Abstract
In the current study, we investigated the emergence of racial-identity latent profiles and the potential for racial-identity profiles to moderate the relationship between race-related stress and trauma symptoms in nontreatment-seeking, trauma-exposed, Black American women (N = 222). Racial-identity profiles emerged from latent profile analyses and supported a three-class solution: undifferentiated, detached, and nationalist. Analysis of variance revealed that the nationalist profile group experienced significantly higher race-related stress compared with the detached and undifferentiated profiles. Moderation analyses revealed racial-identity profile type predicted both total posttraumatic stress disorder (PTSD) symptoms and each PTSD symptom cluster; specifically, the nationalist profile type buffered the effects of race-related stress on PTSD symptoms. This study illustrates how stress from racial discrimination influences PTSD symptoms and the ways racial identity may mitigate this relationship, which has the capacity to inform the adaptation of PTSD treatments for minoritized groups.
For Black Americans, racism is woven into every aspect of society and spans a spectrum of subtle, overt, and systemic manifestations (D. R. Williams & Mohammed, 2013). Negative stereotypes, attitudes, and beliefs about Black Americans have actively translated into policies that oppress Black Americans and allow interpersonal marginalization and abuse (D. R. Williams & Williams-Morris, 2000). Racial oppression of Black Americans is a well-known reality for this marginalized population and has led to a body of literature on racial discrimination. Although the effects of racial oppression on mental-health outcomes is established (Mouzon et al., 2017; Paradies et al., 2015), there continues to be a paucity of research that examines multiple and intersecting forms of oppression, such as racial discrimination and traumatic stress exposure. In particular, these understudied areas have rarely been approached with consideration of factors that may be particularly relevant to Black American women’s resilience, such as racial identity. The current study aims to target these gaps, with emphasis on examining the multidimensional construct of racial identity as potentially protective in the relationship between racial discrimination and trauma symptoms among Black American women.
The Potential Moderating Role of Racial Identity
When investigating the relationship between racial discrimination and trauma outcomes, it is necessary to examine the complexity of factors that may influence this relationship, particularly in the context of risk and resilience. Resilience in reaction to adversity, once considered a purely individual and trait-based concept, has evolved over time and is now recognized as a process that may emerge at individual, familial, community, and cultural levels (Fleming & Ledogar, 2008). Researchers have found that the resilience process is necessary to consider contextually both in relation to existing risk and with attention to potential outcomes (Fleming & Ledogar, 2008). Therefore, in conceptualizing the relationship among racial discrimination, trauma experiences, and posttraumatic stress disorder (PTSD), it is important to examine factors that may mitigate or exacerbate these dynamics among Black American women. Broadly, racial identity is a significant factor in the context of risk and resilience for marginalized groups (Arbona & Coleman, 2008; Ritchey, 2014) and is defined as the way people define themselves with respect to their race. The role of racial identity emerges as a potentially salient and reflective method to examine how appraisal of racial discrimination can shift depending on marginalized membership and how that may influence trauma responses.
To further consider the resilience potential of this construct, racial identity must be considered in the context of identity theory broadly. Scholars have defined identity theory as a conceptualization of how individuals behave, think, and navigate through life. Scholars have focused on various manifestations of identity, leading to research on specific identity classifications, including a body of literature on racial identity. One such model that has emerged as particularly comprehensive is Sellers and colleagues’ (1997) multidimensional model of racial identity (i.e., MMRI), which was developed to understand the dimensional nature of African American racial-identity development. In light of the known limitations of stage-based, unidimensional models of racial identity that typically focus on racial centrality, or the extent to which people define themselves by race, the MMRI considers the multidimensional quality of the African American experience (Sellers et al., 1997). Given this multidimensional consideration of identity, the MMRI lends itself to profile analysis, which has been argued to be a valuable approach for investigating the multiple facets of racial identity in response to complex oppressive experiences (Helms, 1996).
The MMRI proposes that African American racial identity is composed of stable and situational properties that reflect the ways in which racial identity influences behavior both in specific circumstances (molecular level) and across situations (molar level; Sellers et al., 1997). The MMRI considers the significance that African Americans place on race in regard to both defining themselves and defining their racial group as a whole and focuses on understanding African Americans’ beliefs that could influence behaviors across four identity dimensions: salience, ideology, centrality, and regard. Of the four dimensions, salience and centrality reflect an understanding of the significance of race, and ideology and regard illuminate what an individual attributes to experiences in the African American community and their own racial membership. Given that racial identity is better described as multidimensional as opposed to unidimensional, the MMRI lends itself for a unique exploration of potential identity profiles that could emerge through the examination of the African American identity experience.
Researchers testing the MMRI framework often use the Multidimensional Inventory of Black Identity (MIBI) scale (Sellers et al., 1997). Given the multidimensional nature of Black identity, the field calls for the use of mixture-modeling analyses that will highlight the multifaceted quality of how racial identity is experienced in Black American communities. Employing the MIBI to produce racial-identity latent profiles rather than examining single dimensions is minimal in the literature, but some research has been conducted to pinpoint which constellations of identity may promote resilience in Black American populations. For college-age Black American adults, the integrationist profile, characterized by high assimilationist/humanist ideology and lower private regard/centrality, was found to have a significantly stronger association between racial discrimination and depressive symptoms compared with other profiles (Banks & Kohn-Wood, 2007). A similar study with Black American adolescents found that the alienated profile, characterized by low centrality, private regard, and public regard, moderated the relationship between racial discrimination and depressive symptoms such that racial discrimination was positively associated with depressive symptoms for Black American adolescents who fit in this profile group (Seaton, 2009). Research examining the moderating effects of racial-identity profiles on the relationship between racial discrimination and imposter-syndrome phenomenon among Black American college students has also been conducted, but no significant effects were found (Bernard et al., 2018).
Although there has been some research broadly on the relationship between racial identity and PTSD symptoms in adolescents (Tyrell et al., 2019), the literature examining racial-identity profiles in relation to racial discrimination and PTSD is largely nonexistent. However, the research outlined above points to certain dimensions of racial identity, such as low centrality and low private regard, associating with greater negative mood symptoms, suggesting that high centrality or high private regard may be protective. Likewise, racial-identity indicators aligning with strong racial centrality and private regard have been found to buffer the negative effects of racial discrimination on mood, anxiety, and substance use disorders (Woo et al., 2019). So, although the research examining the protective role of racial identity in the association between race-related stress and mental health continues to be minimal, previous research points to racial identity as a potentially significant factor to further investigate. The current literature also mostly uses college samples and does not specifically focus on community samples, trauma-exposed samples, or Black American women (Mekawi, Lewis, et al., 2020). Black American women experience unique stressors that Black American men do not, including high rates of interpersonal violence, sexual assault, and unique forms of racial discrimination, such as gendered racism (Perry et al., 2013; Sue, 2010; Thomas et al., 2008). These experiences may also have a distinct impact on identity development, particularly in the context of risk and resilience in the face of trauma, adverse experiences, and systemic marginalization, illuminating the need to focus specifically on Black American women in the current study.
Racial Identity in the Context of Racial Discrimination and PTSD
When considering the role of racial identity in the context of traumatic and systemic stressors, note that the research literature on racism and trauma has historically been disconnected. Although both experiences of racial discrimination (i.e., the negative treatment of someone because of their race) and trauma influence mental-health outcomes, there is limited research understanding of how these experiences could be connected among Black American women specifically (Mekawi et al., 2021). Several researchers have theorized that race-related stress (i.e., distress that emerges from experiencing racial discrimination) operates as a psychosocial stressor in Black Americans (Castle et al., 2011; R. Clark et al., 1999; Utsey, 1999). According to this framework, distress experienced from racial discrimination has the potential to exacerbate a range of psychological symptoms. Researchers have found racial discrimination experienced by Black Americans to be associated with suicidality (Walker et al., 2017), low self-esteem (Harris-Britt et al., 2007), substance use disorders (T. T. Clark et al., 2015), generalized anxiety (Soto et al., 2011), severe psychological distress (Mouzon et al., 2017), depression (Russell et al., 2018), and PTSD (Mekawi et al., 2021). Scholars have also found that gendered racial discrimination, a combination of racial and gender discrimination, causes significant global psychological distress in Black American women (Thomas et al., 2008).
Although the impact of racial discrimination on trauma symptoms in Black American populations is understudied, a growing body of research literature shows that these experiences are significantly related (Pieterse et al., 2010; Roberson & Carter, 2022). Researchers have found that among Black Americans, racial discrimination is significantly associated with lifetime PTSD compared with other marginalized groups (Brooks Holliday et al., 2020; Chou et al., 2012). Black Americans also experience higher rates of trauma exposure and PTSD than White Americans (Himle et al., 2009; Roberts et al., 2011). Among a Black American and Latinx clinical sample, for example, experiences of racial discrimination significantly predicted PTSD diagnosis longitudinally over the course of 5 years (Sibrava et al., 2019). Psychosocial stressors similar to racial oppression, such as pervasive marginalization (Franklin & Boyd-Franklin, 2000) and victim blaming following trauma (Dukes & Gaither, 2017), may aggravate the severity of PTSD symptoms among Black Americans. Thus, it is clear that experiences of racial discrimination and trauma are highly connected for Black Americans.
Research with Black American women specifically has also revealed a significant relationship between oppression and trauma unique to their lived experiences. Scholars have found that Black American women are significantly more vulnerable to sexual violence compared with White American women and experience racial discrimination at higher rates than women of other racial or ethnic backgrounds (Perry et al., 2013; Sue, 2010). However, despite the documented disproportionate rates of racial and traumatic stressors Black American women face, little research has been conducted on this topic (Thomas et al., 2008). This gap calls for a specific focus on Black American women when examining the relationship between race-related stress and trauma symptoms.
Racial discrimination has also been associated with specific PTSD symptom criteria categories, further highlighting the potential impact of race-related stress on a myriad of PTSD symptoms (Sheehan et al., 2019; Slavin et al., 1991). Racial discrimination has been found to be related to numerous trauma-related symptoms (Kirkinis et al., 2021), including intrusive symptoms, avoidance symptoms, and hypervigilance symptoms (Carter et al., 2020; Sheehan et al., 2019). Black American women who experience racial discrimination have also been found to be fearful and hypervigilant of future attacks on themselves and their children (Hicken et al., 2013; Nuru-Jeter et al., 2009). In a sample of HIV-positive African American women, discrimination positively predicts higher posttraumatic cognitions, or appraisals and thoughts that develop after a traumatic event (Dale & Safren, 2019). Furthermore, the cumulative effect of racial discrimination can produce self-regulatory behavior such as social avoidance and dissociative episodes (Polanco-Roman et al., 2016).
Researchers have also considered the potential need to conceptualize the experience of racism as a traumatic experience (e.g., racial trauma, or PTSD caused by racism; M. T. Williams et al., 2018). However, despite theoretical frameworks that have conceptualized distress from racial discrimination as traumatic (Carter, 2007; Carter et al., 2005; Metzger et al., 2021), minimal empirical research has been conducted to indisputably support these theories. Measures of racial trauma are in need of further validation, and extensive research is needed to gather evidence for or against the notion that race-related stress operates in the same way as traumatic stress or warrants a separate classification. For these reasons, researchers have suggested incremental steps are needed to conceptualize distress from racism (M. T. Williams et al., 2018).
Despite racial identity being understood as a salient factor in the context of risk and resilience related to stressors such as racism and trauma, the identified gaps in the research literature prevent a clear understanding of the role of racial identity in the relationship between race-related stress and PTSD symptoms. Examining certain racial-identity profiles in the context of both race-related and traumatic stress has enormous potential to inform the understanding of what identity-related factors influence the relationship between stressful and traumatic experiences and mental-health difficulties. This knowledge may help modify interventions to address the negative impact of racial oppression on Black American women with trauma-related experiences.
Current Study
In the current study, we assessed whether multidimensional racial-identity profiles alleviate or intensify the negative associations between race-related stress and PTSD symptoms in a nontreatment-seeking sample of trauma-exposed Black American women. We hypothesized that (a) racial-identity profiles would emerge from latent profile analyses on the racial-identity data collected in our sample, (b) group differences in race-related stress or PTSD symptoms would emerge across racial-identity profiles, and (c) racial-identity-profile membership would predict PTSD symptoms in the context of race-related stress, emerging as a significant moderating factor in this relationship. Given that the existing research literature has shown racial identity may serve as a mitigating and exacerbating factor in the relationship between racial discrimination and mental-health outcomes, we hypothesized that the relationship between race-related stress and PTSD symptoms would be exacerbated by certain racial-identity profiles and that group differences may emerge in regard to race-related stress and PTSD; we made no particular claim on which profiles may be protective or exacerbating given the minimal literature on the relationship among racial-identity profiles, race-related stress, and trauma symptoms. Given the inconclusive literature on how racism-related stress influences various types of PTSD symptoms (Mekawi, Kuzyk, et al., 2021), individual PTSD subscales were considered as well as total PTSD symptoms. Examination of individual subscales will also contribute to the literature on how PTSD symptom clusters (e.g., hyperarousal, avoidance, reexperiencing) may manifest differently in Black Americans compared with other racial groups following racism-related stress (Mekawi et al., 2022).
Transparency and Openness
The current study is not preregistered. Readers can access study-related resources on clinicaltrials.gov (NCT03922581) and reporter.nih.gov (MH071537, MH100122, MH102890, MH115174). We report how we determined our sample size, all data exclusions, all manipulations, and all measures in the study. Ethical approval was granted by the university institutional review board (IRB) and the Research Oversight Committee of [Hospital Name Redacted].
Method
Procedure
Participants were recruited from waiting rooms in gynecology, primary care, and diabetes medical (nonpsychiatric) clinics at a publicly funded, nonprofit hospital serving a low-resourced population. Participants were approached at random. To be eligible, participants had to be at least 18 years old, not actively psychotic, and able to give informed consent. If willing to participate, individuals completed informed consent approved by the university IRB and the Research Oversight Committee of [Hospital Name Redacted], and an initial interview was administered with questionnaires regarding trauma history and other psychological variables. Trained research assistants administered this interview (approximately 45–75 min) in private areas of the waiting rooms of the hospital. Participants were compensated $15 for their time. For a detailed description of study methods, refer to Gillespie and colleagues (2009).
Participants
The current study included Black, adult, female participants (n = 222), ages 18 to 64. Reporting of sample size, all data exclusions, all manipulations, and all measures are included in this study. Participants were screened according to the inclusion/exclusion criteria of the larger study. Exclusion criteria includes exhibiting signs or behaviors of active psychosis and intellectual disability. Only individuals who identified as African American or Black, female, and between the ages of 18 and 64 were included in this study.
The majority of participants were single/never married (62.3%) and had a household monthly income of $1,999 or less (66.1%). Regarding education, 13.6% of participants had completed less than 12th grade, 36.2% completed high school or equivalent, 27.6% had some college or technical school, 18.6% graduated from college or technical school, and 4.1% completed graduate school. Approximately 56.5% of participants reported experiencing five or more traumatic events, and only 5% reported not experiencing any traumatic events. Descriptive statistics are presented in Table 1.
Descriptive Information for the Overall Sample
Measures
Demographic information collected from the participants included age, income, education level, gender, and race/ethnicity.
Race-related stress
Distress from racial discrimination (i.e., race-related stress) was measured by gathering participant responses to the Index of Race-Related Stress Brief (IRRS-Brief; Utsey, 1999), which is a 22-item measure of cumulative stress experienced as a result of individual, cultural, and institutional racism. Responses range from 0 (this has never happened to me) to 4 (event happened and I was extremely upset) and indicate exposure to three forms of racism and severity of stress. Higher scores on subscales are indicative of higher levels of race-related stress in each domain, and higher scores overall indicated higher levels of overall race-related stress. For college and community samples, the IRRS-Brief has shown adequate internal consistency (αs = .64–.81) across the three subscales (Utsey, 1999; Utsey et al., 2002; Utsey & Hook, 2007). Internal consistency reliability was assessed in the current sample and found to be strong (α = .93).
PTSD symptoms
Given that 31.6% of participants had completed the Modified Posttraumatic Stress Disorder Symptom Scale (mPSS; Coffey et al., 1998) and 68.4% participants had completed the Posttraumatic Stress Disorder Checklist (PCL) for DSM-5 (PCL-5; Weathers et al., 2013), PTSD symptoms were assessed by generating a combined PTSD score from the mPSS and the PCL-5.
The mPSS is a reliable and well-validated 17-item measure used to assess PTSD symptoms using the criteria from the fourth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV; American Psychiatric Association, 2000). Participants rate the degree to which they experienced symptoms on a scale of 0 (not at all) to 3 (five or more times a week/almost always).
The PCL is a widely used DSM-correspondent self-report measure of PTSD symptoms and was recently revised to reflect changes to the PTSD criteria made in the fifth edition of the DSM (DSM-5; American Psychiatric Association, 2013). The PCL-5 is a 20-item self-report measure that assesses PTSD symptoms on a 5-point Likert from 0 (not at all) to 4 (extremely). The PCL-5 scores exhibited strong internal consistency, test-retest reliability, convergent validity, and discriminant validity (Blevins et al., 2015).
To create the combined PTSD variable for this study, items from the PCL-5 that assessed for symptoms from the mood/cognition criteria group in the DSM-5 were removed. The remaining PCL-5 items were matched to mPSS items. A new Likert scale (0 = not at all, 1 = a little bit, 2 = moderately, 3 = quite a bit/extremely) was developed to reflect the scoring of both measures. When combining the measure, we aligned the mPSS 0 (not at all) with the PCL-5 0 (not at all), the mPSS 1 (once per week or less/a little bit/once in a while) with the PCL-5 1 (a little bit), the mPSS 2 (two to four times per week) with the PCL-5 rating of either 2 (moderately) or 3 (quite a bit), and the mPSS 3 (five or more times a week/almost always) with the PCL-5 4 (extremely). The combined measure resulted in 16 items. Scores were summed to create a total score and scores for reexperiencing symptoms, avoidance symptoms, and hyperarousal symptoms. Internal consistency reliability was assessed in the present study and found to be strong (α = .95). The mPSS reflects PTSD criteria in the DSM-IV (American Psychiatric Association, 2000), and the PCL-5 reflects PTSD criteria in the DSM-5 (American Psychiatric Association, 2013). The DSM-5 was released during data collection, which is why the PLC-5 was used instead of the mPSS after that point. The updated criteria in DSM-5 includes nearly identical avoidance, reexperiencing, and hypervigilance clusters as the DSM-IV and an additional fourth cluster reflecting mood/cognition symptoms. Thus, when creating the combined PTSD measure, the avoidance, reexperiencing, and hypervigilance items measured in the mPSS mapped exactly onto the items measured in the PCL-5 after removing the mood/cognition cluster items.
Racial identity
The MIBI-Short Form (MIBI-S; Martin et al., 2005) is a 27-item shortened version of the measure (Sellers et al., 1997) and reflects three stable dimensions (centrality, regard, and ideology) of racial identity proposed by the Multidimensional Model of Racial Identity (MMRI). The centrality scale measures the extent to which race is central to the respondents’ definition of themselves (e.g., “I am happy that I am Black”). The regard scale, composed of private- and public-regard subscales, assesses the extent to which respondents possess positive feelings toward African Americans and if they feel that other groups have positive feelings toward African Americans (e.g., private regard: “I feel good about Black people”). The ideology scale is composed of four subscales, assimilationist, humanist, minority, and nationalist, and assesses an individual’s opinion about how people with the same race membership should function (e.g., nationalist: “Black people should not marry interracially”). Participants indicated the extent to which they agree or disagree with items on a 7-point Likert scale from 1 (strongly disagree) to 7 (strongly agree). The full MIBI has been shown to have moderate to strong internal consistency, external validity, and structural validity, largely with college samples (Sellers et al., 1997; Simmons et al., 2008). Internal consistency reliability was assessed in the current study and found to be adequate (α = .76).
Trauma exposure
Trauma exposure was applied as a control variable consideration and measured using the Traumatic Events Inventory (TEI; Schwartz et al., 2005). The TEI is a 19-item screening instrument used to assess lifetime trauma history. Each frequency scale for trauma ranges on a scale from 0 or 1 time to greater than 20 times. Participants were asked to indicate the number of times they experienced various traumatic incidents. Traumatic events are summed to obtain an overall trauma-load score. Previous work has demonstrated that this measure is associated with more severe symptoms of relevant constructs, such as PTSD symptoms (Gillespie et al., 2009; Mekawi, Kuzyk, et al., 2021).
Data analysis
Bivariate correlation analyses were conducted in IBM SPSS (Version 25) to determine significant relationships among the variables of interest, including overall race-related stress, overall PTSD symptoms, each PTSD symptom cluster, MIBI subscales, and potential cofounding variables including age, income, education level, and trauma exposure.
Latent profile analyses (LPAs) were conducted using Mplus 8.4 to investigate potential racial-identity profiles (Muthén & Muthén, 1998–2017). Given that identity, including racial identity, is highly influenced by developmental processes, age was added as a covariate variable. Several steps were taken in Mplus to determine the best-suited LPA model. In the first step, k profiles were run to determine the best-fitting model. The factors that were considered in this determination include the log likelihood (LL), Akaike information criterion (AIC), Bayesian information criterion (BIC), sample-size adjusted BIC (SABIC), Vuong-Lo-Mendell-Rubin likelihood ratio test (VLMR-LRT), adjusted Lo-Mendell-Rubin likelihood ratio test (adjusted LMR-LRT), bootstrapped likelihood ratio test (bootstrapped LRT), and entropy (Masyn, 2013; Muthén & Asparouhov, 2008; Nylund et al., 2007). The smaller the AIC, BIC, and SABIC, the better the model fit (D’Unger et al., 1998; Nylund-Gibson & Choi, 2018). Significant p values on the likelihood tests (i.e., VLMR-LRT, adjusted LMR-LRT, and BLRT) suggest that the k-class model is more likely to represent the data than the k − 1 profile model (Asparouhov & Muthén, 2012). Finally, entropy was considered as an indicator to determine certainty of classification (Masyn, 2013). Entropy values range from 0 to 1; values greater than 0.80 indicate greater classification certainty (Celeux & Soromenho, 1996). Dummy coding was automatically applied to each of the profiles that emerged from the LPAs.
Before conducting the main moderation analyses, potential group differences were examined using one-way analyses of variance (ANOVAs) on IBM SPSS (Version 25). Finally, using PROCESS Macro 3.4.1, we conducted moderation analyses with overall PTSD symptoms and each symptom cluster as dependent variables to determine whether racial-identity profiles significantly moderate the relationship between race-related stress and PTSD symptoms. Income, education level, and trauma exposure were evaluated as covariates. The PROCESS macro has continuous variables mean-centered and produces the interaction term as the cross-product term of the centered variables.
Results
Correlations
Bivariate correlation analyses were conducted among race-related stress, overall PTSD symptoms, PTSD symptom clusters, MIBI subscales, age, income, education level, and trauma exposure. The analyses revealed that overall race-related stress was significantly and positively correlated with total PTSD symptoms (r = .29, p < .01), PTSD hyperarousal symptoms (r = .27, p < .01), PTSD avoidance symptoms (r = .27, p < .01), and PTSD reexperiencing symptoms (r = .27, p < .01). Total trauma exposure was significantly and positively correlated with race-related stress (r = .33, p < .01), total PTSD symptoms (r = .59, p < .01), PTSD hyperarousal symptoms (r = .58, p < .01), PTSD avoidance symptoms (r = .54, p < .01), and PTSD reexperiencing symptoms (r = .53, p < .01). Age was significantly and positively correlated with total PTSD symptoms (r = .39, p < .01), PTSD hyperarousal symptoms (r = .31, p < .01), PTSD avoidance symptoms (r = .26, p < .01), and PTSD reexperiencing symptoms (r = .31, p < .01). Education was positively correlated with race-related stress (r = .14, p < .05).
LPAs
Table 2 summarizes goodness-of-fit indices yielded from the LPAs. The entropy values decreased going from the two-, three-, four-, and five-class solution. The entropy values suggested that the two-class solution may be a strong contender for best fit, with the three-class solution as a possible option as well. The two-class solution had slightly higher entropy and slightly lower LL compared with the three-class solution, but consideration of other goodness of-fit indicators shifted the overall assessment of best fit. Specifically, the three-class solution had lower AIC, BIC, and SABIC values and more significant VLMR-LRT and adjusted LMR-LRT p values. This suggested that it was a superior fit compared with the two-class solution. The four-class and five-class solutions were assessed and quickly ruled out as superior solutions given that the VLMR-LRT and adjusted LMR-LRT values were both nonsignificant and entropy was lower. Thus, the three-class solution was retained because it demonstrated good fit compared with other profile solutions and it also reflected profiles discovered in previous research (Banks & Kohn-Wood, 2007; Bernard et al., 2018; Richardson et al., 2015; Seaton, 2009; Thomas et al., 2015). Thus, the three-class solution was retained (Fig. 1). Profiles were automatically dummy-coded as Class 1, Class 2, and Class 3 in Mplus.
Summary Statistics Describing Latent Profile Analysis Fit Indices
Note: AIC = Akaike information criterion; BIC = Bayesian information criterion; SABIC = sample-size adjusted Bayesian information criterion; VLMR-LRT = Vuong-Lo-Mendell-Rubin likelihood ratio test; adjusted LMR-LRT = adjusted Lo-Mendell-Rubin likelihood ratio test; BLRT = bootstrapped likelihood ratio test. *indicates significance at .05.
Racial-identity latent profiles depicted by standardized Multidimensional Inventory of Black Identity subscales means.
Class 1 (n = 23, mean age = 34.13 years) had lower than average scores on public regard (e.g., how one believes others view Black people) and higher scores on nationalism (e.g., placing significance on being of African descent), resembling the race-focused subgroup found by Banks and Kohn-Wood (2007) and the defensive/buffering profile reported by others (Richardson et al., 2015; Seaton, 2009). Particularly in the context of this group also having relatively higher centrality scores (e.g., the extent to which people defines themselves by race) and relatively higher scores on private regard (e.g., how a person feels about being Black), this subgroup appeared to have pro-Black attitudes and was named “nationalist.”
Class 2, the smallest subgroup (n = 18, mean age = 38.5 years), was lower than average on all subscales, with particularly deviant scores on the centrality, private regard, and nationalism subscales. However, scores on public regard were close to average, indicating this subgroup appeared to have negative views about being Black, did not centrally define themselves by a Black identity, and also did not perceive others as having particularly negative attitudes toward Black individuals. This was similar to the detached profile found in previous literature (Mekawi, Lewis, et al., 2020; Seaton, 2009; Thomas et al., 2015). Thus, this subgroup was named “detached.”
Class 3 was the largest subgroup (n = 181, mean age = 36.68 years) and had relatively average scores on all subscales. This subgroup resembled profile findings from other studies (Banks & Kohn-Wood, 2007; Bernard et al., 2018) and consistent with this, was named “undifferentiated.”
Given the small number of individuals classified within the nationalist and detached profiles, we also entertained the possibility that the largest subgroup was better characterized by two or more subgroups by examining the profiles developed in the four-class and five-class solutions (Table 2) and found that the large profile group found in the three-class solution persisted. Furthermore, there is evidence that when trauma is examined, profiles characterizing a small segment of the sample may emerge, which is corroborated by previous research (i.e., < 5%; Contractor et al., 2018; Holt et al., 2017).
Moderation Analyses
Before conducting the main moderation analyses, we examined potential group differences using one-way ANOVAs on IBM SPSS (Version 25). ANOVAs were used to investigate whether any differences among racial-identity profiles exist in relation to race-related stress and PTSD symptoms. Using Tukey’s post hoc tests to correct for Type I error, we found that the nationalist profile had significantly greater levels of overall race-related stress compared with the detached profile (p = .01) and the undifferentiated profile (p = .04). There were no significant differences among the three profiles in overall PTSD symptoms (Nationalist profile, Detached profile, Undifferentiated profile; p = .92, p = .88, p = .10). There were also no significant group differences among the three profiles in reexperiencing symptoms (p = .84, p = .95, p = .90), avoidance symptoms (p = .10, p = .10, p = .98), or hyperarousal symptoms (p = .73, p = .52, p = 1.00). No significant group differences emerged across the profiles in regard to demographic variables in this study, including education, income, employment status, relationship status, and disability status.
Moderation analyses are illustrated in Figure 2. Examining total PTSD symptoms as an outcome (n = 196), we found that the moderation analysis yielded a significant overall model (R2 = .64, p < .001). The results revealed that race-related stress (b = 0.10, t = 2.28, p = .024), racial-identity profile (b = 0.17, t = .13, p = .90), and their interaction (b = 0.15, t = 2.40, p = .02) accounted for significant variance even with accounting for total trauma exposure (b = 2.17, t = 9.62, p < .001) in the model. Racial identity and its interaction with race-related stress was also statistically significant (∆R2 = .02, F = 5.77, p = .02). Follow-up analyses examining conditional effects of different profile of the moderator (racial identity) revealed significant differences among the slopes of the profile whereby the detached profile (p = .02, 95% confidence interval [CI] = [.01, .18]) and the undifferentiated profile (p = .00, 95% CI = [.05, .23]) had a steeper slope than the nationalist profile. Thus, for the nationalist profile, as race-related stress increased, PTSD symptoms stayed steady, and for individuals in an undifferentiated or detached profile, PTSD symptoms increased as race-related stress increased. This pattern held when controlling for education, income, and trauma exposure.
Moderation analyses (total symptoms, reexperiencing symptoms, avoidance symptoms, and hyperarousal symptoms).
In the second model examining reexperiencing symptoms as an outcome (n = 197), the moderation analysis also yielded a significant overall model (R2 = .57, p < .001). Race-related stress (b = 0.03, t = 2.00, p = .05), racial-identity profile (b = −0.04, t = −0.08, p = .93), and their interaction (b = 0.04, t = 2.00, p = .05) accounted for significant variance even with trauma exposure (b = 0.64, t = 7.98, p < .001) in the model. Racial identity and its interaction with race-related stress was statistically significant (∆R2 = .014, F = 4.02, p = .05). Follow-up analyses examining conditional effects of different profiles of the moderator (racial identity) again revealed significant differences among the slopes of the profiles whereby the detached profile (p = .047, 95% CI = [.00, .06]) and the undifferentiated profile (p = .009, 95%CI = [.01, .08]) had a steeper slope than the nationalist profile. Similar to the first moderation, reexperiencing symptoms stayed steady as race-related stress increased for the nationalist profile, and for individuals with an undifferentiated or detached profile, reexperiencing symptoms increased as race-related stress increased. Education, income, and trauma exposure were again added as covariate variables.
The third model examined avoidance symptoms as the outcome (n = 196). The overall model was significant (R2 = .59, p < .001). Race-related stress (b = 0.04, t = 2.12, p = .03), racial-identity profile (b = −0.15, t = −0.26, p = .80), and their interaction (b = .06, t = 2.19, p = .03) accounted for significant variance even with trauma exposure (b = 0.80, t = 8.39, p < .001) in the model. Once again, racial identity and its interaction with race-related stress was statistically significant (∆R2 = .02, F = 4.82, p = .03). Follow-up analyses examining conditional effects of different profiles of the moderator (racial identity) revealed significant differences among the slopes of the profiles whereby the detached profile (p = .03, 95% CI = [.00, .07]) and the undifferentiated profile (p = .01, 95% CI = [.02, .09]) had a steeper slope than the nationalist profile. Thus, members of the nationalist profile experienced steady avoidance symptoms as their race-related stress increased. Undifferentiated and detached profile members experienced increased avoidance symptoms as race-related stress increased. Education, income, and trauma exposure were again added as covariate variables.
The final moderation model investigated hyperarousal symptoms as the outcome (n = 197). The overall model was significant (R2 = .63, p < .001). Race-related stress (b = 0.03, t = 2.09, p = .04), racial-identity profile (b = 0.39, t = 0.84, p = .40), and their interaction (b = 0.05, t = 2.33, p = .02) accounted for significant variance even with trauma exposure (b = 0.74, t = 9.59, p < .001) in the model. Racial identity and its interaction with race-related stress was again statistically significant (∆R2 = .02, F = 5.46, p = .02). Follow-up analyses investigating conditional effects of different profiles of the moderator (racial identity) again revealed significant differences among the slopes of the profiles whereby the detached profile (p = .04, 95% CI = [.00, .06]) and the undifferentiated profile (p = .01, 95% CI = [.01, .08]) had a steeper slope than the nationalist profile. Examination of the interaction plot revealed that the nationalist profile members’ hyperarousal symptoms actually decreased slightly as their race-related stress increased. Once again, the undifferentiated and detached profiles experienced increased hyperarousal symptoms as race-related stress increased. Education, income, and trauma exposure were again added as covariate variables for this final moderation model.
Discussion
In the present study, we investigated the multidimensional construct of racial identity as potentially protective in the relationship between racial discrimination and trauma symptoms among Black American women. These questions were investigated in a nontreatment-seeking sample of trauma-exposed Black American women in an urban setting. Three distinct racial-identity profiles (i.e., nationalist, undifferentiated, and detached) emerged from the sample, and significant moderation effects were found. The results also indicated a particularly strong interaction effect with hyperarousal and avoidance symptoms. When further examining the racial-identity profiles that emerged, we found that the nationalist profile both experienced significantly higher race-related stress than the other two groups and appeared to serve as a buffer against the negative influence of race-related stress on PTSD symptoms, whereas the detached and undifferentiated profiles appeared to be exacerbating the influence of race-related stress on PTSD symptoms.
Racial identity in the context of risk and resiliency
The racial-identity profiles that emerged from the current sample largely reflect profiles found in the literature, with some important points of distinction. The undifferentiated profile, characterized by scores mapping onto the mean for all subscales, is an established profile in the small body of literature that exists using the MIBI to produce racial-identity profiles (Banks & Kohn-Wood, 2007; Bernard et al., 2018; Scottham & Smalls, 2009). Likewise, the detached profile (e.g., lower than average on all subscales with particularly low private regard, centrality, and nationalism scores and higher public regard) mirrors MIBI profiles in other studies (Seaton, 2009; Thomas et al., 2015). The nationalist profile, with scores that are relatively higher on nationalism, private regard, and centrality subscales and low on public regard, also reflects previous literature findings (Banks & Kohn-Wood, 2007; Richardson et al., 2015; Seaton, 2009). In contrast to similar studies, the undifferentiated profile accounted for 81.5% of the current sample, which is a much larger undifferentiated profile than other researchers have found (Banks & Kohn-Wood, 2007; Bernard et al., 2018; Scottham & Smalls, 2009). Thus, the majority of women in this study can be described as having average or midlevel attachment toward being Black, with a distinct group of women deeply connected to their racial identity and the world’s perception of them (nationalist) and another small group actively detached from their Blackness (detached) both in how they view themselves and in how they believe the world sees them.
In contextualizing the racial-identity profiles that emerged from the current study, it is necessary to consider the unique experiences of the sample and how this may influence racial-identity development. Scholars have described racial identity as both a process and journey that is produced from self-reflection on one’s racial existence privately and publicly (Sellers et al., 1997; Vandiver et al., 2002). It is this concept of identity formation that may inform the emergence of the large undifferentiated racial-identity profile in the current study. The sample of Black American women in this study have experienced significant traumatic and systemic stressors that have affected access to education, wealth, and overall stability. These characteristics are a potential departure from typical undergraduate-student study samples that have been used in MIBI profile research, although rates of trauma exposure among these study sample groups are rarely reported (Banks & Kohn-Wood, 2007; Bernard et al., 2018; Seaton, 2009; Shelvin et al., 2014). It is possible that the participants in the current study have not had access to experiences and/or opportunities that naturally allow for self-exploration and critical consideration of racial identity, such as certain educational environments (Lannegrand-Willems & Bosma, 2006; Murray-García et al., 2005). The life experiences of the women in this study may demand a focus on day-to-day survival, which could leave little room for identity exploration. Alternatively, the midlevel identity development could also be viewed as an effective level of racial-identity development within context. Given the multiple stressors that are likely already present in the lives of these women, it could be that an undifferentiated racial identity is adopted as a resiliency response to eliminate the possibility of additional stress associated with holding multiple marginalized identities (Volpe et al., 2019).
These areas of research may also inform why the nationalist profile is so modest in size. Because this profile is characterized by both a deep connection to being Black and a strong awareness of the historical and systemic oppression that informs how American society views Black individuals, an Afrocentric worldview may be viewed as mentally taxing. The detached profile that emerged in the current study also can be contextualized within a greater narrative. The detached profile is distinct because it is low on all subscales but relatively higher on public regard compared with the low private regard, nationalism, and centrality scores. For this group of women, they have a negative perception of themselves and also underemphasize the societal marginalization that Black Americans often face. Black American women navigate statistically more stressors and traumatic experiences, and internally and externally disengaging from racial realities could reflect a need to avoid additional stressors (Mekawi, Lewis, et al., 2020; Thomas et al., 2008).
Racial identity in the context of race-related and traumatic stress
The current ANOVA and moderation analyses used in this study were applied to further explore the risk and resilience potential of the racial-identity profiles described above. In line with our second aim, we found significant group differences in relation to race-related stress such that the women who held a nationalist racial identity experienced significantly higher race-related stress compared with the other two profile groups. The profiles significantly moderated the relationship between race-related stress and overall trauma symptoms, as well as each symptom cluster, with a particularly strong effect found with hyperarousal and avoidance symptoms. Thus, the role of racial identity emerged as a salient factor in the relationship between race-related stress and trauma symptoms for Black American women. It appears that the appraisal of racial oppression is notably affected by one’s relationship with their racial membership, which then influences experiences of trauma symptoms. How women relate to their Black identity significantly influenced the relationship between race-related stress and avoidance and hyperarousal symptoms in particular and suggests the need in their contextual environments to avoid or be hyper-attentive to potential threatening racist experiences. At this point in time, the literature examining racial-identity profiles in relation to racial discrimination and PTSD is minimal, making these findings particularly compelling. Additional research in this area is necessary to substantiate the results from the current study, particularly regarding the significant effects found with hyperarousal and avoidance symptoms.
The implications of the current findings can be more adeptly explored by examining how each racial profile individually functions in these interaction effects (Fig. 2). In the above discussion on the racial-identity profiles that emerged from the current study, the emergence of the large undifferentiated profile was considered in the larger context of risk and resilience and the plenitude of stressors experienced by Black American women. However, when examining the results of moderation analyses for the undifferentiated profile, study findings revealed that race-related stress and PTSD symptoms were positively associated. This finding suggests that an undifferentiated racial identity does not buffer the negative effects of race-related stress on trauma symptoms and may even exacerbate that relationship. The detached profile followed a similar pattern, with overall PTSD symptoms and each symptom cluster increasing with increased race-related stress. Individuals in the detached subgroup hold attitudes that reflect some dislike for being Black and do not experience strong racial belonging, which may mean they are more likely to have poor self-esteem that leads to increased negative outcomes such as PTSD symptoms (Bhar et al., 2008; Mekawi, Lewis, et al., 2020). In fact, for members of both the undifferentiated and detached profiles, limited use of cultural coping strategies in the face of multifaceted stressors could be leading to negative mental-health outcomes such as PTSD, particularly in the context of racism-related experiences (Mekawi, Watson-Singleton, et al., 2020).
Only the nationalist profile demonstrated a notably distinctive effect on the relationship between race-related stress and PTSD symptoms. The detached and undifferentiated profiles showed significant differences from the nationalist profile in follow-up analyses of conditional effects of different profiles; the detached profile and the undifferentiated profile both had significantly steeper positive slopes than the nationalist profile in all four moderation models. To provide further detail, for individuals in the nationalist group, race-related stress was not associated with posttraumatic stress symptoms. Hence, for Black American women, a nationalist worldview, compared with the other profiles, may mitigate the negative effects of race-related stress on PTSD. This finding is even more compelling considering there were significant group differences in race-related stress found through the ANOVA, with the nationalist profile experiencing significantly higher levels of race-related stress compared with the detached and undifferentiated profiles. This finding highlights that even though individuals in the nationalist profile are reporting higher levels of race-related stress, they are also experiencing the lowest PTSD symptoms compared with the other profiles. Furthermore, despite experiencing significantly higher levels of race-related stress, holding an Afrocentric worldview (i.e., centering Black values, ideas, and identity) seemed to provide a protective barrier against the negative effects of such stress on PTSD symptoms for these women, operating as a point of resilience in the larger context of racial and traumatic stressors. There are several potential reasons why these results might have occurred. Given that the women with a nationalist profile experienced significantly higher race-related stress, this may lead to mastery of effective coping responses to manage such stress over time (Neblett & Carter, 2012). Furthermore, the awareness that society does not view Black Americans positively (i.e., low public regard) may lead individuals to develop compensatory coping strategies that counter this reality, such as denouncing racism (Neblett & Carter, 2012). Further research is necessary to confirm the mechanisms involved in these patterns because it is possible that this racial-identity development influencing this coping strategy is one of many factors contributing to the protective potential of the nationalist profile.
The results of the current study align with the existing limited research in this area. Researchers have found that dimensions of racial identity, including low centrality and low private regard, are associated with greater negative-mood symptoms, suggesting that high centrality or high private regard, which are characteristics of the nationalist profile in the current study, may be protective (Tyrell et al., 2019). Likewise, as singular dimensions, the strong racial-centrality and private-regard indicators associated with the nationalist profile have been found to buffer the negative effects of racial discrimination on mood, anxiety, and substance use disorders (Woo et al., 2019). Previous research has also indicated that among Black American individuals for whom race is not central to their identity, their mental-health outcomes are worse, especially in the context of racial discrimination (Mekawi, Lewis, et al., 2020; Neblett et al., 2004; Seaton, 2009). This may be because individuals, especially Black American women, with low centrality are found to use fewer culturally centered coping strategies or fewer effective coping practices in response to racism (Mekawi, Watson-Singleton, et al., 2020). Although the buffering potential of high nationalism-subscale scores has been considered minimally in the literature, it is possible that prioritizing Afrocentric values may add to a sense of racial belonging and racial centrality, which has been found to be protective (Bhar et al., 2008; Mekawi, Lewis, et al., 2020; Neblett et al., 2004; Seaton, 2009; Sellers & Shelton, 2003).
The current results also encourage further research on racial trauma, or the conceptualization of racism as a traumatic stressor, particularly when considering the resilience potential of racial identity (M. T. Williams et al., 2018). Although trauma and racism have conventionally existed as separate bodies of literature, research using a traumatic-stress framework that incorporates systemic experiences of oppression is substantiated by the current study (Carter, 2007; Helms et al., 2012). Given the growing body of literature supporting the association between racism and PTSD, including the current study, critical examination of how researchers conceptualize trauma is necessary. The field must consider whether the existing framework for conceptualization of trauma responses is inclusive of the experiences of minoritized individuals; whether a separate, context-informed framework is required to appreciate the traumatic effects of oppressive experiences on African Americans and other racially marginalized groups; and how researchers should be using such frameworks when considering risk and resilience factors such as racial identity.
Limitations and future directions
There are a few limitations associated with the current study. The study was limited to a single urban community in the South, possibly preventing geographical generalizations. Participants are community members seeking care in a single public-hospital system, which did not allow for data collection with a broader range of participants from the community. The mood/cognition symptom criteria group in the DSM-5 was not examined in this study because of method limitations; however, research does indicate a potential relationship between racial discrimination and posttraumatic cognitions (American Psychiatric Association, 2013; Dale & Safren, 2019). Furthermore, LPA would ideally be used with a larger sample size than the current study but nonetheless has been successfully applied to produce racial-identity profiles with similar sample sizes in other studies (Bernard et al., 2018). Future research should examine this phenomenon with other racially minoritized groups to understand whether the racial-identity profiles and resiliency effects found in this study are reflective of the unique experiences of Black American women. The cross-sectional nature of this study prevents causal conclusions, so researchers should use longitudinal designs to allow for more temporal conclusions and the application of an intergenerational frame when considering the relationship among racism, racial identity, and trauma symptoms over the life course. Application of an intergenerational framework aligns with conceptualizations of distress from racism as a unique result of the historical and intergenerational traumatic contexts that Black Americans navigate (Akbar, 1996; DeGruy, 2017; Halloran, 2019; Opara et al., 2020; Wilkins et al., 2013). Researchers should also investigate mechanisms that may influence the risk and resilience potential of different racial-identity profiles on the relationship between race-related stress and PTSD. In particular, researchers should examine culturally salient regulatory processes (e.g., culturally informed emotion-regulation strategies and shift-and-persist coping) to understand whether such processes may mediate the protective potential of certain racial-identity profiles (Christophe et al., 2021; Mekawi, Watson-Singleton, et al., 2020).
Conclusion
The current study builds on the understanding of Black American women’s mental health by examining relationships among racial identity, race-related stress, and trauma symptoms in a trauma-exposed, nontreatment-seeking sample. Given the minimal literature in this area of research, the current study offers valuable evidence of the highly connected experiences of trauma and race-related stress and the powerful potential for multidimensional profiles of racial identity to buffer or exacerbate this relationship. A concrete grasp of how racial-identity profiles may be protective has the potential to inform adaptation and development of PTSD treatments that address culturally salient factors affecting clinical presentations of trauma. In a population that is deeply overlooked yet essential to consider in the larger question of risk and resilience factors in PTSD and race-related stress, the current study provides evidence of the potent relationship among racism, trauma, and identity and the significant need for additional research in this area.
Footnotes
Acknowledgements
We thank the Grady Trauma Project team for their contributions to data collection and are particularly grateful to our participants for their contribution.
Transparency
Action Editor: Demond M. Grant
Editor: Jennifer L. Tackett
Author Contributions
