Introduction

The high prevalence of racial discrimination reported by African American men in the US has been well-documented [1–5], and growing research showcases significant relations between racial discrimination experiences and negative physical and mental health related outcomes [6–9]. To date, however, much of the existing work on racial discrimination and health with African American communities has primarily focused on chronic health conditions; less is known about how experiences of racial discrimination may influence health behaviors within this population.

One arena of health behavior that is of tremendous concern within African American communities is violence involvement. African American men are disproportionately represented among those involved in and affected by neighborhood and non-intimate forms of violence [10, 11]. Additionally, while intimate partner violence (IPV) exists across all racial/ethnic groups, evidence indicates that African American women are overrepresented among victims of IPV and IPV-related homicide [12–14], with the majority of females experiencing IPV victimized by males of the same race/ethnicity [10]. Recent neighborhood-level research suggests that higher rates of IPV perpetration are more common among neighborhoods characterized by higher rates of neighborhood and gang violence, neighborhoods often characterized by the effects of marginalization, disenfranchisement, and disproportionately populated by racial/ethnic minority residents [15, 16]. Unfortunately, despite its disproportionate impact on African American communities [12, 17], as well as the range of critical public health issues associated with violent experiences (e.g. HIV/AIDS, injury, homicide), no work to date has specifically examined the impact of racial discrimination on violence among African American men.

The study of the impact of racial discrimination on violence involvement is of particular importance given research regarding the co-occurrence of male perpetrated violence (e.g. IPV, neighborhood violence) with men’s exposure to other forms of violence (e.g. child abuse, parental abuse) [18–21]. Further, recent research suggests that racial discrimination can lead to specific health behaviors (e.g., substance use) [22–26] as well as depression and psychological stress [6, 27–30] that may contribute to perpetration of IPV and neighborhood violence.

Community disenfranchisement and marginalization (e.g. including high rates of unemployment and poverty etc.) related to a history of institutionalized racism experienced by African American communities in the US likely serve as structural factors relevant to the high levels of reported neighborhood and intimate partner violence in such communities [31–40]. However, no studies have examined whether individual experiences of racial discrimination contribute to an individual’s risk for neighborhood or intimate partner violence perpetration among racial/ethnic minority men in the US.

The purpose of the current study was to assess whether experiences of racial discrimination are associated with neighborhood violence involvement and perpetration of intimate partner violence among a sample of African American men recruited from urban neighborhoods with high rates of violence. More specifically, the current study investigated the relevance of racial discrimination to the following: (1) perpetration of intimate partner violence, (2) involvement in street violence, and (3) involvement in gangs.

Methods

The current study used data from the Black and African American Men’s Health Study (BAAMH), which involved a cross-sectional survey conducted with sexually active African American men (N = 703) between the ages of 18 and 65 recruited from four urban community health centers and primary and urgent care clinics within a large teaching hospital. All recruitment sites were within Boston neighborhoods characterized by higher rates of violence than that seen in Boston as a whole [41]. The present analyses evaluated data on violence that were collected as part of the larger study to assess sexual risk among a sample of African American men recruited from high risk neighborhood clinic settings. Eligible African American men were those aged 18–65 years, who reported sex with two or more partners in the past year, and demonstrated no cognitive impairment (assessed using the Folstein Mini-mental Exam [42]). Those eligible were invited to complete an audio computer-assisted survey interview (ACASI) assessing demographic variables (income, education, race, immigration, employment), racial discrimination, perpetration of or involvement in different types of violence, and other related variables assessing health risks (e.g. sexual risk, substance use).

Recruitment and Participation

Research staff recruited all African American men attending collaborating health centers and clinics from May 2005 to May 2006 during designated recruitment days and times. Days and times for recruitment to take place were rotated in order to reduce self-selection that could limit generalizability; sample differences were likely attached to time of day of appointment (i.e., interference of work schedules). Among the 2,331 men approached, 85% (n = 1,988) agreed to be screened for study eligibility. Of those screened (n = 1,988), 47% (n = 930) were eligible and 81% (n = 754) of eligible men agreed to participate in the study. The majority of men who were screened and ineligible did not meet eligibility criteria that required participants to have two or more sex partners in the past year. Those who declined participation and those who were ineligible were provided with local relevant social and health service referrals.

Among the 754 surveys collected, 51 (7%) were removed from further data analysis as a result of survey responses that did not meet study criteria based on age and number of sex partners. Current analyses were restricted to men reporting only current female partners (given the focus on male perpetrated IPV against female partners), yielding a final sample size of 569. All 569 men in this sub-sample responded to study questions on exposure variables (7 discrimination items) and the outcome variable (i.e. street violence involvement, gang involvement, IPV perpetration).

Study Procedures and Participant Involvement

Immediately following eligibility assessment, oral informed consent was obtained among participants who agreed to participate. Following consent, participants were administered the 20–25 minute survey using an audio computer-assisted self-interview (ACASI) in a private setting with research staff available to answer questions. ACASI rather than interviewer-administrated surveys were implemented, given evidence of greater response disclosure on sensitive questions [43–47]. Participants were provided with a $35 compensation upon completion of the survey as well as social and health service referrals. All procedures of this study were approved by the Institutional Review Boards of Boston University Medical Campus and the Centers for Disease Control and Prevention. Additionally, a Federal Certificate of Confidentiality was obtained to provide further protections for study participants.

Measures

Demographic variables assessed included age (categorized as 18–34, 35–44, and 45–65 years of age), national origin (US and US territories versus outside of the US; n = 9 participants were from Puerto Rico), education level (high school education without graduation, having received a high school diploma or GED, or at least some college), employment (full/part-time employment or unemployed), and current homelessness (living on the streets or in a housing shelter).

Discrimination experiences were measured using seven items measuring major life events and everyday experiences of racial discrimination [48, 49]. Items were based on the Everyday Discrimination Scale [49] and focused on whether participants experienced the following events because of their race: “looked over for a higher position;” “followed by a security guard or watched by store clerks;” “followed, stopped, or arrested by police more than others;” “house was vandalized;” “someone disrespected me or ignored me because of my race;” “called insulting names related to my skin color or race;” or “was physically attacked or assaulted because of my skin color or race.” Items used a five-item likert response scale either ranging from “strongly disagree” to “strongly agree” or from “never” to “always/daily.” Each likert response was recorded as a score of one (indicating the least reported discrimination) to five (indicating a high level of discrimination). Average discrimination scores were calculated for each participant. Scores were dichotomized based on the median discrimination score of the sample (sample median = 2.3) and categorized as high discrimination (greater than 2.3) or low discrimination (2.3 or less). The seven items had a cronbach’s alpha (α = 0.79) indicating reliability of these items.

Neighborhood violence variables included street violence involvement (having engaged in any street fight or other form of street violence in the past six months) and gang involvement (ever been part of a gang). The IPV perpetration scale comprised four questions on physical abuse, sexual abuse, and injuries from abuse. Physical IPV perpetration was assessed via a single item on whether the respondent had ever “hit, slapped, punched, shoved, choked, kicked, shaken, or otherwise physically hurt” their current partner. Sexual IPV perpetration was measured using two items, asking each respondent, “have you made your partner take part in any sexual activity that she did not want to, including touching that made him or her feel uncomfortable” and asking “have you forced or pressured your partner to have vaginal, oral or anal sex with you.” IPV injury perpetration was assessed via a single item asking whether the respondent’s partner ever “had any injuries, such as bruises, cuts, black eyes, or broken bones as a result of being hurt by you.” Respondents indicating yes on any of these items were defined as having perpetrated IPV in their current relationship.

Data Analysis

Bivariate associations between discrimination (dichotomized as high or low discrimination) and demographics were evaluated using contingency table analysis; Chi-square tests were conducted to assess statistically significant differences. Crude and adjusted logistic regression analyses were used to assess the associations of discrimination (dichotomized) with the violence outcome variables (IPV perpetration, street violence, and gang involvement). Demographic variables associated with an outcome variable in bivariate analyses at P < 0.1 were included in the adjusted regression model for that outcome. Odds ratios and 95% confidence intervals were used to assess effect sizes and significance of associations were measured using Wald Chi-square tests.

Results

Demographic Characteristics and Involvement in Violence

Almost three quarters of the sample were 44 years of age or younger. The majority (83%) were US born. Slightly more than one-fourth of men (28%) reported having less than a high school education; 61% were unemployed, and 22% were homeless in a shelter or on the street (Table 1).

Table 1 Demographics and distribution by average discrimination score
Full size table

Almost the entire sample (96%) reported experiencing some type of racial discrimination (not shown in table). Men reporting high discrimination scores (i.e., a discrimination score above the sample median) were more likely to be among the older groups of participants (χ2 = 25.6, df = 2, P < .0001), homeless (χ2 = 5.2, df = 1, P = .02), have a higher level of education (χ2 = 13.4, df = 2, P = .001), and to be US born (χ2 = 7.2, df = 1, P = .006) compared to men reporting low discrimination scores. Discrimination was not significantly associated with employment in this sample (Table 1).

Approximately one-fourth of the sample (22%) reported perpetration of IPV during their current relationship. About 17% reported physically hurting their partner, 12% reported causing physical injuries, 10% reported forcing sexual activity (e.g. sexual touching), and 12% reported forcing their partner to have sex. Almost one-third of the sample (29%) reported street violence involvement in the past 6 months, and 22% reported ever being involved with a gang (Table 2).

Table 2 Associations between discrimination & IPV perpetration, street violence involvement, and gang involvement (n = 569)
Full size table

Crude and Adjusted Findings: The Relevance of Racial Discrimination to IPV Perpetration and Involvement in Gangs and Street Violence

In crude analyses, men reporting high discrimination were significantly more likely to report IPV perpetration (OR = 2.1, 95%CI = 1.4–3.1); about 28% of men reporting high discrimination score reported perpetrating IPV in their current relationship, compared to only 16% of men reporting lower discrimination. These findings remained significant in adjusted analyses. Although no significant associations were found between discrimination and the gang and street violence variables in crude analyses, a significant relation between discrimination and street violence involvement was found in the age and education-adjusted model (OR = 1.5; 95% CI: 1.1–2.2) (Table 2).

Discussion

The current study is among the first to demonstrate that experiences of racial discrimination are associated with increased likelihood of violence perpetration, above and beyond the influence of socioeconomics, among this sample of urban African American men. Further, this demonstrated link between racial discrimination and violence was found for both intimate partner and street violence perpetration. These findings advance the current state of knowledge by adding violence perpetration and involvement to the growing body of literature on negative health factors that have been associated with this form of social oppression. While such findings are congruent with one existing study by Caldwell and colleagues [50] which found that racial discrimination was related to violence-related behaviors (i.e. group fight, using a knife or gun) among African American youth, no studies have documented the link between racial discrimination and IPV previously.

Literature findings suggest that racial discrimination, recognized as a form of structural violence, may also contribute to other forms of violence perpetration. For example, prior evidence demonstrates the relevance of exposure to violence or violence victimization to violence perpetration; studies show a relation between exposure to family violence and sexual assault victimization and relation to male IPV perpetration [18, 19] as well as between neighborhood violence exposure and IPV perpetration among men [16]. Such evidence across studies suggests that contextual violence supports norms tolerant to violence as well as other factors that contribute to subsequent violence involvement such as substance use and psychological distress. Hypotheses also suggest that male disempowerment (as a result of racial discrimination and marginalization) may result in a desire to demonstrate masculine identity and status within the context of female intimate relationships as well as within the community [38, 51–56]. Future studies are needed to help clarify potential mechanisms for such relations between racial discrimination and violence involvement as well as overlapping experiences of violence.

Unadjusted models examining the relation between discrimination experiences and street violence produced null findings; however, men who reported street violence were more likely to be younger and to have lower educational levels; such groups were also less likely to report discrimination. In other words, the relation between experiencing discrimination and street violence involvement was confounded by age and education. Therefore, upon adjusting for age and education, a statistically significant association between experiencing discrimination and involvement in street violence was evident. Previous studies have similarly found that reports of discrimination often vary by level of education, where those with higher levels of education are more likely to report racial discrimination compared to those with lower levels of education [22, 57]. Reporting of discrimination has been found to vary by socioeconomic characteristics in various previous studies as well [22, 57, 58], therefore further showcasing the importance to consider such variables in investigations related to racial discrimination. For example, higher socioeconomic position, older age, and non-immigrant status may be associated with greater exposure to situations where men may be discriminated against or these men may also be more likely to recognize such situations as discrimination. While socioeconomic factors are associated with both racial discrimination as well as violence perpetration, study findings suggest the existence of a relation between racial discrimination and violence perpetration variables (IPV perpetration and street violence involvement) above and beyond such socioeconomic influences.

Notably, current study findings did not find a significant association between reported discrimination and involvement in gangs. It is likely that involvement in gangs may be more directly influenced by other factors within specific neighborhoods (e.g. social acceptance, camaraderie, protection [59, 60]). Future research is needed, however, to further investigate experiences of racial discrimination and relevance to gang involvement in order to confirm such findings.

Current study findings must be considered with recognition of several study limitations. A primary limitation of the current study is the cross-sectional design; temporal order of the link between racial discrimination and experiences of violence could not be established. In terms of measures, violence measures may be underestimated and were collected during different time periods. For example, the current study only considered IPV perpetration in current relationships with female partners, and gang violence was measured as ever occurring. Further, men may feel stigma attached to reporting perpetration of violence and therefore, may be underreporting these behaviors. However, the use of ACASI has been found to reduce reporting biases of stigmatized behaviors such as illicit drug use and sexual practices [46, 61] and has been recommended for assessing IPV [62]. Despite these limitations, large proportions of men reported current IPV perpetration (22%), street violence involvement (30%), and gang involvement (22%) in the current study.

Racial discrimination is also subject to potential measurement issues, given that the items reflect “perceived discrimination” (e.g. variation in recognition of incidents as discrimination), as described in previous work examining the relevance of discrimination to health [63]. The relevance of age and education as modifying variables (i.e. to examine whether the relation between experiences of discrimination and street violence involvement varied by age or education group) was not possible for further exploration, given limited study power to detect a significant interaction. Findings also have limited generalizability due to the Northeastern US clinic-based sample of African American men reporting a current female main partner and two or more sex partners in the past year. Additionally, as this study included men who were seeking varied types of non-medical programs at recruitment sites, findings cannot be generalized to those seeking traditional or primary care. However, although the current study sample is less generalizable, one major strength of the current study includes the focus on young to middle aged African American men (those most affected by violence) from communities with high levels of violence. Further, given the geographic population of clientele served, the current study sample was also restricted to communities with similar racial/ethnic and sociodemographic profiles. Such restriction reduced the risk for confounding by further “controlling” for the influence of sociodemographics in determining the relation between racial discrimination and intimate and neighborhood violence perpetration and involvement variables (street violence and gang involvement).

Conclusions and Prevention & Intervention Implications

The findings of the current study suggest that the relevance of racial discrimination to the public health of racial/ethnic minority communities extends beyond the effects on opportunities related to employment, education, and other life events. While a history of institutionalized racism has created structural and socioeconomic factors (e.g. segregated neighborhoods; disenfranchisement; high rates of violence, poverty, unemployment, and homelessness) characteristic of the neighborhoods sampled by the current study, study findings further demonstrate that individual experiences of racial discrimination are associated with individual reports of intimate and street violence perpetration and involvement.

While future studies are needed to establish the temporal direction of this association, consideration of the relevance of racial discrimination and further exploration of the context of violence specific to the experiences of African American communities is critical for the development of tailored and effective violence prevention programs. Regarding racial discrimination as a form of structural violence, the current study underscores the need for further study and intervention in order to disrupt overlapping processes of violence, particularly within contexts characterized by high levels of violence, and to ultimately reduce racial disparities in violence linked to this cycle.