The Association Between Discrimination and PTSD in African Americans: Exploring the Role of Gender

Abstract

Background

Research has demonstrated the adverse impact that discrimination has on physical and mental health. However, few studies have examined the association between discrimination and symptoms of posttraumatic stress disorder (PTSD). There is evidence that African Americans experience higher rates of PTSD and are more likely to develop PTSD following trauma exposure than Whites, and discrimination may be one reason for this disparity.

Purpose

To examine the association between discrimination and PTSD among a cross-sectional sample largely comprising African American women, controlling for other psychosocial stressors (psychological distress, neighborhood safety, crime).

Methods

A sample of 806 participants was recruited from two low-income predominantly African American neighborhoods. Participants completed self-report measures of PTSD symptoms, perceived discrimination, perceived safety, and psychological distress. Information on neighborhood crime was obtained through data requested from the city. Multivariate linear regression models were estimated to assess adjusted relationships between PTSD symptoms and discrimination.

Results

Discrimination was significantly associated with PTSD symptoms with a small effect size, controlling for relevant sociodemographic variables. This association remained consistent after controlling for psychological distress, perceived safety, and total neighborhood crime. There was no evidence of a gender by discrimination interaction. Participants who experienced any discrimination were significantly more likely to screen positive for PTSD.

Conclusions

Discrimination may contribute to the disparate rates of PTSD experienced by African Americans. PTSD is associated with a range of negative consequences, including poorer physical health, mental health, and quality of life. These results suggest the importance of finding ways to promote resilience in this at-risk population.

Introduction

Discrimination and unfair treatment have an adverse effect on physical and mental health.¹ Studies have found associations between perceived discrimination and multiple aspects of health, including mental health disorders, self-rated health, cardiovascular health, and mortality (Williams, Neighbors, and Jackson 2003; Troxel et al. 2003; Kessler, Mickelson, and Williams 1999; Karlsen and Nazroo 2002; Guyll, Matthews, and Bromberger 2001; Barnes et al. 2008; Schulz et al. 2000; Ryan, Gee, and Laflamme 2006; Lewis et al. 2006). Of the studies focused on mental health outcomes, most have examined the impact of discrimination on depression and anxiety (Wei et al. 2012). However, few studies have examined its association with posttraumatic stress disorder (PTSD), despite the fact that PTSD is associated with poorer physical health (Pacella, Hruska, and Delahanty 2013), increased rates of comorbid mental health and substance use disorders (Kessler et al. 1995), and poorer quality of life (Olatunji, Cisler, and Tolin 2007).

Understanding the contribution of discrimination to PTSD and PTSD symptoms is important, as some research suggests that PTSD is more prevalent among African Americans (Norris and Sloane 2007; Alim, Charney, and Mellman 2006). In part, this seems to be explained in part by individual socioeconomic disadvantage (Alim, Charney, and Mellman 2006), and may also be due to the increased rates of acute trauma experienced by this population (Turner and Lloyd 2004; Turner and Avison 2003), including exposure to disadvantaged neighborhoods with high rates of crime and violence (Roberts et al. 2011; Alim, Charney, and Mellman 2006). However, there is evidence that the risk of developing PTSD among African Americans remains higher after controlling for exposure to trauma (Roberts et al. 2011), and some have argued that discrimination and race-related stress also contribute to the increased risk of PTSD in African Americans (Carter 2007b, 2007a).

Perceived discrimination is highly prevalent among African Americans. A large national survey found that nearly 50% of African Americans reported experiencing major discrimination during their lifetime, including being passed over for a job or promotion, being hassled by police, and being denied or receiving inferior medical care (Kessler, Mickelson, and Williams 1999). There are multiple pathways by which discrimination and unfair treatment may contribute to symptoms of PTSD particularly for African Americans. In part, it has been suggested that discrimination serves a chronic stressor that can create negative emotional states, lead to negative health-related behaviors, erode protective psychological resources, and result in a heightened stress response (Williams, Neighbors, and Jackson 2003; Pascoe and Richman 2009; Vines et al. 2017; Flores et al. 2008). In turn, these factors may render an individual more vulnerable to developing mental and physical health problems (Williams, Neighbors, and Jackson 2003; Carter 2007b) – for example, increasing the risk that they will develop PTSD after experiencing a traumatic event. Others have argued that discrimination itself may be a chronic trauma (Butts 2002), particularly if experienced as negative, memorable, sudden, and uncontrollable (Carter 2007b). Moreover, though some forms of discrimination may be more chronic or represent the accumulation of minor incidents, other forms may be more acute or traumatic in nature, such as hate crimes or gendered violence.

There is some empirical evidence that unfair treatment is associated with PTSD symptoms in certain minority and marginalized groups. For example, there is evidence that discrimination due to race, sexual orientation, and having HIV are associated with PTSD symptoms in HIV-positive men (Bogart et al. 2011; Wagner et al. 2012), and that microaggressions related to sexual orientation are associated with PTSD symptoms (Wei et al. 2012; Robinson and Rubin 2016). Perceived discrimination has also been shown to contribute to PTSD symptoms in Somali refugees, even after controlling for trauma and acculturative stress (Ellis et al. 2008; Mölsä et al. 2016).

There is also a small body of literature exploring the association between discrimination and PTSD in African Americans. There is evidence that perceived racial/ethnic discrimination is associated with an increased risk of PTSD in African Americans (McLaughlin, Hatzenbuehler, and Keyes 2010). In addition, a study using nationally representative data to explore the association between racial discrimination and PTSD in Asian Americans, Hispanic Americans, and African Americans found that the association between discrimination and lifetime prevalence of PTSD was stronger in African Americans than the other groups (Chou, Asnaani, and Hofmann 2012). Seng and colleagues (2012) examined the association between discrimination, structural inequalities, contextual factors, and PTSD symptoms in a sample of women expecting their first child. They found that frequency of discrimination was associated with PTSD symptoms, even after controlling for structural inequalities and contextual factors (e.g., crime rate, minority status). Although African American women were not more likely to experience discrimination than women of other racial/ethnic groups, they endorsed more PTSD symptoms on average. In addition, a study of undergraduate students found that African Americans reported more race-related events than White and Asian American participants, and that individuals who met criteria for PTSD had experienced significantly more race-related stressors (Waelde et al. 2010). This study builds on an earlier investigation of race-related stressors in undergraduate students that also found a significant association with PTSD (Khaylis, Waelde, and Bruce 2007). Although these studies provide evidence for an association between discrimination and PTSD in African Americans, there are some gaps. First, it is unclear the extent to which the studies conducted in undergraduate students or pregnant women generalize to the larger population. Second, though the study by Chou and colleagues (Chou, Asnaani, and Hofmann 2012) used nationally representative survey data, they focused on lifetime prevalence of PTSD rather than current symptoms, and it is unclear whether the association with discrimination may be different for lifetime prevalence of PTSD versus current PTSD symptoms.

The present study builds on this previous research by examining the association between discrimination and PTSD symptoms in a large sample of predominantly African American, low-income individuals in two disadvantaged neighborhoods – a population particularly at risk for health and mental health problems (Goldmann et al. 2011; Williams 1999). We hypothesized that individuals who reported experiencing more unfair treatment would have more severe PTSD symptoms. In addition, although there have been previous studies of discrimination and PTSD in African Americans, they generally have not explored the role of gender, despite the fact that African American women may experience both race-based and gender-based discrimination and there is evidence that rates of PTSD are higher in women, even following exposure to similar types of trauma (Tolin and Foa 2006). Therefore, we also examined gender as a moderator of this association.

Because participants were recruited from disadvantaged neighborhoods, we also considered the role that neighborhood characteristics may play in contributing to symptoms of PTSD. Researchers have highlighted the importance of including other psychosocial stressors related to disadvantage when examining the health impact of discrimination, including neighborhood conditions (Lewis, Cogburn, and Williams 2015). In particular, individuals in disadvantaged neighborhoods are exposed to high levels of crime and have concerns about personal safety (Goldmann et al. 2011; Krivo and Peterson 1996; Baum et al. 2009); in turn, these factors may contribute to or exacerbate PTSD symptoms. Therefore, we accounted for the potential effect of two neighborhood variables in our analyses: neighborhood crime and perceived safety. This data allowed us to determine whether these neighborhood factors contribute to PTSD symptoms, and also whether any association between discrimination and PTSD persisted after accounting for these factors.

Methods

Participants and Procedures

Data for this study came from the PHRESH Zzz Study (Pittsburgh Hill/Homewood Research on Neighborhoods, Sleep, and Health), part of a longitudinal study designed to examine the effect of changes in the built and social environment on health behaviors and risk factors in two low-income predominantly African American Pittsburgh neighborhoods in Pittsburgh, PA. Households were randomly selected in each neighborhood at baseline (2011), and were followed through three follow-up waves of data collection (2013, 2014, 2016). The present analyses are based on data from the fourth wave (2016), when measures of discrimination and PTSD were added to the survey. Details of the study methods, including recruitment and data collection procedures, are described in depth elsewhere (Dubowitz, Ghosh-Dastidar, et al. 2015; Dubowitz, Ncube, et al. 2015).

A total of 828 participants took part in data collection efforts, which included participation in an in-person interview. Measures of neighborhood crime were obtained via data obtained from the City of Pittsburgh Police Department. For the present analyses, we excluded respondents for whom there was missing data on either the outcome (PTSD symptoms (n=5)) or key predictors (perceived neighborhood safety (n=1), total reported crime (n=7), unfair treatment (n=9)), resulting in a sample size of 806 for all analyses.

Measures

Posttraumatic stress disorder symptoms

Stressful life events were measured using the 6-item PTSD Checklist (PCL-6) (Lang and Stein 2005). Respondents were asked how much they had been bothered by each of the following in the past month: 1) Repeated, disturbing memories, thoughts, or images of a stressful experience from the past, 2) Feeling very upset when something reminded them of a stressful experience from the past, 3) Avoiding activities or situations because they reminded them of a stressful experience from the past, 4) Feeling distant or cut off from other people, 5) Feeling irritable or having angry outbursts, and 6) Difficulty concentrating. Response options for each item ranged from 1 (Not at all) to 5 (Extremely). The scale is scored by summing the responses on each item, with possible scores ranging from 6 to 30. Higher scores on this scale reflect greater severity PTSD symptoms, and a score of 14 or higher is considered a positive screen for PTSD (sensitivity =0.92-0.98, diagnostic efficiency = 0.71-0.75) (Han et al. 2016; Lang and Stein 2005) (alpha in current sample = 0.85).

Unfair treatment

Unfair treatment was measured using the Everyday Discrimination Scale, Short Version (Sternthal, Slopen, and Williams 2011). Respondents were asked how often they had each of five experiences in their day-to-day life: 1) Being treated with less courtesy or respect than other people, 2) Receiving poorer service than other people at restaurants or stores, 3) Being treated like they are not smart, 4) Being treated like other people are afraid of them, and 5) Being threatened or harassed. Response options for each item ranged from 0 (Never) to 5 (Almost every day). Scores reflect the sum of responses and range from 0 to 25, with higher scores reflecting more frequent experience of unfair treatment. Additionally, we calculated a dichotomous indicator to identify those participants who endorsed any of these experiences “a few times a year” or more frequently (hereon referred to as “any discrimination”) (alpha = 0.71).

Psychological distress

The Kessler 6 (K6) is a well-validated self-report instrument including six items that measure psychological distress (Kessler et al. 2002). Respondents were asked how often in the past 30 days they felt 1) nervous, 2) hopeless, 3) restless or fidgety, 4) so depressed that nothing could cheer them up, 5) that everything was an effort, and 6) worthless. Response options for each item ranged from 0 (None of the time) to 4 (All of the time). The measure is scored by summing responses on each item, and higher scores indicate higher levels of psychological distress (alpha = 0.85).

Perceived neighborhood safety

Perceived neighborhood safety was assessed using four items (e.g., “You feel safe walking in your neighborhood during the day,” “Your neighborhood is safe from crime”) (DeSantis et al. 2016). Response options for each item ranged from 1 (Strongly disagree) to 5 (Strongly agree). Items were reverse coded as necessary, and the composite was formed as the mean of responses across items. Higher scores indicate higher perceived neighborhood safety (alpha = 0.69).

Neighborhood crime

We obtained incident-level data on crimes reported to the Pittsburgh Police Department and calculated street network distances from each household in the study to each approximate crime scene using ArcGIS 10.2 software. We calculated the total number of violent and property crimes that occurred within 1/10^th mile of each participant’s home in the month that they were interviewed. Given the high correlations between property and violent crimes (r = 0.50), we created a total reported crime variable that reflects the count of all crimes (including both property and violent crimes).

Sociodemographic characteristics

Participants provided information on age, gender, annual household income, presence of children in the household (recoded as any versus none), and marital status (recoded as married or living with a partner versus all other). We also included an indicator variable for neighborhood.

Statistical Analysis

First, we calculated descriptive statistics of all measures described above for the sample, and examined differences between those with and without a positive screen for PTSD. Significant differences between the two groups were identified using t-tests for continuous predictors and chi-square tests for categorical predictors.

Next, multivariate linear models were estimated in order to assess adjusted relationships between PTSD symptoms and unfair treatment. We estimated a multivariate linear model of PCL-6 on the Everyday Discrimination Scale, K6, controlling for sociodemographic characteristics described above (model 1). We then estimated additional models that included either perceived neighborhood safety (model 2) or total crime (model 3) as predictors, and also both predictors entered simultaneously (model 4). The additional models allowed us to assess the degree to which the estimated relationship between PTSD symptoms and unfair treatment was robust to the inclusion of neighborhood safety and crime. For continuous predictors, we reported standardized regression coefficients, which can be interpreted as effect sizes. Coefficients for categorical predictors can be interpreted as the association between the predictor and the outcome, measured in standard deviations of the outcome. As cited in the literature, effect sizes of 0.20 are considered small, effects of 0.50 are considered moderate, and effects of 0.80 are considered large (Cohen 1988). Because the outcome (PCL-6 scores) are highly positively skewed, we ran two sets of sensitivity analyses for all models: 1) linear regression using log-transformed PCL-6 as the outcome and 2) logistic regression using positive screen for PTSD as the outcome. In addition, to explore whether there were gender differences in the association between unfair treatment and PTSD, we examined a gender by discrimination interaction term in model 1 (with basic covariates).

Finally, we calculated the probability of having PTSD for those experiencing any discrimination compared to those who endorsed no discrimination using the logistic model, and adjusting for psychological distress and the sociodemographic variables described above (Graubard and Korn 1999).

Analyses were performed in SAS software, version 9.4, of the SAS System for Windows (SAS Institute, Inc., Cary NC). An α of 0.05 or less was used to determine significance. Data were cleaned and analyzed from January to July 2017.

Results

The mean age of participants was 58.0 years (SD = 15.4), and the majority were women (79.0%). Most participants (95.4%) were Black or African American; 2.0% were White, 0.4% were American Indian/Alaska Native, and 2.2% identified as “other.” The mean household income was $20,800 (SD = $18,000). For complete demographic information, see Table 1.

Table 1.

Demographic characteristics and bivariate associations with PTSD screening

	Overall (n=806) M(SD) or %	No PTSD (n=596) M(SD) or %	PTSD (n=210) M(SD) or %
PTSD Symptoms
PCL-6 Score	11.0 (4.9) range = 6-30	n/a	n/a
Positive PTSD screen	26.1%	n/a	n/a
Sociodemographics
Age, years***	58.0 (15.4)	59.8 (15.3)	52.8 (14.4)
Male	21.0%	22.0%	18.1%
Annual household income, thousands of dollars***	20.8 (18.0)	22.1 (18.2)	17.2 (17.0)
Any children in household***	24.7%	21.3%	34.3%
Married or living with partner	16.5%	17.6%	13.3%
Black or African American	95.4%	96.0%	93.8%
Discrimination
Unfair treatment***	4.4 (4.6)	3.3 (3.8)	7.3 (5.2)
Any discrimination***	58.6%	50.8%	80.5%
Covariates
Psychological distress***	4.3 (4.6)	2.9 (3.4)	8.4 (5.2)
Perceived neighborhood safety*	3.1 (0.7)	3.1 (0.7)	3.0 (0.8)
Neighborhood crime (# crimes)	2.0 (2.2)	2.0 (2.2)	2.2 (2.3)

^*p < .05,

^**p < .01,

^***p < .001

The mean score on unfair treatment scale was 4.4 (SD = 4.6), and 58.6% reported experiencing at least one type of discrimination at least a few times a year. Approximately 63% of men reported experiencing any discrimination (M = 5.3, SD = 5.3), and approximately 58% of women reported experiencing discrimination (M = 4.1, SD = 4.3). Within the sample, the mean PCL-6 score was 11.0 (SD = 4.9) (Table 1) with 25^th, 50^th, and 75^th percentiles at 7, 10 and 14, respectively. Also, 26% percent of the sample screened positive for PTSD. Participants screening positive for PTSD (i.e., scores of 14 and above) were, on average, younger, had lower income, and were more likely to have a child living in their household. Participants who screened positive for PTSD had a significantly higher mean score on the unfair treatment scale and were more likely to report experiencing discrimination at least a few times a year. On average, those who screened positive for PTSD reported higher levels of psychological distress and slightly lower perceived neighborhood safety.

Table 2 summarizes results from the multivariate linear regression models. In model 1, unfair treatment was a significant predictor of PTSD symptoms with a small effect size (beta = 0.23). Psychological distress was also strongly associated with PTSD symptoms. Collectively, this model accounted for 48% of the variance in PTSD symptoms.

Table 2.

Multivariate linear regression models predicting PTSD symptoms

	Model 1 beta (SE)	Model 2 beta (SE)	Model 3 beta (SE)	Model 4 beta (SE)
Unfair treatment	0.23 (0.03) ***	0.22 (0.03) ***	0.23 (0.03) ***	0.22 (0.03) ***
Psychological distress	0.54 (0.03) ***	0.54 (0.03) ***	0.54 (0.03) ***	0.54 (0.03) ***
Perceived neighborhood safety		−0.03 (0.03)		−0.03 (0.03)
Neighborhood crime			0.02 (0.03)	0.02 (0.03)
Age, years	−0.05 (0.03)	−0.05 (0.03)	−0.05 (0.03)	−0.05 (0.03)
Male	−0.12 (0.07)	−0.12 (0.07)	−0.12 (0.07)	−0.12 (0.07)
Annual household income, thousands of dollars	−0.03 (0.03)	−0.03 (0.03)	−0.03 (0.03)	−0.03 (0.03)
Any children in household	0.12 (0.07)	0.12 (0.07)	0.12 (0.07)	0.12 (0.07)
Married or living with partner	−0.22 (0.07) **	−0.21 (0.07) **	−0.21 (0.07) **	−0.21 (0.07) **
Neighborhood
Hill (ref)
Homewood	0.01 (0.06)	−0.01 (0.06)	0.01 (0.06)	−0.01 (0.06)
Other	0.05 (0.09)	0.05 (0.09)	0.06 (0.09)	0.06 (0.09)
Intercept	0.03 (0.04)	0.03 (0.04)	0.02 (0.04)	0.03 (0.04)
Adjusted R2	0.48	0.48	0.48	0.48

^*p < .05,

^**p < .01,

^***p < .001

Coefficients for continuous predictors have been standardized and can be interpreted as effect sizes.

Coefficients for categorical predictors (Male, Any children in household, Married or living with partner, Neighborhood) have not been standardized. They can be interpreted as the association between the predictor and the outcome, measured in standard deviations of the outcome.

Perceived neighborhood safety (model 2), total reported crime (model 3), and both variables together (model 4) were not significantly associated with PTSD symptoms, and their addition to the model did not improve the proportion of variance explained. Moreover, the estimated relationship between unfair treatment and PTSD symptoms was similar across all four models, indicating that neither perceived neighborhood safety nor total reported crime is accounting for this relationship. There was also no evidence of a gender by unfair treatment interaction effect when the interaction term was added to model 1 (beta = 0.03, SE = 0.04, p > .05).

In Figure 1, we present the probability of screening positive for PTSD when exposed to discrimination (dichotomized), adjusted for psychological distress and demographic variables. Being exposed to any discrimination is associated with almost a 10-percentage point higher adjusted probability of screening positive for PTSD (adjusted probability of 20.0% versus 29.1%, p < .01).

Figure 1.

Adjusted probability of screening positive for PTSD by discrimination

Note: Model adjusted for age, gender, annual household income, any children in household, marital status, and neighborhood

Discussion

This study examined the association between unfair treatment and PTSD symptoms in a sample of predominantly African American, low-income individuals. Approximately 26% of participants screened positive for PTSD. This number is substantially higher than prevalence rates of PTSD in the general population (approximately 8%) (Kilpatrick et al. 2013). However, this finding is consistent with previous research demonstrating that African Americans have higher rates of PTSD (Roberts et al. 2011). Low-income and urban African Americans appear to have particularly high lifetime rates of PTSD, with studies reporting prevalence rates ranging from 17% in the general population to as high as 51% in treatment seeking populations (Goldmann et al. 2011; Alim et al. 2006). There are a number of factors that may put African American populations at higher risk for PTSD, including the experience of acute traumas (Johns et al. 2012). However, chronic stressors, such as unfair treatment, are also pervasive. In the present sample, which is predominantly low-income, nearly 59% of participants endorsed experiencing discrimination in the past year during their day-to-day life.

Our findings indicated that individuals who experienced more unfair treatment had significantly higher PTSD symptoms. This association persisted even after controlling for psychological distress, suggesting that this finding does not simply reflect the known relationship between distress and PTSD (Marshall et al. 2001; Brady et al. 2000), nor does it reflect a general negative affect bias. Moreover, the association remained significant after controlling for individual sociodemographics as well as neighborhood-level characteristics, including perceived neighborhood safety and objective reports of crime – experiences that may serve as acute or chronic traumatic experiences and that have been shown to be associated with PTSD in disadvantaged populations (Gapen et al. 2011).

These results are consistent with previous research that has found that discrimination has a unique association with PTSD symptoms beyond the impact of general life stress (Pieterse et al. 2010; Wei et al. 2012). In addition, although we did not have information about past trauma history, previous research has found that discrimination contributes to PTSD symptoms beyond the impact of trauma alone. In fact, previous research has found that discrimination may account for as much as 20% of the variance in PTSD symptoms among trauma-exposed populations (Ellis et al. 2008). Together, these findings suggest that discrimination may contribute to the disparate rates of PTSD experienced by low-income African Americans.

Understanding the association between discrimination and PTSD symptoms in low-income African American populations has important implications for future health. Discrimination is a chronic and ongoing stressor for many individuals. When individuals who already have heightened trauma-related symptoms continue to experience discrimination, it may have an adverse impact on health. For example, researchers have suggested that ongoing discrimination can serve as a reminder of prior trauma, and may further undermine psychological resilience and coping abilities (Wagner et al. 2012). In turn, this may cause individuals to be vulnerable to future mental health or physical health problems. For example, one study found that discrimination mediated the association between PTSD and adherence to HIV medication (Wagner et al. 2012). Therefore, discrimination may play a role in the development of PTSD, but may also contribute to ongoing health issues in trauma-exposed populations.

This study has certain limitations. First, the measures we focused on in this study were part of a longer battery of self-report measures. For this reason, we selected short form instruments to reduce participant burden and distress. As a result, we were unable to include a more comprehensive PTSD instrument, and do not have an item assessing a specific traumatic event in relationship to these symptoms (i.e., a Criterion A trauma, per DSM-5 diagnostic criteria). Therefore, we are not able to identify individuals who met formal diagnostic criteria for PTSD, only those who screened positive for PTSD. Although we included measures of perceived safety and neighborhood crime in our models to account for threats to an individual’s safety, these are an imperfect proxy for being the victim or target of a traumatic event, which may explain their non-significant association with PTSD symptoms in this sample. That said, even if participants in this study did not meet formal criteria for PTSD, symptoms of PTSD, including avoidance, re-experiencing, and irritability, can have a significant impact on an individual’s day-to-day wellbeing. Therefore, these findings still provide valuable information about the factors that contribute to distress in this population. In addition, at the time that measures were being selected for the present study, there was not a short-form PTSD measure validated against DSM-5 criteria. That said, the PCL-6 was recently validated against the Clinician Administered PTSD Scale (CAPS) for DSM-IV in an underserved, diverse sample of patients seeking care from federal qualified health centers [72], and there is evidence for similar rates of PTSD when using DSM-IV relative to DSM-5 criteria (Kilpatrick et al. 2013; Carmassi et al. 2013).

Though commonly used in the literature, there are also certain limitations associated with our use of the Everyday Discrimination Scale, Short Version. In part, as described above, our selection of a short-form measure was driven by a need to limit participant burden. However, these questions capture broadly-defined unfair treatment, as we did not ask participants to indicate the perceived reason for unfair treatment (e.g., whether it was due to their race/ethnicity). There is evidence that different forms of discrimination may have a similar impact on health (Lewis, Cogburn, and Williams 2015); however, we cannot comment on the impact of racial/ethnic discrimination versus other types of unfair treatment (e.g., gender, socioeconomic status) in this study, nor describe the intersection of these types of unfair treatment. In addition, this measure is designed to measure unfair treatment that is “chronic or episodic but generally minor” (Williams and Mohammed 2009); therefore, it will be important for future research to further explore the contribution of this type of discrimination versus more acute or major experiences of discrimination to PTSD (Lewis, Cogburn, and Williams 2015). Finally, this measure asks participants to consider how often each event happened, but does not initially specify the sources of unfair treatment (e.g., gender and age and race). The level of specificity of survey questions has the potential to affect responses (Schaeffer and Presser 2003); therefore, there may be more variability in responding than there would be on a measure with more specific, well-defined questions.

Regarding the role of gender, 79% of participants in the sample were women, which may have limited our power to detect a significant interaction effect. However, the magnitude of the regression coefficient for this analysis suggests that there is not a substantial difference in the association of discrimination and PTSD for men and women. In addition, the mean age of participants was 58.0 years; therefore, it is unclear the extent to which these results might generalize to younger populations. It would also be interesting to explore whether these findings generalize to other neighborhood contexts, given evidence that African Americans from higher SES backgrounds and those living in primarily non-Black neighborhoods experience more discrimination (Lewis, Cogburn, and Williams 2015; Hunt et al. 2007). Finally, the present analyses are based on cross-sectional data, which precludes inferences regarding causality. It will be critical for future research to explore the longitudinal impact of discrimination on PTSD symptoms.

These findings provide the foundation for future research examining unfair treatment as a chronic stressor that contributes to trauma symptoms. An important next step would be to explore the contribution of discrimination to PTSD beyond the impact of specific traumatic events. In addition, it would be worthwhile to examine the effect that discrimination and PTSD have on subsequent health in this population. Regarding clinical implications, Carter (Carter 2007b) cautioned against pathologizing the psychological impact of racism and discrimination on marginalized populations. Although interventions aimed at reducing individual and institutionalized discrimination would be the ideal approach to mitigating this problem, more immediate individual- and community-level efforts may help to reduce the adverse impact of this discrimination in the short-term. For example, there is research that suggests that there are factors that can buffer the impact of discrimination on health, such as social support and neighborhood cohesion (Chou 2012; Pascoe and Richman 2009). In addition, McLaughlin and colleagues found that an individual’s response to discrimination may impact the effect of discrimination on health: African American women who did not accept and did not disclose discrimination, or who accepted and disclosed discrimination, were less likely to have PTSD than those who did not accept and disclosed. Therefore, to the extent that it is possible to promote buffering factors or protective response styles in at-risk populations, it may be possible to reduce their vulnerability to PTSD.

This study contributes to a burgeoning literature establishing an association between unfair discrimination and PTSD symptoms in African Americans. Together, these studies shed light on potential mechanisms of health disparities in an at-risk population. Although the next step is to explore the temporal association of these effects with longitudinal studies, this research will ultimately be key to the development of prevention and intervention efforts.