male). Recruitment was open to participants who were HIV positive, HIV negative, and unknown HIV status (Marks et al. 2009). Data collection sessions were held in project offices in the CBO office spaces where participants were screened for eligibility, and written informed consent was obtained. Eligible participants completed an audio computer-assisted self-interview (ACASI) in English. The religiosity, spirituality, and high-risk behavior questions were all confidentially via the ACASI system.

Selection of the Data Set

The overall goal of the original ByHS study was to understand differences in undiagnosed HIV infection among Black MSM. The ByHS data set was selected for use in the present study as it is, to our knowledge, the only large data set (N = 1154) of Black MSM respondents that assessed both religion and spirituality, as well as comprehensively assessing HIV status, risk behaviors, and associated variables.

ByHS participants were recruited using respondent-driven sampling (RDS) (Heckathorn 1997), which is a variant of peer-to-peer chain-driven sampling method. Participants who did not self-disclose as HIV positive at intake were provided HIV counseling and testing (OraQuick oral swab); individuals testing preliminary positive for HIV infection received venipuncture blood confirmation testing and returned for results and posttest counseling. Participants were compensated $50 for completion of study procedures.

Description of the Sample of Participants

Table 1 shows the demographic characteristics of ByHS participants included in the present study. A majority of the sample identified as non-Hispanic Black and single/never married; over half identified as homosexual or gay. Nearly half were unemployed, and educational attainment ranged from non-high-school graduate to doctoral level (mean educational level was high-school graduate/GED). Average participant age was 41.5 years (range 18–71, SD = 9.63).

Table 1 Participant characteristics of black men who have sex with men, Brothers y Hermanos Study, 2005–2006
Full size table

Measures (Religiosity, Spirituality, HIV Risk)

Dependent Variables

The dependent variables were religiosity, and spirituality. Measures for these variables religiosity (4 questions) and spirituality (3 questions) are presented with scale responses as noted in Table 2.

Table 2 Independent variable characteristics presented with variable name, number, and corresponding percent of black men who have sex with men, Brothers y Hermanos Study, 2005–2006
Full size table

Outcome Variables

This exploratory study examined relationships among sexual risk (unprotected sex, HIV infection, STD infection) and other high-risk behavior variables (substance use) and religiosity and spirituality (Table 3).

Table 3 Risk variable characteristics of black men who have sex with men, Brothers y Hermanos Study, 2005–2006
Full size table

A religiosity index was created (Hsueh-Sheng 2012) by summing the scale responses of the four religiosity questions from the ByHS data: worship (0–4) + openness (0–4) + religious beliefs (0–4) + choosing religious beliefs (0–4) to develop a composite sum of the responses to the four original ByHS religiosity questions. Figure 1 shows the total number of responses for each number and its corresponding percentages.

Fig. 1
figure 1

Religiosity index responses with corresponding N, and percent from participants

Full size image

Similarly, a spirituality index was created by using participant responses to three questions by summing guidance (0–4) + spiritual connection (0–4) + spirituality and health (0–4). Spirituality responses are noted in Fig. 2

Fig. 2
figure 2

Spirituality index responses with corresponding N, and percent from participants

Full size image

Analysis

Procedures to Identify Correlations of Religiosity and Spirituality

To examine which risk and demographic variables were significantly correlated with both religiosity and spirituality, a correlation matrix demographic and risk variables (i.e., unprotected anal intercourse (UAI), both receptive and insertive, condom use history, HIV infection status, and STD infection status) was created, and Pearson correlations were examined along with the corresponding p values and reported in Table 4.

Table 4 Correlations with religiosity and spirituality indices
Full size table

A series of ANOVAs were conducted to examine the relationship between age and religiosity as well as age and spirituality. Both religiosity (F[1, 1134] = 18.20, p < 0.0001) and spirituality (F[1, 1134] = 4.50, p = 0.0342) were significantly positively associated with age.

To further analyze the risk variables associated with religiosity and spirituality, the Chi-square (χ2) and corresponding p value for each statistically significant risk variable were computed and are reported in Tables 5 and 6. These tables include the individual religiosity and spirituality index subquestions as well and the corresponding p values.

Table 5 Chi-square and p values of religiosity associated with risk variables
Full size table
Table 6 Chi-square and p values of spirituality associated with risk variables
Full size table

Results

The correlation between religiosity and spirituality was 0.1634 (p < 0.0001); however, these were significantly associated with different risk variables indicating that they are distinct constructs.

Religiosity Religiosity was significantly associated with age (r = 0.1331, p < 0.0001), cocaine use (r = 0.0978, p = 0.0010), and crack use (r = 0.1128, p = 0.0001), and significantly negatively associated with ecstasy use (r = −0.0617, p = 0.0375), poppers use (r = −0.0698, p = 0.0185), and HIV infection status (r = −0.1352, p = <0.0001). The religiosity index was positively, but nonsignificantly, associated with non-prescription substance use. Nonsignificant negative associations were found between the religiosity index and both unprotected receptive and insertive anal intercourse.

Chi-square analysis of the religiosity index and various risk variables found that the religiosity index was significantly associated with HIV infection χ2 [(1123, N = 1125) = 17.51, p = <0.0001], cocaine use χ2 [(1135, N = 1137) = 10.29, p = 0.0013], crack use χ2 (1135, N = 1137) = 14.60, p = <0.0001], poppers use χ2 [(1135, N = 1137) = 4.52, p = 0.0330], and marginally associated with ecstasy use χ2 [(1135, N = 1137) = 3.38, p = 0.0661].

Spirituality The spirituality index was positively associated with STD infection status (r = 0.0829, p = 0.0051), and HIV infection status (r = 0.1900, p = <0.0001), and negatively associated with receptive anal intercourse (r = −0.0835, p = 0.0443), unprotected anal intercourse (r = −0.0684, p = 0.0212), alcohol use (r = −0.1245, p = <0.0001), cocaine use (r = −0.0719, p = 0.0153), and crack use (r = −0.1164, p = <0.0001). Chi-square analysis found significant associations between the spirituality index and HIV infection status χ2 [(1123, N = 1125) = 36.98, p = <0.0001], STD infection status χ2 [(1136, N = 1138) = 10.73, p = <0.0001], alcohol use χ2 [(1135, N = 1137) = 30.11, p = 0.0093], and crack use χ2 [(1135, N = 1137) = 4.73, p = 0.0296].

Discussion

Religiosity and spirituality were significantly associated with various high-risk behaviors among Black MSM. Although religiosity and spirituality were strongly associated, their pattern of relationships with other variables was different, suggesting that the two are related, but distinct, constructs.

More religious Black MSM tended to report cocaine and crack use and were more likely to be HIV negative and to participate less in unprotected receptive anal intercourse. This pattern of results suggests a complex relationship between risk and protective factors and HIV infection in this sample. As found in other studies, Rasic et al. (2011), and Cotton et al. (2006), religiosity was significantly negatively associated with certain substance use, such as poppers and ecstasy, and may promote less use of these substances among Black MSM. Religiosity among Black MSM who use crack and cocaine was associated with higher levels of risky behavior as found in other studies and may increase risk of HIV and STD infection (Klobin et al. 2006; Parsons et al. 2012; Reisner et al. 2009; Stall et al. 2003). These men may have higher levels of dissonance, which in keeping with our theory may account for the high-risk behavior.

Similarly, in other studies (Davis et al. 2003; Hampton et al. 2010), MSM who reported a higher degree of spirituality also tended to be HIV infected, and in this study they also tended to report ever having an STD. Given a lack of ability to denote a temporal association in this study, Black MSM’s level of spirituality could have increased upon HIV or STD infection occurred explaining the positive association between spirituality and these constructs.

Spirituality was negatively associated with unprotected anal sex, receptive anal intercourse, alcohol use, cocaine use, and crack use. As in other research (Murray et al. 2007; Wutoh et al. 2011), which found that increased levels of spirituality corresponded to decreased levels of certain risky behaviors, this study found that among Black MSM, spirituality might be protective of unprotected insertive anal intercourse, unprotected casual sex, and non-prescription substance use. One possible inference is that these men may have achieved a level of cognitive consonance, accounting for these differences in behavior.

HIV infection status or STD infection status is not a risk behavior; however, Black MSM who were infected did participate in high-risk behaviors including having unprotected sex with a partner whose HIV infection status was unknown.

Given that religiosity may discourage the use of poppers and ecstasy, and spirituality might discourage the practice of unprotected receptive anal intercourse, unprotected anal intercourse, and the use of alcohol, crack, and cocaine, current and future HIV prevention models targeting Black MSM might incorporate religiosity and spirituality to increase the efficacy of high-risk reduction outcome measures.

Limitations

First, the use of a cross-sectional study design introduces the possibility of measured and unmeasured confounding factors, and causality cannot be inferred from the associations presented in this study. Second, the ByHS survey relied on participant self-report of risk behaviors, as well as the religiosity and spirituality questions, and so responses may not accurately reflect the true risk behaviors or an accounting of religious and spirituality behaviors although collected by confidential computer-assisted methods. The religiosity and spirituality questions were combined and measured independently of self-report Third, although cognitive dissonance and consonance and their relationships with religiosity and spirituality, formed the theoretical framework of the study, the data set did not allow direct measurement of consonance or dissonance. Fourth, while religiosity and spirituality among HIV-infected individuals is reported, the cross-sectional data set does not allow assumptions regarding whether these levels changed with HIV infection status. Better understanding temporal relationships between HIV infection status and religiosity and spirituality is an important avenue for future research. Finally, the religiosity/spirituality items available in the ByHS data set are not part of standardized scales. This limits comparisons of findings from this study to other studies that have relied on standardized religiosity/spirituality measures. However, past research has not identified gold-standard religiosity/spirituality scales (Johnstone et al. 2009), and no standardized measures have been used with Black MSM.

Summary

This exploratory study found that religiosity and spirituality are constructs important and significantly associated with certain high-risk behaviors among Black MSM. This paper illustrates which variables are significantly associated with these constructs. Given the relationships between religiosity and spirituality and certain high-risk behaviors among Black MSM, future risk reduction interventions may achieve increased efficacy by incorporating these constructs in the design, development, and implementation of behavioral models for Black MSM. The relationships identified in this paper support the value of further study of the potential roles of the religiosity and spirituality constructs in risk behaviors among Black MSM. Such developmental work is necessary for incorporation into tailored risk reduction programs for this underserved, high-risk population.

Future Implications

Given that HIV infection rates continue to increase among Black MSM and that other social and cultural factors account for the new incidence among this population, this study used a theoretical framework of cognitive dissonance theory to guide the direction of the research and found religiosity and spirituality to be significantly associated with high-risk behaviors among Black MSM. Religiosity may be associated with more dissonance among Black MSM and may promote more crack and cocaine use, and spirituality may be associated with cognitive consonance and may decrease the practice of unprotected anal intercourse and use of alcohol, crack, and cocaine among Black MSM. Future studies and health behavior theories should consider incorporating these constructs in the design and testing of risk reduction interventions for this population.