Ethnic Disparities in Trauma-Related Mental Illness: Is Ethnic Identity a Buffer?

Abstract

Despite evidence that racial and ethnic characteristics influence the impact of traumatic exposure on psychological health, little is known about how race and ethnic identity can alter, and possibly protect against, the effects of trauma on the psychiatric diagnoses of women. Therefore, the present study examined the moderating role of race/ethnicity and ethnic identity in the link between trauma exposure and psychiatric diagnosis for African American and Caucasian college women. Participants were a sample of 242 women from the Mid-Atlantic region of the United States who self-identified as African American or Black (31%) and European American or Caucasian (69%; M age = 19.5 years). Interviews were conducted over the phone to screen for trauma, followed by longer in-person interviews. Each of the interviewers was supervised, and interviews were reviewed to control for quality. Regression analyses revealed that the number of traumatic events was a stronger predictor of lifetime psychiatric diagnoses for Caucasian women. In addition, ethnic identity served as a protective factor against trauma exposure among participants. The findings suggest that ethnic identity is a relevant buffer against potential psychiatric diagnoses as result of exposure to traumatic events for both Caucasian and African American women.

Keywords

ethnic disparities trauma exposure protective factors ethnic identity African American women

Exposure to traumatic events, characterized by extreme stress (Aldwin, 2007), has been said to be an inevitable part of life (van der Kolk, McFarlane, & Weisaeth, 1996). Epidemiological studies consistently find that more than 50% and up to 90% of the U.S. population has been exposed to some type of trauma (Kessler, Sonnega, Bromet, Hughes, & Nelson, 1995; Kilpatrick et al., 2013). For African American women, the rates of interpersonal trauma are particularly high (Bent-Goodley, 2004). Studies show that African American women are more likely to experience severe abuse (El-Khoury et al., 2004), domestic violence (Rennison & Welchans, 2000), forcible rape (Kilpatrick, Resnick, Ruggiero, Conoscenti, & McCauley, 2007), and revictimization (Carlson, Harris, & Holden, 1999) than are Caucasian or Latina women. African American women must also contend with multiple systems of oppression (racism, sexism, classism, etc.) that intersect to increase overall stress and trauma (Townsend, Batts, & Hawkins, 2007). Certainly, racial oppression in U.S. law enforcement is not a new occurrence; however, recent media coverage of the mass incarceration of Black and Brown people in the United States and the tragic killing of unarmed African Americans at the hands of police has contributed to a heightened experience of alienation, stress, and trauma among African American populations (Westcott, 2015).

Although not every trauma victim responds to trauma exposure in the same way, the effects of trauma can be debilitating, often including the experience of depression, posttraumatic stress disorder (PTSD), anxiety, dissociation, and substance abuse (Kilpatrick et al., 2003), as well as suicidal ideation and low self-esteem (Thompson et al., 2000). Given that African American girls and women are more likely to be victims of interpersonal violence and to experience societal trauma, they may be at a greater risk to develop mental disorders associated with traumatic exposure. According to Alim et al. (2006), 35% of African Americans with trauma histories have suffered from a major depressive disorder during their lifetime, while 51% reported experiencing symptoms of PTSD. There also appears to be racial/ethnic disparities in the risk of PTSD following other traumatic events. Perilla, Norris, and Lavizzo (2002), for example, observed that 38% of African Americans suffered from PTSD after a natural disaster versus 21% of their Caucasian counterparts. Overall, Roberts, Gilman, Breslau, Breslau, and Koenen (2011) found that lifetime prevalence of PTSD was highest among African Americans. In addition, African American adolescent girls are 2 to 3 times more likely to develop PTSD following a traumatic event than are African American boys (Afifi, 2007).

Why is there a racial/ethnic disparity in the risk for PTSD? Scholars suggest that the answer to this question may lie in how people react to and interpret their traumatic experiences (Graves, Kaslow, & Frabutt, 2010; van der Kolk et al., 1996). Specifically, traumatic events are explained and given meaning based on the sociocultural context in which the person and the events are embedded (van der Kolk et al., 1996). Hence, culture and context play major roles in how a trauma survivor copes with and responds to his or her traumatic experiences (Aldwin, 2007; Bryant-Davis, Belcourt-Dittloff, Chung, & Tillman, 2009; Perilla et al., 2002).

Trauma in Context: Why African American Women May Have More Severe Reactions to Trauma

For African American women, their trauma experience must be viewed within the larger historical and sociopolitical context of racism, discrimination, political oppression, unemployment, and enduring economic inequities (Waelde et al., 2010). For example, socioeconomic status (SES), which is often defined as a combination of education, income, and occupation, accounts for a large percentage of the variance in health disparities, documented in the literature among populations of color (Wiltshire, Person, Kiefe, & Allison, 2009). Specifically, African Americans are disproportionally concentrated in lower income communities where resources are limited (Jones, 2007). These poorly resourced environments often expose African American women to more severe traumatic events (Aldwin, 2007; Graves et al., 2010; Norris, 1992; Perilla et al., 2002), which increase their vulnerability to trauma-related mental illnesses. African Americans’ concentration in a lower socioeconomic stratum also makes it difficult to disentangle the influence of ethnicity from the effects of SES on trauma response. In fact, when SES was held constant in some studies, the higher prevalence of trauma/abuse exposure and certain mental health disorders among African American women were reduced or even eliminated (Rennison & Planty, 2003).

Given that culture plays a critical role in shaping the way an individual understands and manages exposure to traumatic events, it follows that it would also be important to consider cultural variables when examining African American women’s response to trauma (Perilla et al., 2002). In particular, ethnic identity has been found to serve a protective function against the experience of racism and discrimination (Neblett, Shelton, & Sellers, 2004) and, therefore, may serve a similar role against the effects of other forms of trauma, such as interpersonal violence, and abuse, particularly for African American women.

Shielding Against Interpersonal Trauma and Abuse: The Protective Role of Ethnic Identity

Identifying with one’s race/ethnicity has been widely recognized as a major normative task for adolescents and young adults of color (Seaton, Yip, Morgan-Lopez, & Sellers, 2012; Whaley, 2003). Phinney (1996) defined ethnic identity as a sense of belonging, pride, and attachment to ethnic group membership. As African Americans are identified and frequently marginalized in this country as racially and ethnically different, Whaley (2003) suggested that African American youth are forced to at least acknowledge their racial/ethnic group membership and incorporate it into their identity. Accordingly, ethnic identity was found to be a crucial domain in the self-concept and self-definition of African American girls and women (Townsend, 2002).

Although few studies have directly examined ethnic identity and trauma exposure, there is evidence that ethnic identity is associated with increased psychosocial functioning among African Americans with a stronger ethnic identity. For instance, ethnic identity has been found to serve as a psychological buffer against environmental stressors (Phinney, 1996), and a strong ethnic identity has been linked to positive self-concept and pro-social behaviors among African American adolescents and young adults (Thomas, Townsend, & Belgrave, 2003; Townsend, 2002). In one of the few studies that examined the protective function of ethnic identity in the context of trauma exposure among adolescents, Bruce and Waelde (2008) found that ethnic identity moderated the relationship between posttraumatic stress symptoms and delinquency, such that greater levels of ethnic identity were related to less delinquency in the face of increasing levels of posttraumatic stress symptoms. This moderating relationship was strongest among the ethnic minority adolescents in the sample relative to their Caucasian counterparts.

A model put forth to explain the buffering effects of ethnic identity suggests that adolescents and young adults of color who have examined the meaning of their ethnicity and developed a positive connection to their ethnic group are able to accept themselves more fully and are less likely to internalize negative self-perceptions, which could result from a negative or hostile environment that is intolerant of racial differences (Martinez & Dukes, 1997). Thus, a positive, healthy identity provides a strong internal foundation, reducing the risk that external factors (e.g., discrimination, history of interpersonal violence/trauma, low access to resources, etc.) will negatively affect or even damage a young woman’s sense of self and emotional well-being. As negative self-concept and damaged self-esteem have been associated with some forms of trauma exposure (Karyl et al., 1995), and there is a positive association between ethnic identity and self-esteem (Thomas et al., 2003), it seems intuitive that ethnic identity could serve as a buffer between the experience of trauma, the development of psychological distress, and ultimately the occurrence of psychiatric disorders among African American girls and women.

Examining Ethnic Disparities in Trauma Exposure and Psychiatric Diagnoses

The current study explored trauma exposure and psychiatric diagnoses among a sample of African American and Caucasian college women in the Mid-Atlantic region of the United States. Due to documented disparities in exposure and psychological reaction to trauma experienced by African American girls and women when compared with their Caucasian counterparts (Hatch & Dohrenwend, 2007), trauma exposure and psychiatric diagnoses were compared based on race. In addition, the buffering effects of ethnic identity were explored. Based on the aforementioned literature, it was expected that the African American women in the sample would report significantly more psychiatric diagnoses in response to trauma exposure than the Caucasian women. However, ethnic identity was expected to moderate this relationship, such that those African American women reporting a stronger connection and identification with their ethnic group were expected to report fewer psychiatric diagnoses, even in the context of high trauma exposure. This was not expected for the Caucasian women in the sample.

As previously mentioned, confounding of race/ethnicity with social class in many studies makes it difficult to separate the influence of ethnicity from the effects of SES. The current study attempted to address this limitation by including two strong indicators of SES (i.e., parental education and parental occupation) as covariates in the analyses. Additional factors known to influence response to trauma, such as age, and parental psychological history (i.e., parental history of substance abuse, alcoholism, and psychiatric hospitalization) were also included in the analysis as covariates to examine the unique effects of ethnicity and ethnic identity.

Method

Participants

The present study focused on the African American and Caucasian subsamples of a larger study of 363 college women. The participants were recruited from six colleges or universities in the Washington, D.C., area over a 2-year period. Inclusion criteria of the participants included an age of 24 years or younger and current enrollment in at least 9 credit hours during that semester at the college or university. The larger study was designed to investigate the outcomes related to exposure to different types of traumatic histories. The self-identified ethnicity distribution of the 363 women interviewed for the larger study was 58% Caucasian, 25% African American, 7% Asian American, 3% Latino, and 7% Other or bi/multiracial. Due to the limited representation of other racial/ethnic groups in the larger sample, the present study focused on the European American/Caucasian and African American/Black women.

Twenty-two participants (20 Caucasian and two African American) were omitted because they did not complete the ethnic identity scale. In addition, 40 participants (11% of original sample) were excluded because their response to the SES questions could not be coded. Due to violations in assumptions for independent t tests, chi-square and Mann–Whitney U tests were conducted to determine if there were statistically significant differences in demographic characteristics between those participants who were included in the study versus those who were excluded. There was a statistically significant difference in ethnicity for those who were excluded from the sample for not completing the ethnic identity measure; excluded participants tended to be Caucasian, χ²(1, 310) = 4.34, p < .05. However, no other statistically significant differences were found between the two groups.

The resulting sample consisted of 242 college women of which 76 (31%) self-identified as African American or Black and 166 (69%) as European American or Caucasian. The mean age of the participants was 19.5 years (range = 17-24), and 98% of the participants were single. Due to their status as college students, their education and occupational attainment had not stabilized or been well established. Therefore, in accordance with recommendations of Diemer, Mistry, Wadsworth, López, and Reimers (2013), parental education and occupation were used to define SES. For the most part, participants reported that their parents had at least a college education and were professionally employed. Furthermore, a majority of participants reported living in a two-parent household and most denied having a parent with a history of emotional or psychological illness that required hospitalization. As the study was concerned with the influence of race/ethnicity on response to trauma, descriptive information on the sample was computed by race/ethnicity and is provided in Table 1. Differences in SES indicators that emerged as a function of race/ethnicity are described in more detail in the “Results” section.

Table 1.

Percentages, Means, and Standard Deviations Among Study Variables by Race.

	African American %	Caucasian %
Age
M (SD)	19.67 (1.15)	19.37 (0.79)
17	1.3	0.0
18	5.3	6.1
19	44.7	60.1
20	35.5	27.0
21^†	13.2	6.7
Marital status
Single	98.7	96.9
Living together/married	1.3	1.8
Divorced	0.0	1.2
Parent marital status
Married	65.3	78.5
Separated/divorced	34.7	20.9
Father’s education***
High school/high school graduate	42.0	15.9
College/college graduate	35.6	43
Postgraduate or higher	22.4	41.1
Mother’s education***
High school/high school graduate	32.9	12.3
College/college graduate	42.1	48.5
Postgraduate or higher	25.0	39.3
Parental psychiatric hospitalization
None	97.4	93.9
Yes	2.6	6.1
Parental alcoholism
None	81.6	82.2
Yes	18.4	17.8
Parental substance abuse**
Yes	11.8	4.9
None	88.2	94.5
Variable	M (SD)	M (SD)
Trauma exposure	1.86 (1.80)	2.13 (1.60)
Ethnic identity***	3.48 (0.44)	2.73 (0.56)
Psychiatric diagnoses**	0.73 (1.03)	1.40 (1.53)

†

p < .10. *p ≤ .05. **p ≤ .01. ***p ≤ .001.

Procedure

The study procedures can be found more detailed in Green and colleagues (2001) and Krupnick et al. (2004). College women at six universities or colleges in the Washington, D.C., area, aged 24 or younger, taking at least 9 credit hours during the school semester were mailed packets over a 2-year period which included self-report questionnaires, a thorough explanation of the study, confidentiality assurance, and a self-addressed postage paid envelope. A total of 10,722 questionnaires were mailed, and 2,568 (24%) were returned. Participants were told they would receive a monetary incentive for participation in the interview phase of the study and a separate consent form was provided upon which the participants could provide their name and telephone number if they were interested in participating in this second stage. Of the questionnaires returned, 65% of these participants granted permission to be contacted for the interview stage. That telephone interview screened further for relevant details of trauma exposure to help with classification into the appropriate trauma groups. Female clinical psychology graduate students interviewed roughly 700 women. For quality control, the first 52 interviews were observed by an investigator. For subsequent interviews, experienced Structured Clinical Interview for DSM-IV–Nonpatient version (SCID) interviewers observed recently trained interviewers, making sure that the instrument was being administered properly. Supervision was provided for all interviewers by an investigator. Based on responses to the trauma-focused telephone interview, 363 women constituted the final interview sample. Participants signed a second consent for the interview, which was administered in person. Institutional review boards (IRBs) at each institution approved the study. The interview was approximately 2.5 to 3 hr, and the participants were given US$25 for their time.

Measures

Demographics Questionnaire

Background information on each participant was gathered through a demographic questionnaire designed by Green and colleagues (Krupnick, Green, Stockton, Goodman, Corcoran, & Perry, 2004). This instrument was used to collect basic demographic information on each of the participants and to determine if each participant met the inclusion criteria previously discussed. The participants were asked about their age, marital status, race/ethnicity, and parental psychological history (i.e., parental history of substance abuse, alcoholism, and psychiatric hospitalization), in addition to indicators of SES, including parental education and parental employment information.

Parental education was assessed by asking participants to indicate the highest level of formal education completed by their father and mother, respectively. On each item, participants responded to a 6-point ordinal scale, which was rank ordered based on a U.S. schooling structure, ranging from 1 = some high school to 6 = postgraduate degree. Higher scores represent more formal education.

Similarly, parental occupation was assessed by two items that asked participants to identify their parents’ occupation. Mothers’ and fathers’ occupation were measured on a 9-point ordinal scale ranging from 1 = professional (e.g., chief executive officer, dentist, physician, lawyer, scientist, college professor, clergyman, etc.) to 9 = unemployed. Lower scores on the occupation items represent more prestigious occupations that are higher skilled and require more specialized education/training. Rank orders roughly aligned with the Nakao and Treas’s (1994) prestige scores.

Stressful Life Events Screening Questionnaire (SLESQ)

Trauma exposure was assessed using the SLESQ (Goodman, Corcoran, Turner, Yuan, & Green, 1998). The SLESQ surveys participants on 14 events intended to assess exposure to trauma and traumatic loss. The purpose of the SLESQ is to screen for a history of stressor events that are expected to be associated with PTSD as outlined in Criterion A of the Diagnostic and Statistical Manual of Mental Disorders (4th ed.; DSM-IV; American Psychiatric Association, 1994) for PTSD. Some examples of the events included a life-threatening illness, loss of a family member, physical assault, life-threatening accident, and so on. If the event was not experienced, the participant checked “No.” The SLESQ can be scored by adding the number of different types of traumatic events reported by the participants and computing a sum score of event types, with a high score indicating a high number of reported traumatic events experienced. In a similar college sample, reliability and validity of the SLESQ were evaluated; a median kappa of .73 (range = .31-1.00) was found for test–retest reliability of the questionnaire and a median kappa of .64 (range = .26-.90) was found for the validity between the questionnaire and a face-to-face interview.

Multigroup Ethnic Identity Measure (MEIM)

Ethnic identity was assessed using the MEIM (Phinney, 1992). The MEIM is a 20-item self-report measure. After an initial explanation of the concept of ethnicity, along with examples, the respondents are asked “In terms of ethnic identity, I consider myself to be _____.” This is followed by specific questions on four dimensions: affirmation and belonging (e.g., I have a strong sense of belonging to my own ethnic group), ethnic identity achievement (e.g., I have a clear sense of my ethnic background and what it means for me), ethnic behaviors (e.g., I participate in cultural practices of my own group, such as special food, music, and customs), and other group orientation (e.g., I am involved in activities with people from other ethnic groups). These items are rated on a scale from 1 (strongly disagree) to 4 (strongly agree). Some items are reversed scored. Based on recommendations from Phinney (1992), the composite score is computed based on the first three subscales (Phinney, 1992). A higher score indicates stronger identification. The composite score was used in this study. Cronbach’s alpha for the composite score using this sample was .83.

SCID

Psychiatric diagnoses were assessed using the SCID (First, Spitzer, Gibbon, & Williams, 1996). The SCID, based on DSM-IV, is a structured interview, which determines lifetime and current Axis I diagnosis. Disorders involving somatization, anxiety, substance abuse, eating disorders, and mood disorders are all assessed in the SCID. For this study, a sum score was calculated by adding each lifetime diagnosis reported to obtain the total number of lifetime diagnoses for each participant, and the highest possible score, indicating a higher number of clinical diagnoses received in a lifetime, is 16. Reliability for the rating of diagnoses was conducted on a similar sample; kappas for the reliability of the ratings for lifetime disorders were .81 for acute stress disorder, 1.00 for PTSD, .92 for major depressive disorder, .78 for alcohol abuse, .79 for alcohol dependence, and .48 to 1.00 for the remainder of the diagnoses. The psychometric properties of the measure were deemed appropriate for this sample. Although the DSM has been revised (Diagnostic and Statistical Manual of Mental Disorders [5th ed.; DSM-5; American Psychiatric Association, 2013]) and the diagnostic criteria for this study are based on the DSM-IV, it is important to note that Kilpatrick and colleagues (2013) found that the criteria for PTSD in the DSM-IV and DSM-5 were essentially equivalent.

Results

Preliminary Analyses

Prior to hypothesis testing, Pearson correlation analyses were performed to identify significant relationships among study variables (see Table 2). As shown, significant correlations were found between demographic characteristics and background information of the participants (i.e., participant age and parental psychological history) and study variables. Therefore, these characteristics were retained in subsequent analyses as covariates. To help separate the effects of ethnicity from SES, indicators of SES were examined to determine the best proxy for SES in this sample. A MANOVA was used to look at the difference between the two racial/ethnic groups in indicators of SES, which included the occupation and education of each parent. Differences emerged between the two ethnic groups for father education and occupation, F(1, 240) = 26.72, p < .001, and F(1, 240) = 24.84, p < .001, respectively, and mother’s education, F(1, 240) = 16.37, p < .001, with Caucasians reporting that their fathers and mothers achieved a higher level of education and their father’s had more skilled occupations than their African American counterparts. However, no significant difference was reported in the occupation of the mother. Due to the significant effects, these indicators were retained in subsequent analyses as a proxy for SES and included as covariates.

Table 2.

Mean, Standard Deviations, and Correlations Among Study Variables for Full Sample.

			Correlations
Variables	M	SD	1	2	3	4	5	6	7	8	9	10	11	12	13
1. Trauma	2.05	1.67	—
2. Ethnic id	2.96	0.63	−.11	—
3. Diagnoses	1.19	1.42	.42**	−.26**	—
4. Age	19.47	0.92	.16*	.09	.07	—
5. MStatus	—	—	−.06	−.02	.03	.13*	—
6. PMStats	—	—	−.01	.10	.10	.01	.09	—
7. F Educ	—	—	.06	−.12	.06	−.10	−.05	−.16*	—
8. F Occup	—	—	−.02	.11	−.09	−.14*	.08	.18**	−.61**	—
9. M Educ	—	—	.01	−.16*	.02	−.18**	−.02	−.15*	.48**	−.33**	—
10. M Occup	—	—	−.01	.02	−.01	.03	−.02	−.03	−.17**	.11	−.45**	—
11. Hospital	—	—	.04	−.03	.13*	−.01	−.03	.17**	.01	.04	−.01	.04	—
12. Alcohol	—	—	.19**	−.09	.21**	.20**	−.07	.20**	−.13*	.12	−.02	−.06	.24**	—
13. SAbuse	—	—	.19**	−.01	.04	.02	−.04	.18**	−.07	.12	−.07	−.07	.25**	.25**	—

Note. N = 242. Trauma = trauma exposure; Ethnic id = ethnic identity; Diagnoses = psychiatric diagnoses; MStatus = marital status; PMStats = parent marital status; F Educ = father’s education; F Occup = father’s occupation; M Educ = mother’s education; M Occup = mother’s occupation; Hospital = parental psychiatric hospitalizations; Alcohol = parental alcoholism; SAbuse = parental substance abuse.

p ≤ .05. **p ≤ .01. ***p ≤ .001.

Descriptive information on each of the scales used to address study hypotheses is provided at the bottom of Table 1 and on Table 2. Extent of trauma exposure, number of lifetime diagnoses, and ethnic identity were compared by race/ethnicity to identify any differences that may exist between the African American and Caucasian participants. A series of ANOVAs revealed a significant difference between the two racial/ethnic groups in number of lifetime diagnoses and ethnic identity, F(1, 240) = 12.22, p = .001, and F(1, 240) = 105.74, p < .001, respectively, such that African American participants reported a lower number of total lifetime diagnoses than their Caucasian counterparts, and African American women in the sample endorsed a stronger ethnic identity than the Caucasian women in the sample. However, no significant difference was found between the two groups in the number of traumatic events reported.

Trauma, Ethnicity, and Diagnoses

It was expected that there would be an interaction between trauma exposure and race/ethnicity such that the African American women in the sample would report significantly more psychiatric diagnoses in response to trauma exposure than their Caucasian counterparts. To test these hypotheses, a hierarchical linear regression was conducted with age, SES proxy, and parental psychological history (i.e., parental history of hospitalization, substance abuse, and alcoholism) entered in the first step as covariates. Race/ethnicity was dummy coded (0 = African American and 1 = Caucasian) and entered as an independent variable in the second step of the analysis, along with trauma exposure, followed by the interaction term between ethnicity and trauma in the final step. Psychiatric diagnosis (total number of lifetime diagnoses) was entered as the dependent variable. As suggested by Aiken and West (1991), all variables used to compute interaction terms were centered, except for race/ethnicity, which was dichotomous (see Table 3 for regression analyses).

Table 3.

Interactions Between Ethnic Identity and Trauma and Ethnicity and Trauma on Psychiatric Diagnosis.

		Final Step
	R ²	B	SE B	b
Ethnicity and trauma
Step 1: Age		0.10	0.10	0.04
Father occupation		−0.07	0.06	−0.07
Father education		0.02	0.08	−0.05
Mother occupation		0.01	0.04	−0.04
Mother education		0.01	0.08	−0.04
Parental psychiatric hospitalizations		0.48	0.46	0.06
Parental alcoholism		0.63	0.26	0.09*
Parental substance use/abuse	.05	−0.07	0.38	−0.04
Step 2: Trauma exposure		0.60	0.09	0.41***
Ethnicity	.25	0.56	0.20	0.18**
Step 3: Ethnicity × Trauma Exposure	.27	0.44	0.18	0.24*
Ethnic identity and trauma
Step 1: Age		0.10	0.10	0.07
Father occupation		−0.07	0.06	−0.10
Father education		0.02	0.08	0.02
Mother occupation		0.01	0.04	0.01
Mother education		0.01	0.08	0.01
Parental psychiatric hospitalizations		0.48	0.46	0.07
Parental alcoholism		0.63	0.26	0.17
Parental substance use/abuse	.05	−0.07	0.38	−0.01
Step 2: Trauma exposure		0.59	0.08	0.41***
Ethnic identity	.26	−0.28	0.08	−0.20***
Step 3: Ethnic Identity × Trauma Exposure	.28	−0.20	0.08	−0.15**

†

p < .10. *p ≤ .05. **p ≤ .01. ***p ≤ .001.

After controlling for age, SES, and parental psychological history, of which only parental history of alcoholism was significant, F(10, 235) = 1.56, p > .05, R² = .05, adjusted R² = .02, both race/ethnicity (β = .18, p < .01) and trauma exposure (β = .41, p < .001) emerged as significant predictors in the second equation. However, parental history of alcoholism was no longer significant, F(10, 235) = 7.30, p < .001, R² = .24, adjusted R² = .21. In general, higher trauma exposure was related to a higher number of psychiatric diagnoses. In addition, after controlling for age, SES, and parental psychological history, African American women in the sample continued to report fewer lifetime psychiatric diagnoses than the Caucasian women.

As previously mentioned, the interaction term between race/ethnicity and trauma exposure was entered in the third and final step. The final, overall regression equation was significant, F(11, 235) = 7.36, p < .001, R² = .27, adjusted R² = .23, ΔR² = .02, p < .05. The two main effects noted in the previous equation remained significant predictors in the final equation. In addition, the interaction also emerged as a significant predictor (β = .24, p < .02). Simple slope regression analyses (Aiken & West, 1991) were performed to probe the significant interaction and to plot the interaction effect. Close examination of the interaction showed that the relationship between trauma exposure and lifetime psychiatric diagnoses was positive for both racial/ethnic groups, but the positive relationship was stronger among the Caucasian women.

Ethnic Identity as a Moderator of Trauma Exposure and Psychiatric Diagnoses

Another interaction effect was explored to determine if ethnic identity moderated the relationship between trauma exposure and psychiatric diagnosis. Similar to the previous analysis, the covariates were entered, followed by the main effects, and finally an interaction term between trauma and ethnic identity was included in the regression to test for moderation effects. Trauma exposure (β = .41, p < .001) and ethnic identity (β = −.20, p = .001) emerged as significant predictors. Participants who reported a strong connection to their ethnic group reported a lower number of lifetime diagnoses. The final, overall regression equation was significant, F(11, 235) = 7.84, p < .001, R² = .28, adjusted R² = .24, ΔR² = .02, p < .001. The main effects of trauma exposure (β = .40, p < .001) and ethnic identity (β = −.20, p = .001) remained significant in the final equation. In addition, the interaction emerged as a significant predictor (β = −.15, p = .01).

A simple slope regression analysis (Aiken & West, 1991) was performed to interpret the interaction. Similar to the interaction noted between race/ethnicity and trauma, examination of the interaction between ethnic identity and trauma exposure shows that the relationship between trauma exposure and lifetime psychiatric diagnoses was positive at both levels of ethnic identity, but the positive relationship was weaker for those women who reported a strong identification to their ethnic group.

A final three-way interaction effect was explored to determine if the buffering effect of ethnic identity was stronger among the African American women than among the Caucasian women in this sample (see Table 4). An interaction term between race/ethnicity, trauma, and ethnic identity was included in the regression. Although the final, overall regression equation was significant, F(15, 235) = 5.87, p < .001, R² = .28, adjusted R² = .24, the interaction term did not emerge as a significant predictor. In other words, the buffering effect of ethnic identity was not moderated by race/ethnicity. Contrary to the study hypothesis, a strong ethnic identity seemed to serve as a buffer against trauma exposure regardless of the race/ethnicity of the participant.

Table 4.

Interactions Between Ethnicity, Trauma, and Ethnic Identity on Psychiatric Diagnosis.

		Final Step
	R ²	B	SE B	b
Ethnicity, ethnic identity, and trauma
Step 1: Age		0.10	0.10	0.04
Father occupation		−0.07	0.06	−0.09
Father education		0.02	0.08	−0.03
Mother occupation		0.01	0.04	−0.01
Mother education		0.01	0.08	−0.03
Parental psychiatric hospitalizations		0.48	0.46	0.07
Parental alcoholism		0.63	0.26	0.10
Parental substance use/abuse	.05	−0.07	0.38	−0.06
Step 2: Trauma exposure		0.58	0.05	0.40***
Ethnicity		0.28	0.24	0.09
Ethnic identity	.26	−0.22	0.10	−0.16*
Step 3: Ethnic Identity × Trauma Exposure		−0.14	0.10	−0.10*
Ethnicity × Trauma Exposure		0.19	0.23	0.11
Ethnic Identity × Ethnicity	.29	−0.14	0.23	−0.07
Step 4: Ethnic Identity × Ethnicity × Trauma	.29	−0.06	0.27	−0.03

†

p < .10. *p ≤ .05. **p ≤ .01 ***p ≤ .001.

Post Hoc Exploratory Analyses

Results suggest that the lower rate of psychiatric diagnosis reported by African American women in this sample may be explained by the strong ethnic identity reported by these women. To test this speculation, the mediation effect of ethnic identity was explored. Results of the ANOVA support the presence of a significant relationship between race/ethnicity and ethnic identity, which was confirmed by a main effect in a regression analysis (β = −.61, p < .001). In addition, a hierarchical linear regression revealed significant main effects between race/ethnicity and psychiatric diagnoses (β = .23, p < .01) and ethnic identity and psychiatric diagnoses (β = −.19, p < .05). However, once ethnic identity was entered into the equation, the effect of race/ethnicity was no longer significant. The final equation was F(10, 235) = 2.95, p < .01, R² = .12, adjusted R² = .08. As suggested by Baron and Kenny (1986), these results are key indicators of a mediation effect. The Sobel test indicated that ethnic identity (z = 2.30, p < .05) was a significant mediator. Specifically, the effects of race/ethnicity on psychiatric diagnosis in this study were mediated and can be explained by ethnic identity.

Discussion

The effects of trauma exposure were explored among African American and Caucasian college women to identify a factor that could serve a protective function, among African American women, who may be more susceptible to negative psychiatric outcomes resulting from their exposure. Consistent with previous studies (Cougle, Resnick, & Kilpatrick, 2009), the results showed a positive relationship between traumatic exposure and lifetime psychiatric diagnoses, and as expected, race/ethnicity was also found to be significant predictor of diagnosis.

Contrary to the hypothesized relationship, African American women reported fewer lifetime psychiatric diagnoses than the Caucasian women in the sample, although both groups reported experiencing a similar number of traumatic events. Interestingly, the results showed a significant interaction between race/ethnicity and traumatic events on lifetime diagnoses, such that the number of traumatic events seemed to be a more significant predictor for the Caucasian women than for the African American women. This differed from expectation and was inconsistent with previous studies, which found that even though Caucasians reported more trauma than African Americans, African Americans had greater distress and psychiatric illness as a result of trauma (Norris, 1992; Perilla et al., 2002). Norris (1992) attributed the increased distress to African Americans’ greater exposure to severe trauma in the form of violence and other hazards, and to the greater resources readily available to the Caucasians, which could help to mitigate the effects of trauma on emotional well-being. Although Norris (1992) attempted to draw from demographically similar neighborhoods across four cities in the south, it was not possible to completely disentangle race/ethnicity from SES in that study, and SES was not considered in the analysis of the findings.

In contrast, the influence of SES (i.e., parental education and occupation) was considered in the current study to identify the presence of racial/ethnic differences in the socioeconomic backgrounds of the study participants. Consistent with the literature, the African American women in this sample reported that their fathers and mothers completed less formal education and that their fathers had less prestigious occupations than were reported by their Caucasian counterparts. As a result, these indicators were used as a proxy for SES and were entered into subsequent analyses to control for any effects that could be accounted for by SES.

At first glance, it would appear that the results of this study support previous work, which found statistically controlling for SES reduced the higher prevalence of trauma/abuse exposure and certain mental disorders among African American women (Rennison & Planty, 2003). However, on closer examination, it becomes clear that the young African American women in this sample reported fewer lifetime psychiatric diagnoses than the Caucasian women in the sample, even before SES was held constant. There seemed to be a third factor that was accounting for the lower rates of psychiatric diagnoses among the African American women in the sample. Results of the mediation analysis suggest that the third factor was ethnic identity. When ethnic identity was entered as a predictor of psychiatric diagnosis, the significant relationship between race/ethnicity and psychiatric diagnosis disappeared. In other words, it was not their race or ethnicity that was explaining the lower rates of psychiatric diagnoses among the African American women in this sample, but rather their strong ethnic identity.

As expected, ethnic identity served a protective function, moderating the relationship between trauma exposure and psychiatric diagnosis, such that those women who reported a stronger ethnic identity also reported a lower number of psychiatric diagnoses, even in the context of high levels of trauma exposure. Due to the significance of race/ethnicity to the identity development of African American, it was expected that this relationship would emerge among the African American women in this study. Interestingly, this effect was significant across both racial/ethnic groups. Although the African American women tended to report a stronger ethnic identity, which was consistent with previous research (Koutrelakos, 2013; Phinney, 1996), the buffering effects of ethnic identity were also found among the Caucasian women who reported a strong ethnic identity in the sample.

It is important to mention here that Phinney’s (1992) original version of the MEIM, which was used in this study, was developed on an ethnically diverse sample. In an attempt to ensure the relevance of this measure for a variety of groups, differential experiences that are unique or distinctive to particular groups are minimized. This results in a measure that may not adequately capture the way in which African American women’s experiences, histories, and values shape the development of their ethnic identity (Cokley, 2007). It is the understanding and incorporation of this unique history that likely serves a protective function for African Americans (Neblett et al., 2004). Instead of examining the buffering effects of identification with a specific ethnic group or culture on psychological outcomes, this study is likely assessing a more universal concept of belonging and affirmation that seems protective in the context of trauma, regardless of ethnicity. In fact, Ponterotto and Park-Taylor (2007) suggested that the single scale score used in Phinney’s (1992) original MEIM may confound ethnic identification with feelings of belonging and affirmation with one’s ethnic group.

Belonging and connection to a group, in this case, one’s ethnic group, helps to build solidarity with similar peers, which could provide much-needed social and emotional support during difficult times (Stein, Kiang, Supple, & Gonzalez, 2014). For women who have experienced trauma, identification with a larger group may reduce social isolation and feelings of alienation that can foster depression or other psychiatric symptoms (Steger & Kashdan, 2009). In addition, affirmation and positive feelings toward one’s ethnic group have been linked to positive psychological outcomes, such as higher self-esteem, increased satisfaction with life, and decreased symptoms of depression and anxiety (Ghavami, Fingerhut, Peplau, Grant, & Wittig, 2011). Thus, for women whose ethnicity is salient, a strong ethnic identity may provide a foundation for the development of a healthy self-perception, providing psychological resistance to other symptoms of psychological distress.

Results of this study suggest that a strong connection to a group, in this case, an ethnic group, may serve to protect African American women against the development of psychiatric symptoms. An interesting finding of the study is that connection to an ethnic group may play an important role in the psychological protection of Caucasian women as well. Caucasian women whose ethnicity is salient may also gain some psychological benefit from the bonding and affirmation they experience when they are strongly attached to their ethnic group.

Study Limitations

While interesting findings emerged, they will need to be interpreted with caution due to the study limitations. The first limitation is the lack of generalizability of the study. Participants were African American and Caucasian college students who were largely from a working to middle class background. In addition, only a subgroup of the sample responded, resulting in a biased, convenient sample. It is possible that those individuals who chose to respond to a questionnaire about trauma may have had more traumatic experiences than those who chose not to respond. Thus, the current sample represents a subset of college women that may have experienced more trauma and been more distressed than a more representative sample. However, reports of trauma exposure in this study are consistent with similar studies in the literature (Resnick, Falsetti, Kilpatrick, & Freedy, 1996), and given the dearth of research in this area, examination of this sample provides a valid first step to exploring the role of ethnic identity in the context of trauma.

It is important to remember that perceptions of trauma are socially constructed, and different ethnic groups may not always consider similar incidents as traumatic. In fact, a study examining the validity of the SLESQ among low income African American women (Green, Chung, Daroowalla, Kaltman, & DeBenedictis, 2006) found that participants failed to identify potential traumatizing events (i.e., robbery and attempted rape) as traumatic. However, participants did identify these events as stressful and they tended to identify most other SLESQ events as traumatic. Hence, Green and colleagues (2006) found the SLESQ to be a valid measure to use among similar African American populations. Although there was not a significant difference between the African American and Caucasian participants in the number of traumatic events reported, findings should be interpreted bearing the limitation of the SLESQ in mind.

Another limitation was the inadequate specificity regarding the ethnicity of the participants. Due to the simple racial categorizations used in this study, nuances within race that result from ethnic distinctions could not be examined (e.g., distinctions between Blacks of African American heritage vs. Caribbean heritage or Caucasians with Jewish heritage vs. Italian). Finally, severity of trauma was not considered in the current investigation. Instead, analysis in this study focused on the number of traumatic events reported. It is possible that more severe diagnoses could have resulted from more severe trauma and not necessarily an increased number of traumatic events.

Despite these limitations, this study makes interesting assertions regarding the importance of belonging and affirmation in the mental health of both African American and Caucasian women. Similar to previous studies, African American women in this sample endorsed stronger levels of ethnic identity. However, ethnic identity served a protective function among both the African American and Caucasian participants by moderating the relationship between trauma exposure and the endorsement of psychiatric symptoms and diagnoses. The protective effects noted across ethnic groups in this study are likely due to the measure of ethnic identity used. Phinney’s (1992) original MEIM seems to be less of an assessment of identification with a specific culture and more an assessment of a universal concept of connection and/or belonging to a social group, in this case, an ethnic group. Based on these findings, service providers and mental health practitioners are encouraged to ask about social group membership and affiliation. For some, as is the case for many African American women, their most salient social group membership may be their ethnic group. Fostering attachment and affirmation to salient social groups may prove to be therapeutic for any woman, regardless of her race. Future studies should use an identity measure developed to assess identification to specific African American culture to determine the unique role that African American ethnic identity may play in reducing trauma-related distress among African American women. In addition, the literature regarding the protective role that ethnic identity may play in trauma exposure could be further developed by examining the impact of ethnic identity on specific behaviors (i.e., social justice advocacy vs. social withdrawal or community disengagement) in the context of traumatic stress.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by a grant from the National Institute of Mental Health of the National Institutes of Health (R0150332) to Bonnie Green.

Author Biographies

Tiffany G. Townsend, PhD, is an adjunct associate professor of psychiatry in the School of Medicine at Georgetown University. Her work involves the implementation of community-based research and prevention programs to decrease health and mental health disparities among ethnic minority women, children, and families. She has received most of her funding to support HIV prevention research among youth populations of color. Specifically, she served as the principal investigator (PI) on four federally funded grants totaling over US$3,000,000.

Stacey Kaltman, PhD, is a professor in the Department of Psychiatry and assistant director of the Center for Trauma and the Community (CTC) in the School of Medicine at Georgetown University. She is a licensed clinical psychologist whose research focuses on the integration of mental health services into primary care settings, with a focus on treatment for trauma-related mental disorders and the intersection between physical and mental health intervention. In addition to teaching medical students about the impact of trauma, her teaching focuses on doctor–patient communication, behavioral and psychosocial aspects of medical care, and the mind–body connection.

Farzana Saleem, MPhil, is a doctoral student in clinical psychology at George Washington University. Her research examines risk and protective factors for racial and neighborhood-level stressors. In particular, she is interested in how parents prepare and socialize their children to manage racial discrimination.

Dionne S. Coker-Appiah, PhD, MAED, is an assistant professor, psychologist, and adolescent health researcher in the Department of Psychiatry in the School of Medicine at Georgetown University. Her research program focuses on using community-based participatory research approaches to prevent adolescent dating violence, and the associated health implications among adolescents of Color in the US and Ghana, West Africa. Her current NIMHD funded research project focuses on the development of the “Project LOVE: Letting Our Voices Empower Adolescent Dating Violence Prevention Intervention”.

Bonnie L. Green, PhD, is a professor emeritus in the Department of Psychiatry in the School of Medicine at Georgetown University. Her research at Georgetown focused on the psychological and physical health consequences of individual traumas experienced by women, including breast cancer, traumatic bereavement, and interpersonal violence. Her most recent work addressed the mental health needs of poor women who receive their health care in primary care settings that serve low-income/safety net populations, and treatments for PTSD among veterans. She is past Editor of the Journal of Traumatic Stress, and past President of the International Society for Traumatic Stress Studies.

References

Afifi

(2007). Gender differences in mental health. Singapore Medical Journal, 48, 385-391.

Aiken

L. S.

West

S. G.

(1991). Multiple regression: Testing and interpreting interactions. Thousand Oaks, CA: Sage.

Aldwin

C. M.

(2007). Stress, coping, and development: An integrative perspective (2nd ed.). New York, NY: The Guilford Press.

Alim

T. N.

Graves

Mellman

T. A.

Aigbogun

Gray

Lawson

Charney

D. S.

(2006). Trauma, posttraumatic stress disorder and depression in an African-American population. Journal of the National Medical Association, 98, 1630-1636.

American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorder (4th ed.). Washington, DC: Author.

American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Arlington, VA: American Psychiatric Publishing.

Baron

R. M.

Kenny

D. A.

(1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality & Social Psychology, 51, 1173-1182. doi:10.1037/0022-3514.51.6.1173

Bent-Goodley

T. B.

(2004). Perceptions of domestic violence: A dialogue with African American women. Health and Social Work, 29, 307-316.

Bruce

Waelde

L. C.

(2008). Relationships of ethnicity, ethnic identity, and trauma symptoms to delinquency. Journal of Loss and Trauma, 13, 395-405.

10.

Bryant-Davis

Belcourt-Dittloff

Chung

Tillman

(2009). The cultural context of trauma recovery: The experiences of ethnic minority women. In Chin

J. L.

(Ed.), Diversity in mind and in action, Vol. 3: Social justice matters (pp. 127-147). Santa Barbara, CA: Praeger.

11.

Carlson

Harris

Holden

(1999). Protective orders domestic violence: Risk factors for re-abuse. Journal of Family Violence, 14, 205-226. doi:10.1023/A:1022032904116

12.

Cokley

(2007). Critical issues in the measurement of ethnic and racial identity: A referendum on the state of the field. Journal of Counseling Psychology, 54, 224-234. doi:10.1037/0022-0167.54.3

13.

Cougle

J. R.

Resnick

Kilpatrick

D. G.

(2009). PTSD, depression, and their comorbidity in relation to suicidality. Depression & Anxiety, 26, 1151-1157. doi:10.1002/da.20621

14.

Diemer

M. A.

Mistry

R. S.

Wadsworth

M. E.

López

Reimers

(2013). Best practices in conceptualizing and measuring social class in psychological research. Analyses of Social Issues and Public Policy, 13, 77-113. doi:10.1111/asap.12001

15.

El-Khoury

M. Y.

Dutton

M. A.

Goodman

L. A.

Engel

Belamaric

R. J.

Murphy

(2004). Ethnic differences in battered women’s formal help-seeking strategies. Cultural Diversity & Ethnic Minority Psychology, 10, 383-393. doi:10.1037/1099-9809.10.4.383

16.

First

M. B.

Spitzer

R. L.

Gibbon

Williams

J. B. W.

(1996). Structured clinical interview for DSM-IV Axis I disorders. New York, NY: Biometrics Research Department, New York State Psychiatric Institute.

17.

Ghavami

Fingerhut

Peplau

Grant

Wittig

(2011). Testing a model of minority identity achievement, identity affirmation, and psychological well-being. Cultural Diversity & Ethnic Minority Psychology, 17, 79-88. doi:10.1037/a0022532

18.

Goodman

L. A.

Corcoran

Turner

Yuan

Green

B. L.

(1998). Assessing traumatic event exposure: General issues and preliminary findings for the Stressful Life Events Screening Questionnaire. Journal of Traumatic Stress, 11, 521-542.

19.

Graves

Kaslow

Frabutt

(2010). A culturally-informed approach to trauma, suicidal behavior, and aggression in African American adolescents. Aggression & Violent Behavior, 15, 36-41.

20.

Green

Chung

Daroowalla

Kaltman

DeBenedictis

(2006). Evaluating the cultural validity of the stressful life events screening questionnaire. Violence Against Women, 12(12), 191-213.

21.

Green

Krupnick

Stockton

Goodman

Corcoran

Petty

(2001). Psychological outcomes associated with traumatic loss in a sample of young women. American Behavioral Scientist, 44, 817-837. doi:10.1177/00027640121956511

22.

Hatch

S. L.

Dohrenwend

B. P.

(2007). Distribution of traumatic and other stressful life events by race/ethnicity, gender, SES and age: A review of the research. American Journal of Community Psychology, 40, 313-332. doi:10.1007/s10464-007-9134-z

23.

Jones

J. M.

(2007). Exposure to chronic community violence: Resilience in African American children. Journal of Black Psychology, 33, 125-149. doi:10.1177/0095798407299511

24.

Karyl

Waelde

Hodges

W. F.

Starek

Heidt

Min

(1995). Retrospective reports of parental partner abuse: Relationships to depression, trauma and self-esteem among students. Journal of Family Violence, 10, 177-202. doi:10.1007/BF02110599

25.

Kessler

R. C.

Sonnega

Bromet

Hughes

Nelson

C. B.

(1995). Posttraumatic stress disorder in the National Comorbidity Survey. Archives of General Psychiatry, 52, 1048-1060. doi:10.1001/archpsyc.1995.03950240066012

26.

Kilpatrick

Resnick

H. S.

Milanak

M. E.

Miller

M. W.

Keys

K. M.

Friedman

M. J.

(2013). National estimates of exposure to traumatic events and PTSD prevalence using DSM-IV and DSM-5. Journal of Traumatic Stress, 26, 537-547. doi:10.1002/jts.21848

27.

Kilpatrick

Resnick

H. S.

Ruggiero

Conoscenti

McCauley

(2007). Drug-facilitated, incapacitated, and forcible rape: A national study. Charleston, SC: National Crime Victims Research & Treatment, Medical University of South Carolina.

28.

Kilpatrick

Ruggiero

Acierno

Saunders

Resnick

Best

(2003). Violence and risk of PTSD, depression, substance abuse/dependence, and comorbidity. Journal of Consulting & Clinical Psychology, 71, 692-700. doi:10.1037/0022-006X.71.4.692

29.

Koutrelakos

(2013). Ethnic identity: Similarities/differences in white groups based on cultural practices. Psychological Reports, 112, 745-762. doi:10.2466/17.10.PR0.112.3.745-762

30.

Krupnick

J. L.

Green

B. L.

Stockton

Goodman

Corcoran

Petty

(2004). Mental health effects of adolescent trauma exposure in a female college sample. Psychiatry, 67, 264-279. doi:10.1521/psyc.67.3.264.48986

31.

Martinez

R. O.

Dukes

R. L.

(1997). Effects of ethnic identity, ethnicity, gender on adolescent well-being. Journal of Youth & Adolescence, 26, 503-516. doi:10.1023/A:1024525821078

32.

Nakao

Treas

(1994). Updating occupational prestige and socioeconomic scores: How the new measures measure up. Sociological methodology, 24, 1-72. doi:10.2307/270978

33.

Neblett

E. J.

Shelton

J. N.

Sellers

R. M.

(2004). The role of racial identity in managing daily racial hassles. In Philogène

(Eds.), Racial identity in context: The legacy of Kenneth B. Clark (pp. 77-90). Washington, DC: American Psychological Association. doi:10.1037/10812-005

34.

Norris

F. H.

(1992). Epidemiology of trauma: Frequency and impact of different potentially traumatic events on different demographic groups. Journal of Consulting and Clinical Psychology, 60, 409-418. doi:10.1037/0022-006X.60.3.409

35.

Perilla

J. L.

Norris

F. H.

Lavizzo

E. A.

(2002). Ethnicity, culture, and disaster response: Identifying and explaining ethnic differences in PTSD. Journal of Social and Clinical Psychology, 21, 20-45. doi:10.1521/jscp.21.1.20.22404

36.

Phinney

J. S.

(1992). The multigroup ethnic identity measure: A new scale for use with diverse groups. Journal of Adolescent Research, 7, 156-176. doi:10.1177/074355489272003

37.

Phinney

J. S.

(1996). When we talk about American ethnic groups, what do we mean? American Psychologist, 51, 918-927. doi:10.1037/0003-066X.51.9.918

38.

Ponterotto

J. G.

Park-Taylor

(2007). Racial and ethnic identity theory, measurement, and research in counseling psychology: Present status and future directions. Journal of Counseling Psychology, 54, 282-294. doi:10.1037/0022-0167.54.3.282

39.

Rennison

C. M.

Planty

(2003). Nonlethal intimate partner violence: Examining race, gender, and income patterns. Violence and Victims, 18, 433-443. doi:10.1891/vivi.2003.18.4.433

40.

Rennison

C. M.

Welchans

(2000). Bureau of Justice Statistics special report: Intimate partner violence (NCJ No. 178247). Washington, DC: U.S. Department of Justice.

41.

Resnick

H. S.

Falsetti

S. A.

Kilpatrick

D. G.

Freedy

J. R.

(1996). Assessment of rape and other civilian trauma-related post-traumatic stress disorder: Emphasis on assessment of potentially traumatic events. In Miller

T. W.

(Ed.), Stressful life events (pp. 231-266). Madison, WI: International Universities Press.

42.

Roberts

A. L.

Gilman

S. E.

Breslau

Koenen

K. C.

(2011). Race differences in exposure to traumatic events, development of PTSD, and treatment-seeking for PTSD in the US. Psychological Medicine, 41, 71-83. doi:10.1017/S0033291710000401

43.

Seaton

E. K.

Yip

Morgan-Lopez

Sellers

R. M.

(2012). Racial discrimination and racial socialization as predictors of African American adolescents’ racial identity development using latent transition analysis. Developmental Psychology, 48, 448-458. doi:10.1037/a0025328

44.

Steger

M. F.

Kashdan

T. B.

(2009). Depression and everyday social activity, belonging, and well-being. Journal of Counseling Psychology, 56, 289-300. doi:10.1037/a0015416

45.

Stein

G. L.

Kiang

Supple

A. J.

Gonzalez

L. M.

(2014). Ethnic identity as a protective factor in the lives of Asian American adolescents. Asian American Journal of Psychology, 5, 206-213. doi:10.1037/a0034811

46.

Thomas

Townsend

T. G.

Belgrave

F. Z.

(2003). The influence of cultural and racial identification on the behavioral adjustment of inner city African American youth. American Journal of Community Psychology, 32, 217-228. doi:10.1023/B:AJCP.0000004743.37592.26

47.

Thompson

M. P.

Kaslow

N. J.

Kingree

J. B.

Rashid

Puett

Jacobs

Matthews

(2000). Partner violence, social support, and distress among inner-city African American women. American Journal of Community Psychology, 28, 127-143. doi:10.1023/A:1005198514704

48.

Townsend

T. G.

(2002). The impact of self-components on sexual attitudes among African American pre-adolescent girls. Sex Roles, 47, 11-20. doi:10.1023/A:1020675518819

49.

Townsend

T. G.

Batts

A. L.

Hawkins

S. R.

(2007). Stress and stress reduction among African American women: A brief report. Journal of Primary Prevention, 28, 569-582. doi:10.1007/s10935-007-0111-y

50.

van der Kolk

B. A.

McFarlane

A. C.

Weisaeth

(1996). Traumatic stress: The effects of overwhelming experience on mind, body, and society. New York, NY: Guilford Press.

51.

Waelde

L. C.

Pennington

Mahan

Kabour

Marquett

(2010). Psychometric properties of the Race-Related Events Scale. Psychological Trauma: Theory, Research, Practice, and Policy, 2, 4-11. doi:10.1037/a0019018

52.

Westcott

(2015). Race, criminalization, and historical trauma in the United States: Making the case for a new justice framework. Traumatology, 21, 273-284. doi:10.1037/trm0000048

53.

Whaley

A. L.

(2003). Cognitive-cultural model of identity and violence prevention for African American youth. Genetic, Social, and General Psychology Monographs, 129, 101-151.

54.

Wiltshire

J. C.

Person

S. D.

Kiefe

C. I.

Allison

J. J.

(2009). Disentangling the influence of socioeconomic status on differences between African American and White women in unmet medical needs. American Journal of Public Health, 99, 1659-1665. doi:10.2105/AJPH.2008.154088