|Home | About | Journals | Submit | Contact Us | Français|
A growing literature has shown a high prevalence of intimate partner violence (IPV) in Latina women (lifetime: 21% to 35%;1–3 past-year: 4% to 33%); 2–7 and increased prevalence of substance abuse,4, 8 injury,9 HIV infection,6 depressive symptoms,10–12 posttraumatic stress disorder,2, 11, 13 and poor physical and mental health5, 14 in Latina women with abuse histories. However, prior studies did not delineate the period-prevalence of IPV over multiple time periods (lifetime, past 5 years and past year) in a single population of Latina women. Moreover, with the exception of one study showing higher rates of suicidal ideation in abused Latina compared to abused non-Latina women,14 studies did not compare the health of abused Latina women to the health of abused non-Latina women using multiple health indicators. The present investigation examined IPV period prevalence (lifetime, past 5 years and past year) and the association between lifetime IPV exposure and multiple health indicators among Latina and non-Latina women.
The study was approved by the institutional review board of Group Health Cooperative, a large health care delivery system in the Pacific Northwest U.S. The study population comprised English-speaking women ages 18 to 64, randomly sampled from enrollment files to participate in a telephone survey to assess IPV and health status.15, 16 A letter was mailed to women describing the study’s focus on women’s health issues, followed by telephone contact to ascertain interest and consent to participate.15, 16 Details of the study protocol, including safety procedures, were published previously.15, 16
Of 6,666 women sampled, 345 were excluded because they: did not meet the sampling criteria (209); were deceased (3); were too ill (15); or did not speak English or had a hearing impairment (118). Of the 6,321 remaining women, 1829 (28.9 percent) refused participation when initially contacted, 539 (8.5 percent) started but did not complete the interview, 385 (6.1 percent) could not be located, and 3,568 (56.4 percent) completed the interview. A propensity score analysis showed that the probability of participation was similar for women with and without an IPV history.17
Of the 3,568 survey respondents, 139 women were excluded because they never had an intimate partner, and 3 women were excluded for not responding to the question on Hispanic ethnicity, reducing the analytic sample to 3,426. Consistent with the western Washington metropolitan area,15 4% (n=139) of the sample reported Hispanic ethnicity.
Women were first asked about their health and then about their IPV history.16
Women reported on their physical, social and psychological well-being by responding to 20 questions from the Short Form-36 (SF-36) Health Survey, version 2.18 These 20 questions were used to create four of the eight SF-36 validated subscales (vitality, mental health, emotional functioning, and social functioning in the past four weeks) and two overall health component summaries (physical component and mental component).16 One question from the SF-36 was used to assess women’s general health.16 The SF-36 subscale scores and the physical component summary (PCS) and mental component (MCS) continuous scores were standardized to have a mean of 50 and standard deviation of 10, with higher scores indicating better functioning; these standardized scores allow for easy comparisons across subscales and clinical populations.18 The general health item was dichotomized (fair/poor versus good/very good/excellent health).19
Women rated the frequency of depressive symptoms (0=less than 1 day per week to 3=five or more days per week) using five validated questions from the 20-item Center for Epidemiological Studies-Depression (CES-D) scale.20, 21 Scores for each of the five items were summed and the summary score was dichotomized to categorize women according to their depressive symptom status; a summary score of four or higher (range, 0–15) indicated minor depressive symptoms, and six or higher indicated severe depressive symptoms.20
Using questions from the National Institute of Mental Health Diagnostic Interview Schedule, women indicated how frequently they were bothered by 14 common physical symptoms in the past six months (range, 1=none of the time to 5=all of the time).22 We estimated the mean number of symptoms experienced by women at least “some of the time.”
Women were asked about their age, household income, employment status, educational level, and number of children living in the home using questions from the U.S. Census Bureau.23 Women were asked about their history of childhood physical abuse (“before you were 18, was there any time when you were punched, kicked, choked or received at more serious physical punishment from a parent or other adult guardian”) and childhood sexual abuse (“before you were 18, did anyone ever touch you in a sexual place or make you touch them when you did not want them to”) using two questions from the Behavioral Risk Factor Surveillance System (BRFSS).15
IPV victimization since age 18 was assessed using the Women’s Experience with Battering (WEB) Scale, and 5 questions from the BRFSS on physical (1 question), sexual (2 questions) and psychological abuse (2 questions) (Table 1).24, 25 The 10-item WEB Scale was designed to ascertain women’s experience of loss of power and control in relation to an abusive partner. To minimize respondent burden, the WEB questions were asked for women’s three most recent heterosexual or homosexual intimate partners, including their current partner.15 Women who scored 20 or higher on the WEB (score range, 10 to 60) for any given partner were considered positive for abuse.25 After women completed the WEB for each partner, they were asked what year and month they first started and stopped feeling this way about their partner; this information was used to construct past-year and past 5-year abuse exposure according to the WEB.
To establish period-prevalence for each type of IPV assessed by the BRFSS questions, women were first asked if they ever experienced each particular abuse type since age 18; if they had, they were asked if the abuse occurred during the past 5 years, and during the past year. To further comment on the type of abuse women reported, we defined two categories of abuse based on the BRFSS questions. Women were defined as having experienced “physical IPV” if they reported physical and/or sexual abuse, and they were defined as having experienced “psychological IPV” if they reported threats and/or controlling behavior. The data were collapsed into these two broad abuse categories (physical and psychological) in order to provide meaningful estimates due to the small number of Latina women.
Chi-square tests were used to compare the demographic characteristics of Latina versus non-Latina women. IPV prevalence (including 95% confidence intervals) was estimated for Latina and non-Latina women for lifetime, past 5 years and past year. Lifetime prevalence was estimated using the BRFSS questions only (since the WEB questions were only asked with regards to the three most recent partners), and past-year and past 5-year IPV estimated using both the BRFSS and WEB questions. Prevalence estimates with relative standard errors (RSE) over 30% are considered unstable. Although estimates with RSE > 30% are reported, unstable estimates are marked with an asterisk; caution should be exercised when referring to these estimates.
Multivariable models included indicator variables for the main effects of IPV exposure and Hispanic ethnicity and their interaction term to allow estimation of the relationship between lifetime IPV and current health separately for Latina and non-Latina women. In these models, the exposed group included women with any IPV since age 18 according to the BRFSS or WEB questions and the unexposed (reference) group comprised women without such histories. Generalized linear models with a log link were used to obtain prevalence ratios (PR) for dichotomous health indicators for women with a lifetime IPV history compared to women without a lifetime IPV history. Multivariable ordinary least squares regression was used to estimate mean differences in SF-36 scores and number of symptoms. Models were adjusted for age and income, factors, which could confound the relationship between IPV history and health.
Compared to non-Latina women, Latina women tended to have lower household income (50.4% versus 38.0% reported annual income of less than $50,000), were less likely to have completed at least some college (77.7% versus 87.9%), were younger (30.2% versus 17.9% were less than age 35), and were more likely to have experienced physical or sexual child abuse (47.8% versus 33.2%) (Table 2).
As ascertained by the BRFSS questions, lifetime IPV prevalence was comparable in Latina (44.6%) and non-Latina (44.0%) women (Table 3). Similar lifetime prevalence rates were observed for both physical IPV (38.8% vs. 34.0%) and for psychological IPV (37.4% vs. 35.3%) According to the BRFSS and WEB questions, IPV prevalence tended to be higher for Latina versus non-Latina women in the past 5 years (20.1% vs. 14.5%, p=0.06) and past year (11.5% vs. 7.8%, p=0.11), but the differences were not statistically significant. The exception was for physical abuse within the past 5-years, which was significantly higher among Latina women (11.5% vs. 4.9%, p<.05).
In adjusted models, women who reported any exposure to IPV (lifetime BRFSS or WEB for any of 3 most recent partners) reported worse health compared to non-abused women (Table 4).
Short Form-36 subscale scores for abused Latina women ranged from a mean of 5.62 (mental health) to 7.77 (vitality) points lower than scores for non-abused Latina women, and the Mental Component Summary score was 7.52 points lower for abused Latina women versus non-abused Latina women. Moreover, Latina women with a lifetime IPV history had significantly more physical symptoms, depression prevalence more than twice that of non-abused Latina women, and were more likely to report distrust of people in their residential community (prevalence ratio: 1.84).
Non-Latina women with IPV histories also had worse health across many indicators compared to non-abused women, but the differences were not as pronounced. For example, for non-Latina women, SF-36 scores ranged from a mean of 0.96 (PCS) to 3.87 (MCS) points lower for women with abuse histories compared to non-abused women.
The significance of the interaction term between Hispanic ethnicity and IPV exposure was tested, to determine if the association between IPV and health differed by ethnicity. Latina women suffered significantly more adverse IPV-related mental health issues compared to abused non-Latina women, in their overall mental health functioning (Mental Component Summary) (p<0.02) and the specific areas of vitality (p<0.01) and role emotional functioning (p<0.01) (last column, Table 4).
Similar lifetime IPV rates were found for Latina and non-Latina women. Rates of recent abuse (past year and past 5 years), however, tended to be more common in Latina versus non-Latina women, but the differences were not statistically significant. In models adjusted for race/ethnicity, women with a lifetime IPV history had compromised health compared to non-abused women. Adverse IPV-related mental health issues were more pronounced in Latina women.
IPV prevalence varies widely depending on how, where and when women are asked about abuse, how abuse is categorized, and according to the characteristics of women; these methodological considerations constrain cross-study comparisons. These limitations noted, the lifetime IPV rate we observed among Latina women (44.6%) was higher than the rate reported for roughly 250 Latina women (35%) recruited from community hospital emergency departments,3 and was within the range of lifetime IPV reported in a survey of 292 Latina women receiving community health services (33.9% physical assault; 20.9% sexual coercion; and 82.5% psychological aggression).26 The rate of past-year IPV in Latina women (11.5%) was consistent with prior estimates (4% to 33%),2–7 particularly one study which showed past-year physical or sexual IPV of 10.8%.5
The results corroborate prior findings of higher rates of depression, 10–12 and poor mental health and somatic symptoms5,12 among Latina women with IPV histories. For example, Hazen found a significant relationship between depression and physical assault and psychological maltreatment and between somatization and emotional abuse in Latina women.12 The results also confirm poor overall health associated with IPV in women in general.16, 25, 27–29
In addition, findings suggested that abused Latina women had more compromised mental health than abused non-Latina women, in overall mental health functioning, vitality and emotional functioning—a significant addition to the sparse literature focused on IPV-related health in Latina compared to non-Latina women. In one other study, suicidal ideation tended to be more common in abused Latinas compared to other abused women.14 While Latina participants had access to mental health services through their insurance, they may perceive greater barriers to accessing these services.30–32
Caution should be used in generalizing the findings because of the small number of Latina women, the inability to assess subgroups and acculturation status of Latinas, the insured nature of the sample, and the data collection method. While telephone surveys are widely used to assess violence and health,33–35 it is possible that the most severely abused women do not participate in such surveys because they are isolated.36 Our response rate was low; however, a propensity score analysis showed the likelihood of response was similar for women with and without IPV histories.17 The cross-sectional nature of the study precluded statements about causality.
The present study provides information to warrant larger investigations of abuse in Latina women, and supports the ongoing case for screening and intervention development in women with abuse histories.37 We recommend screening Latina women for abuse in health care settings, particularly those presenting with mental health concerns, and the development of abuse interventions that are sensitive to the cultural needs of Latinas.38–48
Amy E. Bonomi, Department of Human Development and Family Science, Ohio State University, Columbus, Ohio.
Melissa L. Anderson, The Center for Health Studies, Group Health Cooperative, Seattle, Washington.
Elizabeth A. Cannon, Department of Human Development and Family Science, Ohio State University, Columbus, Ohio.
Natasha Slesnick, Department of Human Development and Family Science, Ohio State University, Columbus, Ohio.
Michael A. Rodríguez, Department of Family Medicine, David Geffen School of Medicine at UCLA.