| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ORIGINAL ARTICLES |
From the Alcohol Research Group (E.A.L.), Berkeley, California; and Institute for Quality, Research, and Training (W.A.V.), Robert Wood Johnson Medical School–University of Medicine and Dentistry of New Jersey, New Brunswick, New Jersey.
Address reprint requests to: E. Anne Lown, DrPH, 2000 Hearst Ave., Suite 300, Berkeley, CA 94709-2167. Email: ALown{at}arg.org
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
METHODS: Participants were women (N = 1155) with current male partners enrolled in a household survey of 3012 Mexican-origin adults, ages 18 to 59 years, living in urban, town, and rural areas of Fresno County, California. Crude and adjusted odds ratios (ORs) were calculated for four self-assessed health measures, seven chronic diseases, and 32 somatic symptoms.
RESULTS: In multivariate analyses, women reporting previous-year physical or sexual IPV were more likely to report 1) fair/poor overall health (OR, 1.9; confidence interval [CI], 1.0–3.7), physical health (OR, 2.1; CI, 1.2–3.9), and mental health (OR, 3.4; CI, 1.9–6.1), as well as worse comparative health (OR, 4.4; CI, 2.3–8.3); 2) a history of heart problems (OR, 17.0; CI, 4.3–66.7); 3) persistent health problems (OR, 3.3; CI, 1.5–7.0); and 4) numerous somatic symptoms.
CONCLUSIONS: Physical or sexual IPV was associated with poorer self-assessed health and many health symptoms among this culturally distinctive Mexican American population.
Key Words: domestic violence • partner abuse • health status.
Abbreviations: CI = confidence interval; IPV = intimate partner violence; OR = odds ratio.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
Health problems among abused women are not limited to the acute consequences of abuse but include a variety of long-term health complaints. A significant body of research documents an association between stressful life events and the development of disease (6, 9–14). Consequences of specific stressful events such as physical and sexual abuse include physical health problems (6, 9, 10, 15–22), physical functioning limitations (17, 23, 24), more diagnostic tests and frequent surgeries (9, 25–28), and increased healthcare utilization (6, 11, 18, 29).
In addition to health complaints and healthcare services use, self-assessed health indicators have been used to evaluate general health status among abused women. In three previous studies victims of sexual assault rated their health more poorly than did nonvictims (10, 11, 30). Single measures of general health are powerful predictors of morbidity and mortality. People who rate their health as fair or poor have consistently higher rates of mortality (31–35), increased bed days, and more physician visits (34). A recent large study found that Hispanic women who rated their general health as fair or poor vs. excellent, very good, or good (adjusting for age) had more than twice the odds of death (OR, 2.2; CI, 1.9–2.6), a higher average number of bed days (23.8 vs.3.3 days), and an average of 11.6 vs. 3.6 physician visits during the previous year (34).
The study reported here addressed three limitations in the current research examining the association between IPV and women’s health status. First, research on medical health and IPV has been conducted primarily in medical settings with notable exceptions (11, 30). Second, the primary population-based studies examining self-assessed health and chronic health problems did so in relation to lifetime sexual assault (10) and lifetime criminal physical and or sexual assault (24). Third, cross-cultural comparisons on the health correlates of partner abuse have not been made using either medically based or population-based samples.
Mexican Americans represent a culturally distinct group from the general US population, and they are the largest Hispanic ethnic group in the United States with a population of about 7.5 million in California (36). The prevalence of IPV among Mexican American women is comparable to the rate for white women, though US-born women report higher rates of abuse than do immigrant women (37). Abused Mexican American women might be expected to have even more serious health problems because of their disproportionate representation among low socioeconomic persons in the United States, a factor that has been associated with health problems in general populations (38). In addition to economic stresses, Mexican-origin women living in the United States often experience social stresses related to their minority status and degree of acculturation. Many Mexican Americans experience a sense of relative deprivation (39) compared with the majority population. This sense of deprivation has also been associated with poorer health and higher mortality compared with other populations (40, 41). Previous research reported that Hispanics are 3.6 times more likely to report fair or poor health compared with whites, even after controlling for illness factors (42). Despite their poorer self-assessed health status, Mexican Americans tend to underutilize healthcare services (43). The combination of IPV, low healthcare utilization, and minority status with its associated economic and social stresses may lead to more serious health problems for Mexican American women compared with abused women in general.
To our knowledge, no study has examined the correlates of sexual or physical abuse by intimate partners in a general population sample. The present study uses a large population sample of culturally distinctive Mexican-origin women to examine the association between recent physical or sexual IPV and self-assessed health status, chronic health conditions, and somatic symptoms.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
Data Collection and Measures
Data were collected by trained lay interviewers in participants’ homes using a computer-assisted personal interview system. Human subjects approval was obtained from the University of California. All participants gave informed consent. The interview was available in either English or Spanish and took approximately 1 hour to complete.
Three sets of health indicators were collected: 1) self-assessed health status, 2) chronic health conditions, and 3) somatic symptoms. Health status was measured using four self-assessed health measures. "General health" was measured by asking "In general, would you say your health is excellent, very good, good, fair, or poor?" Responses were dichotomized into "excellent, very good, and good" and "fair or poor." Responses grouped as fair or poor are referred to as poor self-assessed general health. This measure has widely accepted validity (32, 34, 45) and has a demonstrated association with mortality (31, 32, 34, 46). It primarily reflects physical health problems (including physical functioning limitations and chronic and acute conditions) and mental health to a lesser degree (47).
Two additional questions rated physical and mental health: "How would you rate your overall physical health?" and "How would you rate your overall emotional or mental health?" Responses were grouped as above. The physical health question was developed for this survey to differentiate between mental and physical health. The mental health question was adapted from the National Comorbidity Survey (NCS) (48). Comparative health was assessed by a NCS question modified for use here, ie, "Would you say your overall health is better or worse than other people your age?" "Better" or "same as others" was combined and contrasted with "worse than others."
Chronic health conditions were assessed for seven conditions: respiratory problems, diabetes or high blood sugar, heart attack or other serious heart trouble, high blood pressure or hypertension, arthritis or other severe bone or joint diseases, visual or hearing impairment, and stomach or gall bladder trouble. Responses were coded "yes" for each problem experienced during the previous 12 months. These questions, previously used in the NCS and the Los Angeles Epidemiologic Catchment Area (ECA-LA) study, were selected for comparability with ECA-LA data examining sexual assault and physical health in a multiethnic population.
Somatic symptoms were assessed using the Composite International Diagnostic Interview (CIDI) based on DSM-III-R (49). The CIDI was developed by the World Health Organization and the former US Alcohol, Drug Abuse, and Mental Health Administration for use in large international psychiatric epidemiologic research. It uses core questions from the Diagnostic Interview Schedule with modifications to make it more appropriate for cross-cultural populations with various levels of literacy and education (50). The CIDI uses a screener for a diagnosis of somatization: "Did you persistently have a lot of physical problems or illnesses over a period of several years?" The 95 women who answered "yes" were asked additional follow-up questions about 32 health symptoms that are used to diagnose somatization disorder (51). Women who had both persistent health problems and a given symptom were scored positive for that somatic symptom. Women with persistent health problems who denied having had a given symptom or women without persistent problems were scored negative for that symptom. A formal diagnosis of somatization was beyond the scope of this assessment.
Physical or sexual IPV by a current male partner during the last 12 months was measured by the following questions, "Has your current spouse/partner ever pushed you, hit you with a fist, used a knife or gun, tried to choke or burn you?" and "Has your current spouse/partner ever forced you to have sex against your will?" These questions were adapted from the Abuse Assessment Screen (52) and the NCS. Those who responded "yes" to either were then asked if such an event had occurred in the previous 12 months. Women who reported either type of abuse in the previous 12 months were considered positive for IPV.
Other independent variables included age, socioeconomic status (measured by the greatest number of years of education attained), and health insurance status (any health insurance (public or private) vs. no health insurance). Age was included because younger women are more likely to report IPV and less likely to have health problems. Education, although not consistently associated with IPV, is associated with self-assessed health (34) and health outcomes. Health insurance status is a measure of access to care and consequently may impact perceived and actual health.
Analysis
All bivariate and logistic regression procedures were first calculated using the statistical software SPSS, version 7.5 (SPSS Inc, Chicago, IL). Weights were applied using SPSS to ensure comparability of the final sample to the actual distribution of county residents by urban, town, and rural residence; by block; and by household size. SUDAAN 6.5 (53) was used to adjust the standard errors for stratified and clustered sampling.
Bivariate frequencies were used to describe the prevalence of IPV and each health condition. Crude ORs, 95% CIs, and p values were determined using logistic regression of one row category and one response. Multivariate logistic regression models were used to describe differences between women reporting and not reporting particular health outcomes. Adjusted ORs and 95% CIs are reported.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
|
Self-Assessed Health
In bivariate analyses, two self-assessed health measures indicated significantly worse health among abused women (Table 2). About half of abused women reported poor mental health, compared with about one-quarter of nonabused women. Similarly, 26% of abused women reported health worse than others their age, compared with about 8% of nonabused women. General health and overall physical health were also poorer among abused women, although the differences were not statistically significant.
|
Chronic Health Conditions
In multivariate analysis, only one of seven health conditions was significantly associated with abuse (Table 3). Heart attack or heart problems were more common among women reporting abuse (3.9%) than among women not reporting abuse (0.5%). In multivariate analyses, the odds of reporting heart problems among abused women was 17.0. Abused women were only marginally more likely to report chronic stomach or gall bladder trouble. Although small cell sizes create large confidence intervals, the data imply a much higher odds ratio among those who were abused.
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
Two groups of studies have provided data that describe health problems among women reporting physical or sexual abuse. The first and largest group includes studies carried out in healthcare settings such as internal medicine, family practice, obstetrics/gynecology, and gastrointestinal clinics and emergency departments. These studies have documented 1) the high prevalence of abuse among medical-service seekers (2, 3, 9) and 2) the frequency of abuse histories in women who report certain medical symptoms (6, 9, 17, 21, 27, 54). Abused women in the current sample described types of symptoms similar to those described by abused women in research conducted among primarily white women in medical settings, such as gastrointestinal disorders, pain, shortness of breath, and sexual and reproductive symptoms.
A second group of publications using population-based samples examined health problems and health perceptions in women reporting sexual assault (10, 11, 23, 30, 55, 56) and physical or emotional abuse (19). For example, an analysis of 1610 women from the ECA-LA study reported that assaulted women were only half as likely to report excellent health as women with no history of sexual assault (10). In addition, women who reported sexual assault had higher odds of diabetes, arthritis, and physical disability as well as more somatic symptoms compared with nonassaulted women. In a second example, a meta-analysis of seven general-population studies (N = 10,001) found that sexual assault was associated with poor subjective health regardless of gender, ethnicity, or sample site (30).
Our finding of poor self-assessed general health among abused women is consistent with findings from previous studies on the effects of sexual assault in ethnically diverse populations (10, 11, 30). The importance of poor self-assessed health is highlighted by the strength of this measure in previous research in predicting poor health outcomes. In the present study, abused Mexican American women were more likely to report poor general health and thus may be at greater risk for morbidity and early mortality and may use more health services than nonabused women.
In the NCS, fair or poor self-assessed mental health was associated with almost three times the odds of mental health service use and with higher perceived mental health needs (48). Based on these findings, it is likely that abused women in our sample who were significantly more likely to report poor self-assessed mental health would have higher perceived need and higher mental health service utilization than nonabused women. However, Mexican Americans are known to be low mental health service users (57); thus, increased psychiatric symptomatology (which may benefit from mental health treatment) may not result in increased mental health service use. Barriers to accessing mental health care among women reporting IPV and mental health symptoms would be an important area for future research.
The study reported here found a much stronger association between IPV and "heart attacks or other serious heart trouble" than was apparent in the other ethnically diverse population-based study. Sexually abused women were marginally more likely to report chronic heart disease than nonabused women (adjusted OR, 1.59; p = .06) as well as cardiovascular somatic symptoms (10). Similarly, in a primarily white gastroenterology clinic sample, physically or sexually abused women were more likely to report pain around their hearts (OR,= 2.41; CI, 1.22–4.79) (9). Although this association has been found in two other surveys, the strength of the association in the present sample leads us to believe this symptom may be manifested more strongly among this Mexican American sample.
Complaints of "heart attack" are likely to be a manifestation of distress, including anxiety disorders or ataques de nervios (58), which is a cultural-bound syndrome and an expression of distress described among Puerto Ricans (59). Consistent with this picture, the majority (60%) of physically or sexually abused women reporting heart attacks or heart problems in the present study were below the age of 30, whereas only 20% of nonabused women reporting heart attacks were below this age. Women below age 30 are extremely unlikely to have had an actual heart attack. This hypothesis is consistent with the increased odds among abused women of reporting two cardiopulmonary somatic symptoms: chest pain and shortness of breath, which are commonly associated with anxiety disorders.
Findings on somatic symptoms may be limited because a screener for persistent health problems was used. Only women with self-reported persistent health problems were asked questions about 32 somatic symptoms. Use of this screener may have kept us from capturing the true extent of these health symptoms, which are certain to exist in some women who do not describe themselves as having persistent health problems. As a result, this measure probably conservatively estimates the true association between these symptoms and abuse.
Somatic symptoms were widely reported by abused women in the present study. This is consistent with previous research (6, 9, 10, 12, 17). Replicating the findings of previous population-based (10, 13, 56) and clinic-based studies (21, 60), the current research found that this population of Mexican American abused women reported numerous reproductive symptoms, sexual symptoms (16, 26, 61), and pain (9, 17, 62, 63). Differing from some earlier studies, cardiovascular symptoms and some neurological symptoms (such as urinary retention and amnesia) were reported by the abused Mexican American women in this sample. These complaints may also be characteristic of ataques de nervios. Heart problems, amnesia, and paralysis are common among people reporting ataques de nervios. However, other symptoms common to this syndrome were not significantly more common among abused women, including seizures, fainting, palpitations, and dizziness.
It is possible that symptoms found to be strongly associated with IPV in the present study but not in previous studies are culturally specific manifestations of stress. However, findings would need to be replicated in other studies of Mexican American women to confirm these explanations, and questions would need to assess ataques de nervios directly.
Strengths and Limitations
Given the population-based design of this study, we relied on self-report for chronic medical problems and health symptoms. Some women may have been unaware of certain health problems or misunderstood their diagnoses. However, the health problems described by this population are likely to correspond with presenting complaints in a medical setting, which can form the basis for diagnostic tests and treatment. Recognition of abuse as an underlying contributor to a variety of health conditions may be important in the diagnosis and treatment of women.
Our use of a brief measure to assess abuse may have resulted in an undercount of the actual number of abused women. Repeated questions tend to yield higher prevalence rates for abuse, suggesting that increased trust and behaviorally specific questions make admission of abuse easier (64, 65). Our measure might underestimate the true association between abuse and poor health.
The cross-sectional design of the study does not allow for assessment of temporal order of health and abuse conditions. A history of childhood abuse has been associated with partner abuse as well as later medical problems (9, 66). Although we did not measure (and therefore control for) childhood abuse, the association between partner abuse and a spectrum of medical problem indicators suggests a robust effect for partner abuse. Some of the statistically significant findings may have been due to chance given the multiple comparisons made. Hence, caution should be exercised in interpreting the results, and replication of these findings in other population-based studies is important.
Despite these limitations, IPV was consistently associated with poorer self-assessed health status, chronic heart diseases, and many somatic symptoms in a culturally distinctive Mexican American population. Though the prevalence of abuse in a general population sample of women is significantly lower than that reported in medical settings, it seems that women who report abuse in either setting are likely to report poorer self-assessed health and numerous health complaints. Health complaints among abused Mexican American women seem to be consistent with the complaints of ethnically diverse populations of abused women. However, Mexican American women also may have a culturally specific response to abuse, with a higher likelihood of reporting heart problems and chest pain, amnesia, and paralysis, along with more typical somatic symptoms such as gastrointestinal, neurological, sexual, and reproductive problems.
Abused women clearly suffer a disproportionate burden of disease compared with nonabused women. However, it is unclear whether healthcare utilization by abused Mexican American women in a general population sample is higher than that of their nonabused counterparts. Future studies should examine whether abused women with poorer health are able to access health care.
Increased attention should be paid in medical settings to the identification and prevention of abuse among Mexican American women as well as among other ethnic groups. Knowledge of the association between abuse and certain health complaints can alert healthcare providers to inquire about and address issues related to IPV before considering expensive, invasive, or painful diagnostic tests. Future research might examine medical utilization patterns among a population of abused Mexican American women. Lastly, policy makers should consider the importance of protection for this population, including the provision of culturally appropriate medical and social services, available shelters, and legal services.
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
Received for publication April 17, 2000.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  S. Sanchez-Lorente, C. Blasco-Ros, C. L. Coe, and M. Martinez Recovery of Immune Control Over Herpes Simplex Virus Type 1 in Female Victims of Intimate Partner Violence Psychosom Med, January 1, 2010; 72(1): 97 - 106. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Nicolaidis, B. McFarland, M. Curry, and M. Gerrity Differences in Physical and Mental Health Symptoms and Mental Health Utilization Associated With Intimate-Partner Violence Versus Childhood Abuse Psychosomatics, July 1, 2009; 50(4): 340 - 346. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Prospero Mental Health Symptoms Among Male Victims of Partner Violence American Journal of Men's Health, December 1, 2007; 1(4): 269 - 277. [Abstract] [PDF]
|
|
|
|
 |
|
 R. Caetano, S. Ramisetty-Mikler, P. A. Caetano Vaeth, and T. R. Harris Acculturation Stress, Drinking, and Intimate Partner Violence Among Hispanic Couples in the U.S. J Interpers Violence, November 1, 2007; 22(11): 1431 - 1447. [Abstract] [PDF]
|
|
|
|
 |
|
 D. Holmes, G. W. Alpers, T. Ismailji, C. Classen, T. Wales, V. Cheasty, A. Miller, and C. Koopman Cognitive and Emotional Processing in Narratives of Women Abused by Intimate Partners Violence Against Women, November 1, 2007; 13(11): 1192 - 1205. [Abstract] [PDF]
|
|
|
|
 |
|
 I. Ruiz-Perez, J. Plazaola-Castano, and M. del Rio-Lozano Physical health consequences of intimate partner violence in Spanish women Eur J Public Health, October 1, 2007; 17(5): 437 - 443. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Lipsky and R. Caetano Impact of Intimate Partner Violence on Unmet Need for Mental Health Care: Results From the NSDUH Psychiatr Serv, June 1, 2007; 58(6): 822 - 829. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Klevens An Overview of Intimate Partner Violence Among Latinos Violence Against Women, February 1, 2007; 13(2): 111 - 122. [Abstract] [PDF]
|
|
|
|
|
|
A. C. Denham, P. Y. Frasier, E. G. Hooten, L. Belton, W. Newton, P. Gonzalez, M. Begum, and M. K. Campbell Intimate Partner Violence Among Latinas in Eastern North Carolina Violence Against Women, February 1, 2007; 13(2): 123 - 140. [Abstract] [PDF]
|
|
|
|
|
|
D. J. Whitaker, C. K. Baker, C. Pratt, E. Reed, S. Suri, C. Pavlos, B. J. Nagy, and J. Silverman A Network Model for Providing Culturally Competent Services for Intimate Partner Violence and Sexual Violence Violence Against Women, February 1, 2007; 13(2): 190 - 209. [Abstract] [PDF]
|
|
|
|
|
|
A. Interian, L. A. Allen, M. A. Gara, J. I. Escobar, and A. M. Diaz-Martinez Somatic Complaints in Primary Care: Further Examining the Validity of the Patient Health Questionnaire (PHQ-15) Psychosomatics, October 1, 2006; 47(5): 392 - 398. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Loxton, M. Schofield, R. Hussain, and G. Mishra History of Domestic Violence and Physical Health in Midlife Violence Against Women, August 1, 2006; 12(8): 715 - 731. [Abstract] [PDF]
|
|
|
|
|
|
M. A. Dutton, B. L. Green, S. I. Kaltman, D. M. Roesch, T. A. Zeffiro, and E. D. Krause Intimate Partner Violence, PTSD, and Adverse Health Outcomes J Interpers Violence, July 1, 2006; 21(7): 955 - 968. [Abstract] [PDF]
|
|
|
|
|
|
R. J. Waldinger, M. S. Schulz, A. J. Barsky, and D. K. Ahern Mapping the Road From Childhood Trauma to Adult Somatization: The Role of Attachment Psychosom Med, January 1, 2006; 68(1): 129 - 135. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. Ruiz-Perez and J. Plazaola-Castano Intimate Partner Violence and Mental Health Consequences in Women Attending Family Practice in Spain Psychosom Med, September 1, 2005; 67(5): 791 - 797. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. B. Plichta Intimate Partner Violence and Physical Health Consequences: Policy and Practice Implications J Interpers Violence, November 1, 2004; 19(11): 1296 - 1323. [Abstract] [PDF]
|
|
|
|
|
|
M. I. Garcia-Linares, S. Sanchez-Lorente, C. L. Coe, and M. Martinez Intimate Male Partner Violence Impairs Immune Control Over Herpes Simplex Virus Type 1 in Physically and Psychologically Abused Women Psychosom Med, November 1, 2004; 66(6): 965 - 972. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. H. Porcerelli, R. Cogan, P. P. West, E. A. Rose, D. Lambrecht, K. E. Wilson, R. K. Severson, and D. Karana Violent Victimization of Women and Men: Physical and Psychiatric Symptoms J Am Board Fam Med, January 1, 2003; 16(1): 32 - 39. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
