Severe Physical Violence and Black Women’s Health and Well-Being

Abstract

Objectives. We evaluated the association between intimate partner violence and the mental and physical health status of US Caribbean Black and African American women.

Methods. We used 2001 to 2003 cross-sectional data from the National Survey of American Life—the most detailed study to date of physical and mental health disorders of Americans of African descent. We assessed participants’ health conditions by the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (Washington, DC; American Psychological Association) Composite International Diagnostic Interview.

Results. We found differences in health conditions between abused African American and Caribbean Black women. There were increased risks for lifetime dysthymia, alcohol dependence, drug abuse, and poor perceived health for African American victims of partner abuse, and binge eating disorder was associated with partner violence among Caribbean Black women.

Conclusions. Severe intimate partner violence was associated with negative mental and physical health outcomes for US Black women, with different patterns between African American and Caribbean Blacks. Understanding intimate partner violence experiences of US Black women requires recognition of key intragroup differences, including nativity and immigrant status, and their differential relationships to women’s health.

Intimate partner violence (IPV) is a serious public health problem that has devastating consequences for the health and well-being of women.¹ Nearly 28% of women in the United States have experienced IPV.² The risk is heightened for Black women, with an estimated 4 in 10 experiencing physical abuse by a partner in their lifetimes.² The long- and short-term effects of partner violence may be greater for women within this population, who not only experience violence at much higher rates than do other ethnic groups (e.g., White, Hispanic)^1,3–9 but also are exposed to external factors and social conditions that increase their chances for poorer outcomes.^4,9–14

Previous studies have indicated that associated mental conditions of IPV include depression, posttraumatic stress disorder, anxiety, suicide, and tendencies for substance use (both legal and illegal).^15–27 Along with these possible mental disorders, IPV victims are prone to physical health problems, such as increased risk for back, limb, gastrointestinal, stomach, and gynecological problems.^22,28–31

Despite the various health problems that are associated with IPV toward women in general, research devoted to understanding the influences on US Black women is limited.^19,32–34 The few studies conducted have primarily used unstructured clinical assessment and are determined by clinical and community-based samples.^33,35 Studies using national samples and structured clinical tools are rare, limiting valid assessments and definitive statements on the resulting effects of IPV on women within this population.

In addition to these shortcomings, previous research has typically aggregated Blacks into a single category, which may obscure key intragroup differences.^5,11,36 This is especially problematic because groups may have culturally distinct behaviors, practices, and experiences that may exacerbate certain health conditions or, conversely, serve as protective factors. Evidence suggests that health conditions may vary according to race and ethnicity.^{10,13,28,37–39} These differences, in particular, have become more apparent between African Americans and Caribbean Blacks.⁴⁰ To date, however, we have less knowledge about the differences in health outcomes that may exist among abused women within these populations. Importantly, the impact of IPV on the health and well-being of US Caribbean women, one of the largest and fastest growing ethnic groups in the United States,⁴¹ has not been explored in depth.

We addressed 2 underlying questions: (1) What are the associations between IPV, mental health disorders, and the physical health of African American and US Caribbean Black women? and (2) Are there differences in health outcomes between abused African American and US Caribbean Black women?

METHODS

Focusing on women of African descent, we examined cross-sectional data from the National Survey of American Life, which as part of the Collaborative Psychiatric Epidemiological Study, is the most detailed study to date of mental disorders and the physical health of Americans of African descent and 1 of the first national probability samples that includes Blacks of different ethnic groups living in the same context.^34,42 Data were collected over 3 years—2001 through 2003—using multistage probability sampling methods.

In-person interviews were the primary method of data collection, with a smaller percentage of the sample (i.e., 10%) collected by telephone interviews. Interviews typically lasted an average of 2 hours and 20 minutes in the general sample. A $50 honorarium was given to each respondent for participation in the study. The study surveyed 6082 adults aged 18 years and older, including 3570 African Americans, 1623 Caribbean Blacks, and 890 non-Hispanic Whites. An overall response rate of 72.3% was obtained in general: 70.7% for African Americans, 77.7% for Caribbean Blacks, and 69.7% for non-Hispanic Whites.

Measures

We used a single measure to assess severe lifetime physical abuse. The question posed to participants was “Were you ever badly beaten up by a spouse or romantic partner?” The measure included response categories of “yes” and “no” (1 = yes; 0 = no). In total, 505 Black women in the sample reported having experienced severe lifetime physical violence in an intimate relationship. We compared the IPV measure with the US National Comorbidity Survey Replication dichotomously defined severe partner violence Conflict Tactic Scale measure^43a,43b within the Collaborative Psychiatric Epidemiological Study.¹⁹ To address the proposed measure’s validity, we conducted 2 tests—probability of agreement (odds ratio [OR]) and area under the curve (AUC)^44–46—and found that they have fair association among African American women (data were not available for Afro-Caribbean women) across different approaches to estimating agreement (OR = 4.5; confidence interval [CI] = 1.49, 14.98; P < .001; AUC ≥ 0.6).

We derived the criteria for mental health disorders from the World Health Organization Composite International Diagnostic Interview defined by the Diagnostic and Statistical Manual of Mental Disorder, Fourth Edition (DSM-IV).^47,48 The National Survey of American Life used this structured diagnostic interview to assess lifetime mental disorders and conditions. We examined participants that met criteria for the following disorders: mood disorders (major depressive episode, dysthymia, major depressive disorder), anxiety disorders (panic, agoraphobia, generalized anxiety disorder, obsessive compulsive disorder, posttraumatic stress disorder), substance disorders (alcohol abuse and dependence, drug abuse and dependence), and eating disorders (anorexia, bulimia, binge eating).

Questions related to suicide ideation and planning were also administered to participants. These questions included “Have you ever seriously thought about committing suicide?” and “Have you ever attempted suicide?”

“Any disorder” was a composite of all mood, anxiety, substance, and eating and other disorders (e.g., childhood, conduct disorder). Participants who experienced “any disorder” were those who had at least 1 of the disorders listed.

The perceived overall general physical health of respondents was assessed by the question “How would you rate your overall physical health at the present time?” The response categories were excellent, very good, good, fair, and poor. We coded the measure in accordance with previous studies to reflect poor health status in comparison with others (1 = poor; 0 = else).^35,49

We examined 2 ethnic groups: African Americans and Caribbean Blacks. African Americans comprised all individuals of African heritage but without Caribbean ancestry. Caribbean Blacks were individuals with African heritage who reported Caribbean background or ancestry from defined geographical areas constituting the Caribbean region. Specifically, these individuals either had parents or grandparents born in the Caribbean or were from the list of Caribbean countries presented by interviewers.⁴²

Analyses

We used bivariate logistic regression analytic procedures to address the associations between IPV and various mental health disorders and general physical health. We computed adjusted odd ratios (AORs) and 95% CIs. The cut-off for significance was set at 0.05 α levels.

Because the National Survey of American Life used a multistage sample design, we adjusted the CI reported for both clustering and stratification. We conducted the analyses using Stata analytic software employing the Taylor expansion approximation technique for calculating the complex design-based estimates of variance.⁵⁰

RESULTS

The sample consisted of women aged 34 to 49 years (41.7%) who were either separated or divorced (33.6%) and had less than a high school education (34.1%; Table 1). A majority of participants were employed (58.8%) and earned incomes less than $25 000. The geographic location of women varied, with most living in the Southern region (46.2%) and in major metropolitan or urban areas (91.4%).

TABLE 1—

Sample Characteristics of Severely Physically Abused Black Women (n = 505): National Survey of American Life, United States, 2001–2003

Characteristics	No. (%)
Age of respondents, y
18–24	41 (8.9)
25–34	121 (20.7)
35–49	193 (41.7)
50–64	109 (22.7)
≥ 65	41 (6.0)
Marital status
Married	81 (20.6)
Partnered	48 (10.0)
Separated or divorced	187 (33.6)
Widowed	49 (8.9)
Never married	140 (26.9)
Education
Less than high school	165 (34.1)
High school graduate	176 (31.2)
Some college	114 (23.9)
College graduate	50 (10.8)
Household income, $
< 25 000	328 (62.5)
25 000–34 999	70 (12.5)
35 000–49 999	47 (10.7)
50 000–74 999	41 (9.8)
≥ 75 000	19 (4.6)
Occupational status
Employed	291 (58.8)
Unemployed	81 (16.1)
Not in the labor force	133 (25.1)
Region
Northeast	146 (20.7)
Midwest	85 (23.3)
South	247 (46.2)
West	27 (9.8)
Urbanicity
Major metropolitan or urban	468 (91.4)
Suburban	31 (7.4)
Rural	6 (1.2)

Associated Physical and Mental Conditions

As shown in Table 2, we found an association of increased risk for any mood disorder with severe physical IPV among Black women in general (AOR = 2.5; 95% CI = 2.00, 3.22). We found this association for specific types of mood disorders, including major depressive episodes (AOR = 2.7; 95% CI = 2.10, 3.44), dysthymia (AOR = 3.4; 95% CI = 1.86, 6.32), major depressive disorder (AOR = 2.7; 95% CI = 2.10, 3.44), and bipolar disorder (AOR = 2.1; 95% CI = 1.19, 3.70).

TABLE 2—

Correlates of Lifetime Abuse and Physical and Mental Disorders Among US Black Women: National Survey of American Life, 2001–2003

Risk Factor Variable	AOR (95% CI)
Mood disorder
Any	2.53*** (2.00, 3.22)
Major depressive episode	2.69*** (2.10, 3.44)
Dysthymia	3.43*** (1.86, 6.32)
Major depressive disorder	2.67*** (2.10, 3.44)
Bipolar	2.09* (1.19, 3.70)
Anxiety disorder
Any	2.68*** (2.18, 3.30)
Panic disorder	3.60*** (2.16, 6.00)
Agoraphobia	3.74*** (2.44, 5.72)
Generalized anxiety disorder	1.99*** (1.32, 2.99)
Obsessive compulsive disorder	2.65** (1.30, 5.42)
Posttraumatic stress disorder	3.50*** (2.70, 4.54)
Substance disorder
Any	4.45*** (2.73, 7.26)
Alcohol abuse	4.91*** (3.14, 7.66)
Alcohol dependence	5.09*** (2.80, 9.28)
Drug abuse	4.15*** (2.24, 7.70)
Drug dependence	6.56*** (2.77, 15.53)
Eating disorder
Any	1.94** (1.26, 2.97)
Anorexia	1.03 (0.83, 12.73)
Bulimia	2.87** (1.52, 5.42)
Binge eating	1.95** (1.28, 2.99)
Suicide
Attempts	3.36*** (2.07, 5.51)
Ideation	2.84*** (2.13, 3.79)
Any disorder	3.39*** (2.75, 4.18)
Poor health	3.00*** (1.97, 4.71)

Note. AOR = adjusted odds ratio; CI = confidence interval. We compared the intimate partner violence measure with the National Comorbidity Survey Replication dichotomously defined severe partner violence measure from the severe physical violence subscales of the Conflict Tactics Scale within the Collaborative Psychiatric Epidemiological Study and found it to have a fair agreement across different estimates (odds ratio = 4.5; 95% CI = 1.49, 14.98; P < .001; area under the curve ≥ 0.6).

*P < .05; **P < .01; ***P < .001.

There was also an association between increased risk for any anxiety disorder (AOR = 2.7; 95% CI = 2.18, 3.30) and experiences with partner abuse. Greater risk for panic disorder (AOR = 3.6; 95% CI = 2.16, 6.00), agoraphobia (AOR = 3.7; 95% CI = 2.44, 5.72), generalized anxiety disorders (AOR = 2.0; 95% CI = 1.32, 2.99), obsessive–compulsive disorder (AOR = 2.7; 95% CI = 1.30, 5.42), and posttraumatic stress disorder (AOR = 3.0; 95% CI = 2.70, 4.54) was associated with severe intimate victimization.

We found associations between IPV and heightened risk for substance disorders. We found significant associations of IPV with any substance disorders (AOR = 4.5; 95% CI = 2.73, 7.26), alcohol abuse (AOR = 4.9; 95% CI = 3.14, 7.66), alcohol dependence (AOR = 5.1; 95% CI = 2.80, 9.28), drug abuse (AOR = 4.2; 95% CI = 2.24, 7.70), and drug dependence (AOR = 6.6; 95% CI = 2.77, 15.53).

Abused US Black women were also vulnerable to eating disorders. With the exception of anorexia nervosa, we found associations of increased risk for any eating disorder (AOR = 1.9; 95% CI = 1.26, 2.97), bulimia (AOR = 2.9; 95% CI = 1.52, 5.42), and binge eating (AOR = 2.0; 95% CI = 1.28, 2.99) among victims of intimate partner abuse. This association further extended to increased risk of suicide ideation (AOR = 2.8; 95% CI = 2.13, 3.79) and attempts (AOR = 3.4; 95% CI = 2.07, 5.51) among IPV victims. Overall, we found an association between IPV and any mental disorder (AOR = 3.4; 95% CI = 2.75, 4.18), and poor perceived general physical health (AOR = 3.0; 95% CI = 1.97, 4.71).

Health Outcomes of Ethnic Intimate Partner Victims

Table 3 presents a comparison of mental health disorders and physical health correlates separately between abused African American and Caribbean Black women. In general, African American and Caribbean Black women had similar associations. There were noticeable differences, however, between the groups within some of the health categories. For example, among African American abused women there was increased risk for dysthymia (AOR = 3.4; 95% CI = 1.80, 6.47), alcohol dependence (AOR = 5.0; 95% CI = 2.71, 9.36), drug abuse (AOR = 4.2; 95% CI = 2.2, 8.0), and poor perceived general health (AOR = 3.2; 95% CI = 2.00, 5.15)—all associated with severe IPV. The opposite was true for Caribbean Black women. Instead, among Caribbean abused Black women there were increased risks for certain eating disorders, such as binge eating (AOR = 3.7; 95% CI = 1.03, 13.50), unlike abused African American women.

TABLE 3—

Correlates of Intimate Partner Violence by Ethnicity: National Survey of American Life, United States, 2001–2003

Risk Factor Variable	African Americans, AOR (95% CI)	Caribbean Blacks, AOR (95% CI)a
Mood disorder
Any	2.52*** (1.96, 3.24)	2.97* (1.27, 6.94)
Major depressive episode	2.70*** (2.08, 3.51)	2.57* (1.18, 5.57)
Dysthymia	3.41*** (1.80, 6.47)	3.14 (0.82, 12.02)
Major depressive disorder	2.70*** (2.09, 3.51)	2.57* (1.18, 5.57)
Bipolar	1.92* (1.05, 3.54)	10.34*** (3.48, 30.71)
Anxiety disorder
Any	2.62*** (2.11, 3.25)	3.81*** (1.91, 7.59)
Panic disorder	3.46*** (2.02, 5.95)	7.07** (2.03, 24.68)
Agoraphobia	3.88*** (2.46, 6.11)	1.84* (1.01, 3.37)
Generalized anxiety disorder	1.89** (1.23, 2.92)	4.68** (1.49, 14.71)
Obsessive compulsive disorder	2.51* (1.18, 5.34)	8.40** (1.75, 40.39)
Posttraumatic stress disorder	3.37*** (2.56, 4.45)	6.73*** (3.85, 11.79)
Substance disorder
Any	4.45*** (2.67, 7.36)	2.85* (1.02, 7.94)
Alcohol abuse	4.91*** (3.11, 7.76)	3.16* (1.04, 9.60)
Alcohol dependence	5.03*** (2.71, 9.36)	4.80 (0.54, 42.55)
Drug abuse	4.21*** (2.21, 8.01)	2.23 (0.58, 8.66)
Drug dependence	6.62*** (2.64, 16.60)	5.57* (1.43, 21.75)
Eating disorderb
Any	1.86** (1.18, 2.94)	3.73* (1.03, 13.50)
Anorexia	1.01 (0.08, 12.87)	NA
Bulimia	2.94** (1.45, 5.94)	3.03* (1.03, 8.91)
Binge eating	1.88 (1.19, 2.96)	3.72* (1.03, 13.50)
Suicide
Attempts	3.20*** (1.91, 5.36)	10.62*** (3.48, 32.36)
Ideation	2.78*** (2.06, 3.77)	3.98** (1.54, 10.27)
Any disorder	3.36*** (2.69, 4.19)	3.68*** (1.84, 7.33)
Poor health	3.21*** (2.00, 5.15)	1.51 (0.29, 7.84)

Note. AOR = adjusted odds ratio; CI = confidence interval; NA = not available.

^aSubsequent comparisons between abused African Americans and second-generation Caribbean Black women largely revealed no significant differences in health outcomes between cohorts with the exception of very few disorders (e.g., bipolar, any eating disorder, binge eating) in which higher rates were found for Caribbean Black women.

^bThe National Survey of American Life adolescent data set reveals a very low rate of eating disorders among women within this population (< 1%), supporting our interpretation.^62,63

*P < .05; **P < .01; ***P < .001.

DISCUSSION

We addressed the associations between IPV and the physical and mental statuses of US Black women. In general, we have provided additional evidence that US Black women, like other ethnic women, are prone to poor health outcomes resulting from IPV. Consistent with other studies, victims of abuse were more likely to experience various mood, anxiety, and substance disorders and suicide tendencies.^{15–19,28,30,32} Abused women were also more vulnerable to poor perceived general physical health.

We also found that partner violence might have different effects on Black women from different ethnic groups. For example, we found an association between eating disorders (e.g., binge eating) among Caribbean Black, but not African American, women victims of abuse. These differences in behavior could be driven by different coping strategies for Caribbean Black versus African American women. Research suggests that cultural differences strongly influence coping patterns among victims of IPV.^51–53 Disordered eating is considered a passive, emotional coping strategy and may represent women’s desire to exert greater personal control over their lives in cases in which they feel an abusive partner is impeding it.^51,54

Because Caribbean cultural norms place great value on family privacy and the separation of public and private life, Caribbean victims of abuse may be discouraged from publically addressing personal experiences with IPV.^55,56 African American women may have greater cultural support for using more active and public strategies for coping with abuse. Among Caribbean women, disordered eating could be an immediate behavioral response to the stressors of IPV in the context of constrained culturally acceptable behavioral alternatives.^55,56

Conversely, we found a relationship between dysthymia, alcohol dependence, drug abuse, and poor perceived general health among African American victims of severe IPV; this was not the case for Caribbean Black women. These differences in health outcomes are difficult to explain simply and warrant further exploration. This finding may reflect the cultural differences between these groups related to coping with long- and short-term traumatic experiences. It is possible that abused Caribbean Black women are more protected from some disorders and health conditions than are their African American counterparts as a result of migratory experiences.^38,57 There is a notion that voluntary international migrants, such as US Caribbean Blacks, may have relatively higher levels of self-efficacy, resilience, coping strategies, and resources associated with migration.

These factors along with the maintenance of strong ties to formal and informal networks among other Afro-Caribbean members may provide certain health benefits to IPV victims that African American women are not afforded. Research shows that IPV victims who received greater interpersonal support had fewer depressive symptoms and better mental health.⁵⁸ Additionally, differing cultural frameworks between African American and Afro-Caribbeans may shape perceptions of abuse and influence distinct emotional and behavioral responses. For example, African American women may engage in certain behaviors (e.g., drug or alcohol abuse) as a means of coping with the abuse of an intimate partner. This might not be the same for Caribbean Black women who, because of cultural norms, may restrict such behavior and find other more culturally acceptable ways of coping.⁵⁹ This could come in the form of abnormal eating disorders (e.g., binge eating) or changes in normal eating habits.

This study, like many others, is not without shortcomings. One shortcoming in particular is that we attempted to address only the association between possible health outcomes and severe lifetime IPV. As we are aware, association does not confirm causality. Because of the cross-sectional nature of the data, some women may have already had existing health conditions before being abused. We were unable to assess the temporal relationship between intimate partner abuse and physical and mental conditions. Longitudinal and comparative analyses are needed to fully evaluate these relationships. Moreover, we did not control for moderating or mediating factors that may influence certain outcomes.

Factors associated with immigrant selectivity such as self-efficacy, resiliency, resourcefulness, and generational status may play a determining role in the health benefits Caribbean women experience. Additionally, we focused solely on severe physical violence. We did not evaluate other forms of abusive acts (e.g., sexual, psychological) that could have detrimental effects on the overall well-being of women. Finally, the measure we used to address severe IPV may have different meaning for different cultural groups, resulting in selection bias.

Despite the limitations, this study enhances our understanding of the associated health risks of IPV among US Black women through the use of a very extensive and detailed national sample of mental health disorders.³⁴ We, further, used structured clinical health assessment measures, providing more valid assessment of DSM-IV disorders. Such assessment also made it possible to evaluate associations between IPV and eating disorders, which have rarely been explored in empirical studies, especially among US Black women.

Although there are benefits to using the highly structured, lay interviewer–administered Composite International Diagnostic Interview to examine the mental health of abused women, the complexity of the Composite International Diagnostic Interview questions and differences in both cross-cultural interpretation of concepts and in expressions of distress may influence the study findings.^37,60,61 Finally, this is one of the first studies in which a national sample was used to evaluate the health differences among abused women of different Black ethnic groups.

Our findings provide useful information for conceptualizing the factors associated with IPV among varied Black cultural groups in the US population, particularly among abused Caribbean women, who remain understudied. Moving forward in our understanding of the nature and health effects of IPV on women of color, it is important to examine the role of culture and nativity, context, and migratory processes as well as what effects these factors may have on women’s outcomes.

As demonstrated in other studies³⁵ and in this study, abused women may differ in health outcomes as a result of varying backgrounds, culture, and experiences. A thorough understanding of these processes may be essential to developing effective prevention and treatment approaches and ultimately may assist in reducing abusive behavior.